sSmCStdP

1Introduction Development economists use data extensively, so it is important to understand how such data are constructed and thereby the limitations of such data when used in empirical work. If the data are faulty, the empirical work cannot be sound no matter how sophisticated the techniques utilized. In other words, the principle of “garbage in–garbage out” applies. The first section of this chapter explores various data collection methods and reflects on how these data might be used in empirical work. In so doing, some of the issues raised in Chapter 1 on the discussion of method utilized by the different approaches are elaborated on. These discussions continue into the second section in which recent macro and micro development economics topics are picked as case studies to elaborate on the use of alternative kinds of data in empirical analysis.
2
3Alternative forms of data collection There are many kinds of data depending on the form of data collection, and each has its own strengths and weaknesses.
4
5Cross-section (secondary) The comparative statistics reported in the Appendix Tables in Chapter 3 represent an example of such data. These data are called cross-sectional because all the relevant units for a given variable, such as primary school enrollment rates, are observed at one point in time (in this case some specified year). They are called secondary because the researchers are not collecting the data themselves but are relying on some agency to procure the relevant data. Secondary data are usually collected by national government agencies.     The compilation of primary school enrollment data is based on all schools informing the relevant government statistical department officials of pupils enrolled in a given year in their primary schools. Generally sub-national (say state or provincial) statistical bureaus gather such information forwarded by lower government tiers (say district) and in turn forward these to national agencies (say federal). 34 Background
6
7   Various ministries like health and education collect and compile these data for national use but also pass them on to the various concerned international agencies. Thus education data are passed on to UNESCO (United Nations Educational, Scientific and Cultural Organization) and health data to WHO (World Health Organization). These international agencies in turn compile international comparative statistical compendiums. Thus part of the World Health Statistics reported every year rely on routine reporting from its 194 member states.    The World Development Indicators, compiled by the World Bank, are the largest current source of development statistics. The World Bank relies on credible international organizations such as the WHO. While international organizations have rigorous standards and work with national statistical bureaus to improve the quality of data compiled, they are ultimately dependent on member governments for providing the secondary data. Inaccuracies in the original source carry through the chain of data compilation reported at an international level.    Errors could creep into the national data due to systematic reporting biases. Consider primary school statistics as an example. If school resources at the local level depend on enrollments, there is an incentive to overstate them. Local governments may also want to overstate reported enrollment rates, especially if they are under pressure to enhance school access by the provincial or state governments.    Countries are ranked internationally based on social and economic performance indicators and so there is an incentive at the national level to overstate achievements. Nations also have an eye on domestic constituents when doing this, particularly when the governments are democratic. The press picks up and reports on the lack of comparative social and economic progress and civil society organizations often base their advocacy campaign on such reporting.
8
9Time series (secondary) If the data reported on above are collected for each year, in the course of time a substantial time series develops. In January 2019, the World Development Indicators reported data on 1,600 variables for 264 countries or country groups going back to 1960.1 For more specialized data such as finance one would consult the International Monetary Fund (IMF) statistics, or for international trade and foreign direct investment (FDI) the United Nations Conference on Trade and Development (UNCTAD), for labor statistics the International Labor Organization (ILO), for industrial statistics the United Nations Industrial Development Organizations (UNIDO), for agricultural statistics the Food and Agricultural Organization (FAO), and as mentioned earlier the WHO for health statistics and UNESCO for education statistics.     In addition to problems pertaining to cross-sectional secondary data the time series in such data could be subject to additional errors. For example, if variable definitions change over time, the data are not comparable (e.g. changes in the definition of literacy) unless the agency concerned revises the earlier reported data, which would be a very time-consuming and expensive endeavor and even                                       Data and their use in development economics 35
10
11then would only represent guesstimates. If definitions become more stringent over time, there would be a systematic understatement of reported progress.
12
13Cross-sectional (primary) quantitative These data are collected by researchers based on special purpose sample surveys designed with specific research questions or hypotheses testing in mind. Generally the study is designed (probability sampling) so that a relatively small sample can be used to generalize the results to the population of interest. Good probability sampling requires starting with a good sampling frame from which the selection is made, and that is the first challenge.     Errors can creep into the data for various reasons other than a poor sampling frame. Many times, the data collection is based on structured questionnaires produced far away (often in a different country) from the research site. The researcher in this case presupposes that the information they plan to gather exhausts the relevant social information on the issue being researched. If there are limitations to the conceptual understanding of the researcher in terms of their perception of social and economic reality at the site in question, the closed ended questions designed would have limitations.     Even if questionnaire limitations are ruled out, the research design could be faulty. The errors could pertain to coverage or sample design. In this case, the sampling errors may be very large and the information unreliable. It is expensive to increase the sample size to reduce sampling errors. Further, to get reliable data the field interviewers need to be very well trained and highly motivated to withstand the hardships of fieldwork often far removed from the basic amenities (clean eating, living and drinking) that they take for granted. Getting such a team in place for short-term assignments is a difficult task. Without such a team, errors compound in the field. Unless the field team has a very good understanding of the content of the study, their ability to secure the desired information will be wanting.     Even if there is an excellent study and sample design and field team in place, errors are still inevitable. Recall errors are more likely if the sample population is illiterate and do not maintain records. Rough and ready answers might be provided so they can get back to their lives. A badly trained field team might prompt them to give desired answers or the interviewer might glean what the desired answer is and seek approval by responding in a way that they sense is desired. If they have a reason not to provide information that they think might harm them in some way, inaccuracies are inevitable. For example, they may perceive information on income or assets might be passed on to tax authorities or competitors. Even if all this is not a problem, errors can compound during data processing (data entry and cleaning).
14
15Panel (primary – cross-sectional/time series) Panel data are rare and entail going back to the same households and collecting data over a period of time, sometimes decades. As expected, these are enormously 36 Background
16
17expensive studies (millions of dollars). The most accurate information is possible from these if well conducted. While many additional research questions can be addressed with panel data, they still are sample survey studies and hence subject to the same shortcomings as primary cross-sectional data.
18
19Census A census by definition means interviewing every household in the population and hence this is enormously expensive and so can only be done periodically.2 Generally, if the whole nation is the relevant population, they occur every ten years or so if possible.3 Various questions are sometimes appended to the population survey pertaining to shelter, health and education, but adding more questions adds to the cost across the board.     Even though all the households of a population are supposed to be interviewed, the poorest and richest are the ones most likely to be missed out. The poorest because they may be itinerant and the richest because they live in gated abodes and refuse to concede the time needed for the interviews if present at the time of the survey. Successful surveys require training and managing an army of interviewers and, for example in the United States, the monthly employment rate temporarily ticks up for the duration of the survey. Charges of cooked up responses, sitting far from the survey sites, are mostly likely for poorly trained and poorly resourced government surveys in L/LMICs.
20
21Cross-sectional (primary) qualitative To avoid the problems of quantitative survey data, social scientists have developed alternative methods to collect and analyze qualitative data. These techniques use quasi-anthropological or ethnographic approaches. This entails staying a few days or weeks at the site rather than a few years as in the case of cultural anthropology.    Apart from the problems alluded to above, proponents of qualitative methods view sample surveys as extractive with researchers taking something to build their careers without giving anything back.4 The purists who developed this method, like Chambers (1997), view it as a process of metaphorically handing the stick to the poor and empowering them to understand their own reality and change it themselves.    Numerous techniques that have evolved include transect walks for location familiarization, village mapping exercises for ice-breakers, and focus group discussions (FGDs) to hone information via mutual vetting. FGDs are very carefully managed to avoid dominance by the elites or males in mixed FGDs. Often separate FGDs are held to ensure that women, minorities and the poorest have a voice. Even then the coordinator has to ensure that more or less everyone gets a chance to speak. There are some issues that people prefer not to discuss in a group and a household instrument may be better suited for that. In addition, key informant (depending on the subject) interviews are held. The idea is that triangulation (cross-checking) via several sources of information will ensure accuracy.5                                       Data and their use in development economics 37
22
23   No data collection method can avoid problems. In participatory research, based on the Heisenberg uncertainty principle, social and economic reality can change in the process of studying it. Response bias might result if a good rapport is established, as required, between the researchers and respondents. The respondents may be more inclined to give the answers they sense the researchers want to hear. Another problem is that, on the one hand, the time spent in the field is not nearly long enough to understand the social and economic dynamics according to cultural anthropologists, the sharpest critics of such qualitative research, but, on the other, it is too much to cover many sites and so by design the findings cannot be generalized.6 Even so, quantitative and qualitative research methods can be viewed as complementary to enhance triangulation.
24
25Random control trials (primary) Bannerji and Duflo (2011) have done the most to popularize this method in development economics partly to address a selection bias problem (see next section). Abdul Latif Jameel Poverty Action Lab (J-PAL) was founded by them along with Sendhil Mullainathan in 2005 at the MIT Department of Economics for Policy-Oriented Poverty Research in L/LMICs. It has grown into a network of researchers and, by 2016, included 143 affiliated researchers at 49 universities.    The method is borrowed from epidemiological (medical) research whereby one group of patients gets a treatment and the control group, with the same characteristics, gets a placebo. The theory is that since the two groups are otherwise identical, it is possible to tease out the impact of the treatment. Policy issues explored by RCTs include the impact of microcredit, treated bed nets, de-worming, cash incentives for appropriate behavioral responses and better communications strategies to provide important information.
26
27Illustrating data use The various kinds of data described above are extensively used in development economics in exploring both macroeconomic and microeconomic topics. To illustrate the use of these data, I have picked premature deindustrialization as a macroeconomic topic that has gained traction since the turn of the 20th century and microcredit as a microeconomic topic that was popularized in the 1980s. For a while the latter was considered an anti-poverty panacea and it is still part of the poverty alleviation tools advocated by development policy organizations like the World Bank.     Recall the concept of structural change in L/LMICs introduced in Chapter 1. One of the “stylized facts” of economic development identified by Chenery and his research associates was the increase in the share of the secondary sector (industry) in GDP and employment and a concomitant decline in the GDP share of the primary sector (mainly agriculture). As catch-up growth continues and countries 38 Background
28
29attain HIC status, the primary sector shares continue to shrink, industry shares also shrink (deindustrialization), and a sophisticated tertiary sector (services) expands.     Since the last decade of the 20th century and into the 21st century, development economists have noticed the setting in of deindustrialization even in most L/ LMICs. While the tertiary sector has expanded, this is mostly as a process of informalization of the economy (Chapter 1). Since the deindustrialization has set in at much lower levels of PCGDP than historical trends, as identified by Chenery and his associates, development economists have dubbed this macroeconomic phenomenon premature deindustrialization.     Development economists have empirically and theoretically addressed this puzzle of premature deindustrialization and the findings are discussed in detail in Chapter 9. Here, the various stories7 that emerged from that body of research are reviewed to illustrate how various kinds of data have and can be utilized to shed light on this puzzle of premature deindustrialization.     The World Bank, IMF, WTO led neoliberal structural reforms (Chapter 8) are one among several stories for premature deindustrialization. While the details are discussed in Chapters 8 and 9, the summary is that these organizations pushed aggressive trade liberalization and that without protective tariffs L/LMIC fledgling industries were decimated.     Again, these organizations aggressively opened up L/LMICs to capital inflows. The FDI of multinational corporations (MNCs) displaced local industries and portfolio investment caused temporarily booming stock and real estate markets that diverted resources from productive investments. Economic instability and recessions were exaggerated from the herd-like disappearance of foreign capital when profit opportunities shrank (for example, the Asian financial crisis of 1997). To sum up, the story here is that induced economic globalization in the form of inter-linkages between HICs and L/LMICs induced premature deindustrialization in the latter.     Another associated story is that neoliberal reforms in L/LMICs resulted in income inequality and that this resulted in premature deindustrialization by inducing consumption pattern changes. High income groups prefer luxury imports they now had access to due to neoliberal enforced trade liberalization. Demand for mass produced manufactured goods among lower income groups accordingly shrank as their income share was squeezed. This was reinforced by rising income inequality in HICs that resulted in less demand for the cheaper labor intensive manufactured imports from L/LMICs.     The obvious question is why L/LMICs would open themselves up to self-harm. The answer is complex and addressing this issue using several different approaches is one of the themes of this textbook. Summary answers are that the organizations identified above have leverage. This leverage increases when L/LMICs need fiscal or balance of payment assistance because they have mismanaged their economies. Loans come with conditions and heterodox economists believe these conditions to be destructive, including to local industry. A political economic explanation is that elites may gain even as the general population loses, and so they have an incentive                                       Data and their use in development economics 39
30
31to accept loans, FDI, portfolio investment and trade policy that is more generally harmful to the population.     Other stories are more technocratic and refer simply to the working of global market mechanisms. One strand in this literature is that global commodity prices rose at the turn of the 20th century for over a dozen years and this resulted in resources being reallocated to the primary sector relative to the secondary sector in many L/LMICs, particularly in Latin America and Sub-Saharan Africa where premature deindustrialization is more acute. Another strand is that higher productivity growth in HICs due to automation resulted in falling prices of manufactured goods and so L/LMICs could not compete. As told, this technocratic story excludes the possibility of market power or predatory pricing focused on by political economists.     Finally, another story pertains to the “China effect”. The argument here is that China with its huge labor force and rapid technological progress and productivity growth was able to bring down unit labor costs to levels that most L/LMICs were unable to compete with. Thus, the premature deindustrialization in L/LMICs in the last three decades (starting in the last decade of the 20th century) coincides with China taking over export markets of labor intensive manufactured goods.     The data discussed in the earlier section could be utilized with data analysis techniques to explore these stories and more than one story could be at play at the same time. All data have limitations as explained above and similarly data analysis methods have shortcomings. Here only the broad outlines of these shortcomings will be identified and courses in econometrics are required for more in-depth understanding.     Mainstream economists wedded to the deductive method prefer to use econometrics to test the association between variables for testing hypotheses. For now, consider each of the stories above as a hypothesis, with the structural adjustment story having several sub-hypotheses pertaining to the impact of loans, foreign aid, FDI, portfolio investment and trade policy on premature deindustrialization.     Thus premature deindustrialization is the dependent variable that needs to be explained. One option would be to utilize secondary data for cross-country analysis. In this case, each country is a unit of analysis or observation. If data are available for say 124 countries across all income categories for the dependent variable, and all the independent variables reflecting the different variables, then in principle one can estimate the statistical significance and the magnitude of association of these variables with premature deindustrialization. Engaging in econometric analysis requires being able to adequately define and measure variables. Constructing indices for variables such as for example trade policy are extremely challenging.     One major problem with such cross-country estimations that treat each country as a unit of observation is the implicit assumption that all countries in the sample are structurally similar. But as explained in detail in Chapter 1, if this were so, there would be no need for development economics. There are econometric methods for addressing this shortcoming, but none is entirely satisfactory. 40 Background
32
33    Another problem with such studies is that only countries for which data are available can be in the sample. Studies have shown that the results can dramatically change by adding a few more countries. Also, ruling out countries due to unavailable data limits the sample size and this too can change the estimates.8 One solution to the shrinking sample size is to pool secondary cross-section and time series data to increase the sample size. However, while the sample size is increased, the other problems remain.     Time series econometrics is another available tool and this utilizes time series data. In principle, such analysis bypasses having to assume structural similarity of countries of very different income levels, and so is preferable. Time series analysis requires an observation for each variable over a specified period of time (hour, day, month, quarter or year). Time series analysis has its own set of demands for the accuracy of estimates, and one of these is that it requires many observations for each variable. Since most country data are annual, this limits the number of observations that can be available. Also, observations on some variables, for example income inequality, are simply not available annually and so this limits the hypotheses that can be tested.     Ioannidis, Stanley and Doucouliagos (2017) explain why much published economics research is false. Most empirical research focuses on the significance of a statistical association between variables. Suppose the null hypothesis to be tested is that treated bed nets reduce the incidence of malaria. In this case, recall from your statistics courses that a type I error would be incorrectly concluding the null is false while type II would be the failure to reject a false null. In this regard, the power of a test of significance identifies the probability of finding an association if it exists. The power of the test can for example be enhanced by the level of significance chosen for the test and the sample size.     Ioannidis, Stanley and Doucouliagos conducted a meta analysis (study of studies) of 159 empirical economics literatures covering 67,076 parameters drawn from 6,700 empirical studies. They found that in half the research areas 90 percent of the studies were under-powered. Further, due to publication and reporting bias, a simple weighted average of results that were adequately powered (80 percent or more), showed that reported effects were exaggerated, typically by a factor of two. In a third of the studies the effects were overstated by a factor of four or more.9     Various hypotheses (stories) explaining deindustrialization were presented above without explaining how they were arrived at. The deductive and inductive methods as ways of arriving at hypotheses were discussed in Chapter 1. As practitioners of the deductive method, prospective economists learn the art of economic modeling in graduate school as a mechanism for deriving hypotheses for empirical testing. Economists differ in the amount of skill they acquire in modeling but, across the board, modeling entails a narrow focus to be mathematically tractable.     For example Rodrik (2015) used a two-sector model (manufacturing and non- manufacturing) to explore how demand, technology and trade shape the size of the manufacturing sector in countries across the income spectrum. The modeling is designed to interpret empirical findings regarding premature deindustrialization in L/LMICs. Schweinberger and Suedekum (2015) used a two-sector model                                       Data and their use in development economics 41
34
35(agriculture and manufacturing) to explore how an increase in price of manufactured goods (terms of trade effect) can induce premature deindustrialization in L/LMICs.     Such modeling can be valuable in uncovering important mechanisms but, even when built on defensible assumptions, leave out much that is important.     Heterodox economists do not have a penchant for such modeling and they argue that such modeling leaves out too much that is important. For example, they view premature deindustrialization as a historical phenomenon driven by power asymmetries between HIC and L/LMICs. This power they argue is mediated via international organizations like the World Bank, IMF, WTO and regional multilateral banks to serve the interests of HICs with regards to trade and investment. Thus, for them, analysis needs to holistic and include historical, multi- and inter-disciplinary, institutional and political economic perspectives.10     Some heterodox economists are also skeptical of the quality of quantitative data and of the tools used for such data analysis and so shun multivariate or econometric analysis. Some use descriptive methods and compile tables using various kinds of quantitative data in support of their arguments while conceding that their evidence is purely suggestive. Others prefer the inductive method whereby they generalize based on holistic case studies. While they do not avoid quantitative analysis, when they do so, it serves a larger story and is not the story.     All econometric tools used for macroeconomic data analysis can be utilized for microeconomic data analysis and the same shortcomings apply to this analysis. However, while secondary data are used in macroeconomic analysis, primary data are often collected and used in the microeconomic analysis of issues such as the impact of microcredit on household well-being.     The problem in much econometric research that attempts to measure impact is that there might be “unobservables” that cannot be captured by data but which nonetheless affect the nature of the association between the two variables in question. For example, those seeking and getting microcredit may have traits that are more likely to make them succeed. If such individuals are then more likely to show up in the sample of microcredit recipients, their personal success might get reflected as the success of microcredit.     This phenomenon is referred to in the econometrics literature as “selection bias”. There are econometric methods that can be utilized to address this selection problem, but a growing group of empirical economists believe that natural experiments,11 where possible, or RCTs otherwise, are a more “scientific” method (“gold standard”) for collecting the relevant data for empirical analysis that avoids the selection problem.     RCTs address the selection problem with the use of a treatment and control group. If the treatment group gets microcredit and the control group, otherwise identical, does not, the selection bias is taken care of since the result does not depend on who is selected for the study. This method of data collection has taken empirical micro development economics by storm and the usual econometrics tools are then used to analyze data that are considered far more superior.     However, RCTs have also been subject to criticism for many methodological shortcomings that Deaton (2010) has explored.12 They are viewed as technocratic 42 Background
36
37and no overarching conceptual framework appears to be driving the research unless it is individual utility maximization. Critics argue that focusing on individual utility maximization is not what development is about. For them development is more about nation building and adopting the right strategies such as the ones discussed in Part III of this textbook.     Another concern is the ethics of denying a beneficial intervention to the control group that is extended to the treatment group. Behavior of subjects in both the control and treatment group is assumed unaltered by the experiment, but this may not be the case. For example, those in the control group may seek the treatment (say credit) from an alternative source. Finally, treating humans in L/LMICs as guinea pigs for social science experimentation is found objectionable by some observers even if the research might serve a worthy social purpose.13     Many prominent economists are not persuaded by the possibility of replicating these studies (so they are not scientific as claimed) and there are even fewer chances of generalizing from these studies given the variability even across adjacent villages or urban localities.14 Deaton (2010, p. 448) points out that RCTs are neither suited to nor able to explain the underlying mechanisms of why some things work if they do or vice versa and as such are not helpful. Other critics charge that since they have become all the rage, they are therefore more easily funded than other competing research projects which are being crowded out.     Another objection is that a particular technocratic approach to microeconomic development economics is driving the research agenda. If such research projects are more likely to be funded, it creates the incentives to look for topics that can be framed as a natural experiment or an RCT and then decide on a research question. Ideally, the researcher should pursue a research question of interest because it has social relevance and then decide on the best research method to pursue it. Since RCTs are very expensive to set up and execute, the crowding out of other research is very likely an issue.15
38
39Summary and conclusion This chapter pointed out that economists, including development economists, are wedded to the use of numbers and perhaps are not as scrupulous in examining where they come from as they need to be. It pointed out the different ways in which the numbers are generated and that this can be wedded to the analysis techniques that are likely to be utilized for different research topics. For example, mainstream development microeconomics might use quantitative methods or RCTs and utilize primary (survey) data. A very different approach to development microeconomics is based on the use of qualitative research methods. For the purists, data generation is secondary or inconsequential and grassroots empowerment the primary objective. Development macroeconomics is more likely to utilize secondary data, often pooling cross-sectional and time series observations to enlarge the sample size.    Since all the data collection methods have strengths and weaknesses, it is best to be aware of the shortcomings and make the necessary qualifications. Further,                                      Data and their use in development economics 43
40
41when possible, it is best to view them as complementary to get more than one perspective on an issue. Rejecting evidence-based development economics entirely because of flaws in the data and analysis tools is an extreme position. The quality of data and the analytical tools have been improving over time and are likely to continue to do so. However, it is highly unlikely that economics or development economics as a field of the subject will ever legitimately claim to be scientific, as are the natural sciences, the Noble Prize in “economics science” notwithstanding. Economics shares with other social sciences inherent limitations to the understanding of “agents” (e.g. humans, firms, interest groups, classes, governments) that it studies. Thus, while relying on evidence is critical in moving the field along, given inherent limitations of the data, tools and subject, humility in making claims to the addition of knowledge needs to be the hallmark of a development economist.
42
43Questions and exercises 1. Explain the problems associated with the following kinds of data:    a. Secondary cross-sectional    b. Panel data. 2. Pick a subject of study and design a primary research project for it. What kinds    of issues would you need to be wary off? 3. Pick a subject of study and design a qualitative research study for it. State what    you see as potential advantages and disadvantages. 4. Reconstruct Table 9.2 (Chapter 9) for the period 2000–2014 that coincides    with the global commodity boom and tell a story of the identified association    to premature deindustrialization. 5. Identify a plausible story explaining premature deindustrialization and construct    an alternative Table 9.2 to support your story.
44
45More advanced  6. Identify plausible stories to explain premature deindustrialization. Draw on     The World Bank, World Development Indicators to construct indices to test     your hypotheses using time series analyses for the country of your choice.     Identify the shortcomings of your analysis.  7. Identify plausible stories to explain premature deindustrialization. Construct     indices to test your hypotheses using cross-sectional analysis by drawing on     The World Bank, World Development Indicators. The indices will dictate the     countries in your sample. Identify the shortcomings of your analysis.  8. Identify plausible stories to explain premature deindustrialization. Construct     indices to test your hypotheses using pooled cross-sectional and time series     data for analysis drawing on The World Bank, World Development Indicators.     The indices will dictate the countries and time periods in your sample. Identify     the shortcomings of your analysis. 44 Background
46
47 9. Explain how an RCT would address the selection problem in a development     microeconomics study other than microcredit. 10. Pick a micro development economics topic of your choice. Design an RCT     (including sites and sample selection) to conduct an impact analysis. Make     a notional budget to identify the possible cost of the study. Identify the     shortcomings of this method of data collection. 11. You are an intern in the office of a senior official at the UN and your first     assignment is to write a data brief on an L/LMIC country (your choice) that the     official will be visiting (three 1.5 spaced (font 12) typed pages excluding tables,     footnotes and references).    a. Start with a brief background for the country you have selected (one       paragraph).    b. The brief should include a comparison of current economic performance with       at least one other point in time to indicate change. Feel free to report a time       series if showing fluctuations over time is relevant for your selected country.    c. Construct at least one table from your statistical sources that, apart from       providing basic background, highlights the main problems this country       confronts and then describe these data.    d. Your concluding paragraph should contain concise recommendations       based on your country analysis.Hints for constructing table(s)        • Provide context for selected variables e.g. if you report debt statistics,          report it as a percentage of GDP rather than in absolute terms.        • Select variables in constant rather than current prices unless you have a          reason to do otherwise.        • Round out the numbers used to the highest unit possible (billions,          millions) and use two decimal places at most (false accuracy is implied          by using more decimal places since most of the numbers you will report          are probably inaccurate anyway).        • Only select numbers that you will actually discuss in the text.        • Provide a caption for the table and report your data source at the bottom.
48
49Notes  1 This is a truly valuable public resource provided by the World Bank.  2 If there is no sampling frame for selection in a primary cross-sectional survey, say    for a village or urban neighborhood, the first step would then be to do a village or    neighborhood census from which a sample selection can then be made. This obviously    adds to the expense of the survey.  3 The same is true for agriculture censuses.  4 While this sensitivity is laudable, one could argue that such research is needed to    understand social and economic reality to help improve it via policy.  5 Qualitative methods are more popular in the other social sciences. For details on the use    of qualitative methods in economics and development economics, refer to Starr (2014).  6 For a critical perspective refer to Cooke and Kothari (2001).  7 Many economists use the term “story” in recognition of the difficulty of capturing    economic reality with their models and empirics due to the problems identified in this    chapter and Chapter 1.                                          Data and their use in development economics 45
50
51 8 Economists are now expected to engage in “robustness tests” to show, for example, that    the estimates are not sensitive to minor changes in the sample.  9 I plead guilty to not having paid due attention to power issues in my empirical research. 10 An extreme view of economic modeling is that it is pseudo science that obfuscates    rather than clarifies. 11 These opportunities can occur fortuitously and are used by scholars for observation    and analysis. In this case, researchers do not control and establish the experiment, but    it occurs exogenously. For example, if education is a provincial rather than a federal    subject, and an educational reform affects schools in all districts of a province, but not    in other neighboring districts belonging to a different province, it would be possible to    assign children into treatment and control groups for an impact analysis of the reform. 12 Notwithstanding the many methodological, empirical and interpretive problems    identified, Deaton does not call for discarding the experimental method but for an    alteration of the focus on the scientific method where theory guides the experiment    and the experiment aids in the understanding of mechanisms that could have broader    applicability. 13 Silicon Valley entrepreneurs have expressed an interest in a Universal Basic Income    to offset the employment impacts of automation and funded the long-term RCT trials    (funding for 12 years) in Kenyan villages in 2016, www.businessinsider.com/basic-    income-study-kenya-redefining-nature-of-work-2018-1, consulted 3/10/2019. 14 For example, refer to Deaton (2010), Rodrik (2010) and Ravillion (2012). 15 Some years ago my research colleague and I received a very welcome response on a    research grant we applied for but funding was made contingent on converting the study    into an RCT. We withdrew the application since we considered an RCT as wholly    inappropriate for the research study in question and informed the granting authority    accordingly. Khan and Ansari (2018, appendix 7.1) elaborated on these objections.
52
53References Bannerji, A. V. and E. Duflo 2011. Poor Economics: A Radical Rethinking about the Way    to Fight Global Poverty (New York: Public Affairs). Chambers, R. 1997. Whose Reality Counts: Putting the Last First (London: Intermediate    Technology Publications). Cooke, B. and U. Kothari (eds.). 2001. Participation: The New Tyranny? (London: Zed    Books). Deaton, A. 2010. “Instruments, Randomization, and Learning about Development,”    Journal of Economic Literature, 48(2), 424–455. Ioannidis, J. P. A., T. D. Stanley and H. Doucouliagos 2017. “The Power of Bias in    Economics Research,” The Economic Journal, 127(October), F236–F265. Khan, S. R. and N. Ansari 2018. A Microcredit Alternative in South Asia: Akhuwat’s    Experiment (New York: Routledge). Ravillion, M. 2012. “Fighting Poverty One Experiment at A Time: A Review of Abhijit    Bannerjee and Esther Duflo’s Poor Economics: A Radical Rethinking of the Way to    Fight Global Poverty,” Journal of Economic Literature, 50(1), 115–127. Rodrik, D. 2010. “Diagnostics before Prescription,” Journal of Economic Perspectives,    24(3), 33–44. Rodrik, D. 2015. “Premature Deindustrialization,” Journal of Economic Growth, 21(1),    1–33. Schweinberger, A. G. and J. Suedekum 2015. “De-industrialization and Entrepreneurship    under Monopolistic Competition,” Oxford Economic Papers, 67(4), 1174–1185. Starr, M. A. 2014. “Qualitative and Mixed-Methods Research in Economics: Surprising    Growth, Promising Future,” Journal of Economic Surveys, 28(2), 238–264. 3 COMMONALITIES AND DIFFERENCES AMONG LOW AND LOW MIDDLE INCOME COUNTRIES
54
55Introduction The unit of observation in development economics is often the nation state, as mentioned in Chapter 1. There is also an implicit presumption when claiming there is a field of development economics that it applies to a set of nations that are not “developed”. This chapter will start by demonstrating that there is such a set of nations that share common features. It will also be indicated that it is easy to over- generalize and that while L/LMICs share characteristics with others in the same classification, they may also differ in many ways. Since the income variable is used for classification, how it is computed is explained first. In fact, this exercise in and of itself illustrates differences between L/LMICs and HICs.
56
57Classification variable: PCGDP Per capita gross domestic product (PCGDP) represents a division of GDP by population. As defined in introductory economics textbooks (see Chapter 1, endnote 4), GDP is the sum of final goods and services produced for the market place in a specified time period. Thus to calculate GDP, quantities and prices are needed and for L/LMICs this is not straightforward.    Prices in L/LMICs for similar goods can vary because markets are segmented due to poor infrastructure, lack of information, transportation and communication costs. There may be regional variations because internal movement across administrative entities (districts, provinces, states) is difficult for political reasons or due to poor infrastructure. The lack of storage may mean high seasonal price variations.    Output data may be difficult to procure because illiteracy may result in a lack of systematic record keeping. Producers may have to rely on memory and there can be a great deal of recall error. There is also systematic understatement because tax assessment depends on output. The lack of competent statistical agencies also leads to endemic inaccuracies. This is compounded by a much larger size of the informal sector as a percentage of total production and a much larger proportion of home production than in HICs (La Porta and Shleifer, 2014). Also, a smaller                                     Commonalities and differences among L/LMICs 47
58
59proportion of the economy is monetized (barter is the alternative), so many transactions are beyond the purview of GDP statisticians. For all these reasons, GDP is understated.    To get from GDP to GDP per capita, accurate estimates of population are needed. The problems of getting good data, including census data, were explained in Chapter 2. In addition, conducting a census is fraught due to political tensions. These are exacerbated in countries experiencing regional strife because they might reinforce regional conflict if resources are constitutionally distributed based on regional population sizes. All ethnic groups then have a stake in overstating their own population and understating those of the others, and there are often charges that the census is rigged in favor of the dominant ethnic group.    To make comparisons across countries meaningful (i.e. to have all variables in real terms), inflation needs to be adjusted for. For this an accurate index has to be generated to deflate and convert nominal into real values. Again, the problems of getting accurate price data are greater in L/LMICs for the reasons explained above. This is in addition to the problems resulting from insufficient statistical expertise and the inherent problems of getting good price indices as explained in introductory economics textbooks.    All PCGDPs in real terms are then converted into US$ terms using $ exchange rates for international comparisons.1 The conversion using the exchange rates is to the US$ since it is currently the global reserve currency.2 This means, among other things, that it is used by central banks to maintain their reserves in and many key international economic transactions are conducted in $ and key international commodities (like oil and copper) are priced in them. The problem is getting accurate exchange rates since countries may have an incentive to systematically overstate or understate their exchange rate.    Why should LICs have an incentive to maintain an overvalued exchange rate? One reason is that local elite consumption patterns emulate those of HIC elites and are hence based on imported luxury goods, and overvaluation of the exchange rate results in lower local currency prices. This consumption is effectively subsidized by the rest of society, particularly exporters.    Initiating industrialization could be another reason. In local currency terms, the imported capital goods needed are cheaper. Again, local industry is implicitly subsidized by thwarted exporters. The export sector is implicitly taxed since these exports are more expensive in foreign currency terms with an overvalued exchange rate.    Overvaluation overstates GDP per capita in constant $. If there is overvaluation, an active secondary market generally develops. Since secondary market transactions are generally outlawed because they draw away foreign currency from the central bank, there is a risk associated with such transactions. This means the premium over the official exchange rate on the secondary market is likely to overstate the extent of overvaluation.    Countries which are already substantively industrialized and are engaged in catch-up growth may want to encourage their manufactured exports by 48 Background
60
61undervaluing the currency. In this case, exporters are implicitly subsidized by consumers and importers. In some cases, countries have dual or multiple exchange rates for different activities depending on what they want to encourage. For example, foreign travel as a luxury good could be discouraged while studying science subjects abroad could be encouraged.    Such practices are now not common since the IMF is active in most L/LMICs and loan agreements with the IMF make it incumbent on countries to unify the exchange rate and make it flexible or “market driven”. The notion of a market- driven flexible exchange rate is a myth since the central bank influences the interest rate as one of the key macro prices influencing the exchange rate. For example, interest rates lower than other countries would result in a higher demand for foreign currency and exert pressure on the local currency to depreciate. Similarly, higher local inflation relative to other countries would reduce exports and again put pressure on the local currency to depreciate. Since the central bank controls the money supply and influences the interest rates, it has a strong influence on the exchange rate. Even if the exchange rate was market driven in this qualified way, the currency could still be undervalued.    Goods that form part of trade (traded goods) determine currency values, but in LICs the proportion of the non-traded sector of the economy is a much higher proportion of total transactions. These would include land, housing/rents, services (restaurant meals, haircuts), transport and utilities.    The larger the ratio of non-traded goods relative to traded goods, and the lower the price of non-traded goods relative to traded goods (due to lower wages), the higher the undervaluation of an exchange rate. One therefore needs to adjust for real purchasing power across countries that conversion of real GDP via the official exchange rate does not capture.    Exchange rates that adjust for purchasing power are referred to as purchasing power parity (PPP) exchange rates. The idea here is to adjust for systematic difference in the cost of living across countries resulting from the higher ratio of non-traded goods and lower real wage rates that figure prominently in prices of non-traded goods in L/LMICs. The method of adjusting exchange rates for cost of living differences requires estimates of prices and expenditures, and international agencies, regional agencies and local statistical offices share responsibilities for this.    The International Comparison Project was initiated in 1968 by the United Nations Statistical Office to adjust for the undervaluation of the official exchange rate in most L/LMICs (Kravis, 1986). A simple way of thinking about the conversion is to use say US prices (much higher) and quantities of a given L/ LMIC. Thus, as expected, the result of the conversion would be that PPP adjusted GDP in constant $ is generally much greater for L/LMICs than it is without the PPP adjustment (Appendix Table 3.1).3    While GDP adjusted for PPP can be a more accurate representation of comparative purchasing power, it is not a good measure of purchasing power of L/LMICs on the international market for capital, intermediate goods and raw                                      Commonalities and differences among L/LMICs 49
62
63materials. In this regard, GDP per capita in constant $ is a more suitable measure of real LIC economic development prospects.
64
65Commonalities Appendix Table 3.1 uses data on economic and social variables by income classification to identify commonalities. Appendix Table 3.2 then shows that, notwithstanding commonalities, there are variations across regions. Appendix Table 3.3 picks one region and shows how there are variations even within a region. This theme is explored further in the next section on limits to generalization.    Two caveats are worth repeating here before proceeding. First, it was made clear in Chapter 2 that many errors can creep into all forms of data and so although the Appendix Tables appear to contain facts, this is only so with many qualifications. Second, even if we view what is reported in the Appendix Tables as facts, contrary to the common saying, such facts rarely speak for themselves. Instead, social scientists including economists impose their own narratives or stories on the facts. Some are more plausible than others, but to imagine that social scientists attain the absolute truth or capture social or economic reality is a stretch.    The variable in the first row of Appendix Table 3.1 shows average PCGDP in constant $ terms (adjusted for inflation) in HICs was 58 fold greater than in LICs in 2017. The second row provides PCGDP in PPP terms. As expected, the PPP adjusted gap in average per capita real GDP between HICs and LICs is lower and was 23 fold in the same year.    The next six variables pertain to the structural features of the economy referred to in Chapter 1 i.e. the relative sizes of the various sectors in the economy and sector productivities. The agricultural value added per worker as an indicator of agricultural productivity shows that agricultural productivity in HICs was 70 fold higher than in LICs in 2016. This partly explains the next row, which shows agricultural value added as a percentage of GDP of 1.3 percent in HICs compared to 26.3 percent in LICs. Since agriculture is so immensely productive in HICs relative to LICs and LMICs, labor force employed in the agricultural sector can be a very small percentage of the total labor force. The other reason for the low share of agricultural value added in GDP in HICs is that they are highly diversified, and accordingly the shrinking of the sectoral share of agriculture is one of the “stylized facts” of economic development discussed in Chapter 1.    Industry is also much more productive in HICs relative to the other income groups. However, at 25 fold greater than LICs in 2016, this is much less so the case than in agriculture. The next row explains why it is a misnomer to refer to HICs as industrialized economies to distinguish them from other country groups. The industrial sector in all country groups is larger than in HICs and that is also the case for manufacturing as the largest component of industry for MICs relative to HICs.4    The service sector in HICs is almost three-fourths of the whole economy compared to two-fifths for LICs and about a half or slightly more for the MIC 50 Background
66
67groups. However, the bulk of services in HICs are information technology intensive, especially the technical and professional occupations, and this is why the service sector is 27 fold more productive in HICs than in LICs, which is characterized by informality and self-employment (Chapter 1).     One way of assessing the level of prosperity attained is via access to electricity. In 2016, only 30 percent of the rural population in LICs had access to electricity. In urban areas it was much higher at 68 percent, but even then the gap between urban access in LICs and HICs (at 100 percent) was dramatic. A broader gauge of prosperity is energy consumption. Once again, the same linear pattern is evident by country group income classification. Accompanying energy consumption is CO2 emissions (last row of Appendix Table 3.1) and, as expected, per capita emissions in HICs were 33 fold greater than in LICs in 2014.5     For developmentalists, saving and capital accumulation are central to initiating catch-up growth (Chapter 6). In this regard, LICs have low saving rates relative to LMICs and MICs and since LIC investment rates exceed saving rates, the gap is made up by borrowing and aid. LMICs also have a saving/investment gap, but it is much less acute than LICs. Since HIC economies have matured, they do not have the high investment rates of the middle income economies that have experienced or are initiating catch-up growth.     Neoliberal economists privilege the private sector and hence central to the needed structural reforms are those that promote “ease of doing business”. A set of indicators reported in Appendix Table 3.1 addresses this issue and across the board there is a negative association of income level and the “cost of business start-up”. Here, as elsewhere, heterodox scholars suggest that the causality may well run in the opposite direction and once catch-up growth is underway, management and administration also improve in conjunction i.e. that the ease of doing business emanates from the economy as it improves.     Apart from the saving/investment gap and a fiscal imbalance, LICs and LMICs also suffer from an external imbalance on the goods and services account. Only UMICs and HICs had a balance of trade surplus, though as a percentage of GDP in 2016 it was modest at 1 percent. LICs had a deficit of 18.1 percent while LMICs had a deficit of 3.6 percent.     For new developmentalists, technology is central to catch-up growth and all the indicators point to LICs and LMICs lagging. For example, this is the case with the adoption and use of cell phones and the use of the internet and broadband. Adopting a “glass half full” approach, one could argue that the pace of diffusion of these technologies to L/LMICs has been rapid compared to other technologies in the past as evident from recent time series data (not reported).     While it is difficult to standardize publication in scientific and technical journals by country income groups, LICs and LMICs clearly lag. Information on patents secured by domestic residents is not available for LICs, but LMICs lag behind the other country income groups. While these indicators probably reflect what is true for other variables for which data are not available (e.g. research and development), new developmentalists are also open to the same charge of technological change                                     Commonalities and differences among L/LMICs 51
68
69being endogenous as is true of the heterodox critique of neoliberalism with regards to business conditions. The important question in this debate among alternative approaches is where to focus to initiate the process of catch-up growth and this issue is taken up in the last chapter of this textbook (Chapter 15) after a detailed discussion of alternative approaches in Parts II and III.     For the basic human needs/human development approach, the starting point is ensuring adequate investment in humans to make them productive participants in the economy. The next sets of indicators in Appendix Table 3.1 provide information on social investments and human development. The human development indicators and social investments are at comparatively very low levels in LICs based on expenditures on education and health, access to services like health staff, drinking water, sanitation, life expectancy, mortality rates, literacy rates and enrollments. However, there is also good news in that human development indicators vary positively with country group income levels across the board. Also gender gaps improve by country group income level. For example, the gender literacy gap decreases markedly as country group income increases. Also, within the LIC group, scanning the time series data (not reported) show rapid improvement in human development indicators even over the last five years up to 2017.     Having established the “stylized facts” of what it means to be an LIC, the next two Appendix Tables establish the limits to generalizations. In Appendix Table 3.2, LICs and MICs are bunched into regions and variation by region is demonstrated using economic variables and the human development index. For example, Sub-Saharan Africa (SSA) and South Asia (SA) consistently lag on almost all indicators including the human development indicators.     A scrutiny of Appendix Table 3.2 suggests a possible foreshadowing of the economic emergence of South Asia. Its human development index on average is ten points higher than SSA. Its per capita GDP growth of 5.5 percent, post the 2007–2009 global financial and economic crisis up to 2017, has almost matched the 6.6 percent growth rate of East Asia and the Pacific region. The other three regions have stagnated in this period with a growth rate of almost 1 percent. The high growth rate in SA was accompanied by much higher investment and saving rates compared to SSA and the best business conditions by far across the board in 2018.     However, even here there are limits to generalization, and to demonstrate this Appendix Table 3.3 picks the SA region to demonstrate variation by country within the SA region.6 Sri Lanka as the only MIC in the region clearly dominates in most economic indicators and its high human development index (.77 compared to .56 for Pakistan). In fact Sri Lanka initially came into notice for its high social investments, high human development indicators, low poverty rates, and then for being the first in the sub-continent to attain MIC status (Gunatilleke, 2000).     Pakistan was viewed as the “development model” to emulate in the 1960s for having initiated impressive catch-up growth (Papanek, 1967).7 The growth momentum was sustained such that up into the 1980s and 1990s, Pakistan had the second highest per capita GDP growth rate. However, Pakistan lagged in social 52 Background
70
71investments and the associated human development indicators. Experts warned that Pakistan would not be able to sustain its economic growth momentum without concomitant investment in its people, but the latter remained a low priority for Pakistani policy makers, both military and civil. Nonetheless, Pakistan was able to maintain its lead over India in per capita GDP terms until the turn of the century. India overtook Pakistan with a much higher growth trajectory in 2006 (Khan, 2011) and by 2017 Pakistan’s per capita GDP was only about three-fifths of India, as indicated in Appendix Table 3.3.8     Bangladesh (formerly East Pakistan) declared its independence from West Pakistan in March 1971. As a province, it had a larger population but lagged behind West Pakistan on all economic indicators. Post independence, Henry Kissinger is said to have referred to it as a basket case. Yet all social indicators started to dramatically improve post independence in Bangladesh and the drop in its fertility rate excited the attention of demographers (Cleland et al., 1994; Caldwell et al., 1999).     Subsequently, it invested heavily in the textiles and garment industry in the 1980s, and within a decade became a world beater in garment exports despite growing no cotton (Rhee, 1990). While in 2018 Pakistan ranked fourth in raw cotton exports, in 2017–2018 its total textile exports were $12.5 billion while Bangladesh apparel exports alone in the same period were $30.6 billion. As Appendix Table 3.3 shows, Pakistan still has a higher per capita income, but Bangladesh has a much higher economic growth trajectory (5.3 percent between 2010–2017 compared to 2.0 percent for Pakistan) and is very likely to overtake Pakistan fairly soon.9     Pakistan can point to having to cope with the blow back of the Afghan wars and associated terrorism for its poor economic performance. This could have resulted in a poor investment climate and the low domestic investment as shown in Appendix Table 3.3. Alternatively, its stubborn disregard of social investment (lowest human development in SA), pathetically poor saving rates (7 percent compared to 30 percent for Sri Lanka) and lack of structural transformation (12 percent manufacturing as a percent of GDP compared to 17 percent in Bangladesh) are self-inflicted wounds. Of late, it has dramatically improved its business climate compared to other SA countries (6.8 percent of per capita income in terms of cost of business start-ups compared to 21 percent in Bangladesh), but quite clearly this is not a sufficient or even necessary condition for impressive economic performance.10     This detour above into SA economies was simply to illustrate the limits of generalization. Ultimately, detailed historical inductive (case) studies are the only way to understand individual economies.11 Yet, this does not negate the commonalities among L/LMICs that give meaning to the pursuit of development economics and the paragraphs that follow revert to this theme.     To summarize, the stylized facts in Appendix Table 3.1 indicate a large number of commonalities across LICs which represent challenges even as they try to initiate catch-up growth. These include high levels of malnutrition, which                                     Commonalities and differences among L/LMICs 53
72
73is indicative of high poverty rates (Chapter 4). Low levels of literacy and mean education rates in terms of the composition of the labor force suggest a larger proportion of unskilled relative to skilled or professional workers, and these two variables contribute to explaining low levels of labor productivity. Low levels of public investment in health and other social services are associated with high infant mortality and high maternal mortality rates which are indicative of low life expectancy.     The low levels of labor productivity in agriculture, the largest component of rural economic activity, means a large percentage of the population is in the rural sector in L/LMICs. Many LICs export composition is still primary goods intensive (raw material, agricultural or mining based). Some countries like Bangladesh, Cambodia and Vietnam have moved into low-tech labor intensive manufactured goods like textiles or leather goods. As labor costs in China rise further, more opportunities for such specialization in L/LMICs will be created.     Appendix Table 3.1 also points to a number of other challenges. LICs confront high balance of payment deficits and low saving and investment ratios with the latter far exceeding the former. In addition, tax capacity is low and so they also confront fiscal deficits and high debt levels. Structuralists refer to these as the three gaps (i.e. foreign exchange, saving and fiscal) that need to be addressed with the help of foreign aid (Taylor, 1994, chapter 11).     Other challenges include environmental and resource degradation since industrial safeguards are not in place and implementation capacity of existing legislation is weak. In some cases, population pressures could result in resource degradation. To add to the social discontent that may result from high population growth rates, they may face regional and ethnic strife in the competition for limited resources.12 Three commonalities that deserve special attention are population growth, rural to urban migration and market segmentation.     High population growth rates mean high household population dependency ratios.13 High population growth rates and a young population can be a plus as an addition to the labor force (demographic dividend). However, to realize the dividend, the state needs to have the resources and capacity to deliver the needed social and physical infrastructure. If opportunities are not available on the demand side, more labor on the supply side means unemployment and social discontent.     Neoclassical economic theory explains the high demand for children among the less well off in LICs premised on using them for chores, care-giving (girls) and for old age security (boys). Due to high infant mortality rates, high birth rates are a strategy to ensure that the numbers of children desired are attained. This is viewed as a demand for children as “investment goods”. As household income levels increase, children start being viewed more as “consumption goods” and a quality–quantity trade-off kicks in (Becker and Lewis, 1973). The price of providing quality (schooling, health) is high and so the demand for children drops. Higher female education levels complement this trend if it leads to women entering the labor force. In urban areas this is likely because two paychecks are often needed to maintain a barely middle class lifestyle.14 54 Background
74
75    It follows from neoclassical theory that incentives can be used for reducing population growth and this has been tried by a number of countries such as with preference in apartment allocations and school admissions for smaller families.15 Other mechanisms for influencing fertility rates are enhancing female education levels and population programs that diffuse birth control information and meet the demand for contraception.     An unpleasant demographic story that still plays out in many L/MICs has been referred to by Sen (1990) as the puzzle of the “missing women”. The biological sex ratio at birth (M/F) has been estimated to be 1.07. In many L/MICs, particularly India and Pakistan in SA, there has traditionally been a male preference since women were viewed as marrying into other families and therefore an asset to other families and not available for example as old age security to the family they were born into. Sen simulated how many women should be present for the biologically expected ratio to hold and hence his dramatic finding of 100 million missing women. Ultrasound technology is alleged to have made this worse in India.     China put into place a very aggressive “one-child” policy in 1982 and in 1988 extended it to rural areas (with some exemptions). It attained a stationary population growth rate, but not without social repercussions and one of these was the missing women syndrome (Ebenstein, 2010).     In general since rural fertility rates are much higher and living conditions much worse, LICs confront high rates of rural to urban migration. Mainstream economists separately analyze push and pull factors as leading to such migration. Apart from population growth, push factors could include natural disasters, agricultural mechanization, public sector infrastructure projects like dams displacing populations, and predatory landlords. The main urban pull factors include the perception of higher wages and more work opportunities, more security, better living conditions and the glamour of the city.     The term urban bias was coined by Lipton (1977) to describe the asymmetry in investment in social and physical infrastructure in urban relative to rural areas. The reasons could be that the rich and politically powerful live in the cities and also the greater population concentrations in cities compared to the diffused populations in the rural areas are more likely to produce social protest by the poor who are deprived. The high pace of rural urban migration and subsequent high organic population growth rates in the cities far exceed the capacity of urban municipalities to cope and the result is shanty towns and urban blight.     Preston (1979) estimated that the share of urban population in LICs between 1950 and 1975 increased from 17 to about 28 percent. The increase for current HICs between 1875 and 1900 was from 17 to 26 percent. Thus the increase is not unprecedented. However, the rate of internal growth in urban population in LICs between 1950 and 1975 of 188 percent is unprecedented. The corresponding increase between 1875 and 1900 for current HICs was 100 percent. Thus the natural increase in population growth in urban areas now accounts for a greater increase in urban population growth then does rural-urban migration and this contributes to unmanageable megacities in L/LMICs.                                      Commonalities and differences among L/LMICs 55
76
77   Another common characteristic of L/LMICs is the market segmentation and the economic dualism this generates in the product, labor and capital markets.16 The market, based on a number of assumptions, is viewed in mainstream economics as a homogenizer i.e. price differences across regions should disappear due to the mobility of goods since price differentials for the same goods provide profit opportunities to businesses.17 Similarly, wage and interest differentials should shrink due to the mobility of labor and capital.    In LICs, price differentials persist and this phenomenon has been referred to as market segmentation which produces dualism. Poor physical infrastructure hinders transportation and communication. Paved roads and rail links have limited coverage, are of poor quality and are not maintained.18 Poor communications limit information flows (cell phones have redressed this failing to some extent), and the non-existent storage or very high storage costs (warehouse and refrigeration) raise marketing costs. Thus different markets can be physically cut off and one way of inferring the existence of market segmentation and the resulting dualism is from price differences for similar products across region and time.    In both urban and rural areas, two separate interest rates can co-exist. In the formal sector government owned or directed private banks may make loans to large formal sector firms at subsidized interest rates. In the informal sector, capital scarcity could mean very high interest rates. Households have recourse to family and friends, microcredit if available, ROSCAs (Rotating Saving and Credit Associations), or else resort to a money lender.    A similar duality prevails in rural areas. Government-owned or -controlled agricultural development banks provide interest free loans to the so-called “progressive farmers” trying out new agricultural technology. In practice, this often turns out to be the large landholders with “contacts” with the bureaucracy or politicians, even if the loans are intended for the small farmers. Small farmers and other households then have to borrow from moneylenders, shopkeepers, landlords, ROSCAs, or from family and friends. Moneylenders operating in the local environment have access to the relevant local information to assess credit risks and have low or no overheads. The poor infrastructure, lack of information and small size of loans and hence high transaction costs deter private sector banks from opening branches.19    Technological dualism drives segmentation in the labor markets in both urban and rural areas. In the formal sector, large industrial firms are likely to use highly capital intensive technology. Various incentives could drive this. For example, wages in the formal sector may be governed by unions and legislation driving up labor costs. As noted above, interest rates might be subsidized, the exchange rate overvalued and various tax incentives (accelerated depreciation or tax holidays) cheapen capital. By contrast, in the informal sector wages are market determined, entry is easy, labor turnover is high and no benefits are provided. There are myriad activities in the informal sector and some examples include crafts, home-based production, petty retail, food stalls, street vending, letter writing, knife sharpening, street entertainment (bear, snake), junk collection, domestic service and day labor. 56 Background
78
79   In rural areas, large farms are highly mechanized and input intensive (i.e. use tractors, harvesters, threshers and also tube-wells, fertilizer and insecticides), use wage labor and often have monopsonistic power. Juxtaposed with this picture are small peasant farmers who work as share croppers and many still use primitive technology. Land is sometimes “rented in” from absentee owners and generally family labor or landless labor is used.    The non-agriculture rural sector is also characterized by informality with a high prevalence of crafts production, trades and services. At least initially, migration keeps rural-urban links intact. Rural migrants generally find entry into the urban informal sector labor but may return during peak agricultural seasons when there is a labor shortage such as during harvesting. There is also the reverse flow of remittances to support family members who stay in the rural areas. However, given the rural-urban migration and incentives for capital intensity in the formal sector, various categories of employment in L/LMICs are characterized by wages exceeding productivity.    Over-staffing in the public sector might result from nepotism or jobs resulting from political connections. Productivity could also be low because of the lack of complementary inputs. Finally, productivity might be low due to low nutrition, education or training (impaired employment). The commonalities pointed to so far pertain to current conditions. Perhaps more important is the common history of colonialism most current L/LMICs faced.
80
81Colonialism and neo-colonialism20 There are several theories of imperialism and colonialism and the prominent ones, starting with Marx, are explored in Chapter 5. The debates about the impacts of colonialism are still ongoing. Since there is no counter factual (we cannot replay history without colonialism), this question cannot be directly answered as is the case for so many issues in development economics. Nonetheless, argumentation and clever scholarship attempts to build a case one way or the other.     Prominent development economists such as Rostow (1960, pp. 26–27) and Bauer (1984, p. 58) argued that colonialism paved the path to progress. In their view, it provided poor countries with social and physical infrastructure, bureaucracies and modern attitudes which enabled future progress. The dissenting arguments claim that the legacy of colonialism included demographic devastation, surplus extraction,21 arrested development (see below), corruption, low tax effort, poor social infrastructure, income inequality, ecological degradation and genocide.     Cypher (2014, chapter 3) reviewed the literature on colonialism and documented by citing various sources the demographic and resource extraction aspects of colonialism. In Latin America, the indigenous population declined from 70–90 million in 1540 to 3.5 million in 1690. The reasons for this were wars, viruses that native populations did not have immunities against, forced labor, deprivation and malnutrition. In Africa, between 1600 and 1900, 12 million were sold into slavery and 36 million died during the walk to the coast or en route. Colonial agents                                     Commonalities and differences among L/LMICs 57
82
83instigated inter-tribal wars that resulted in the enslavement of defeated tribes. Thus, between 1650 and 1850, Africa’s share of the world population declined from 18 to 8 percent. In the United States, Canada, Australia and New Zealand, the indigenous populations were virtually exterminated.     Between 1757 and 1812, Great Britain extracted about 5 to 6 percent of Indian GNP annually.22 Similarly, there was an extraction of 15.6 percent of net national product (GDP − capital depreciation) from Indonesia by the Dutch as late as 1930. The last twenty-four years of Spanish rule resulted in the extraction of 7.2 percent of the annual income from Mexico. The mechanisms of extraction were brute force, including paying only for part of colony exports based on resource extraction from mines or plantations, trade (prices set by colonial administration) and oppressive taxation. Trade, head (per person) and hut taxes raised revenue but also induced forced labor (to pay the taxes) at exploitative wage rates.     Baran (1957, pp 142–149) argued the colonized countries were placed on a lower path to economic development because they arrested productive forces and transferred resources from the colonized to the colonial powers. The needs of the colonizers shaped economic activity in the colonies, and so infrastructural investments served the purpose of connecting hinterlands, plantations, or mines to ports for primary commodity exports. Indigenous development may have required specialization in different kinds of agricultural and manufacturing commodities and regional networks to enhance trade. Even now, many LIC trade networks look to the ex-colonizers rather than to the region they are located in and the old trade patterns of selling primary products and importing luxury goods has continued long after colonialism and into the post-colonial period. Colonial historians point out that the colonial administrations maintained dominance by promoting or exploiting inter-tribal and ethnic tensions (divide, conquer and rule) and that this also deterred the formation of regional trade networks and a sense of nationhood after independence.     Cypher (2014, pp. 97–99) also documents that colonizers were not interested in competition from the colonies and so engaged in a systematic process of instituting deindustrialization. Examples of such deindustrialization include the elimination of sophisticated textile manufacturing in India via brute force, rules and tariffs, and in Egypt via a manipulated war. A tariff of 70 to 80 percent was placed on Indian textile exports to Great Britain while duty free textiles were imported into India. Since cotton could be imported duty free into Great Britain, there was a move away from exporting textiles into exporting the raw material. Gradually the very sophisticated textile industry that was more competitive than any European industry was decimated. Similarly, the attempt by Muhammad Ali of Egypt to build a flourishing cotton textile industry based on indigenous high quality cotton was thwarted by the British.     Another way of exploring arrested development is by examining the impact of colonialism on future development. Alam (1999) demonstrated that the more sovereign a country was in the initial period, the better the growth outcomes between 1900 and 1950. Colonies and quasi-colonies showed the worst performance due 58 Background
84
85to their inability to structure their policies to favor domestic factors because of this lack of sovereignty. Similarly, scholars have shown that economic growth and physical and social infrastructure in the post-colonial period immediately picked up relative to the colonial period and was enhanced multiple folds in a short span of time.23     To sum up, it appears based on the information provided above that on balance the colonies had to deal with baggage that negatively impacted their future economic growth trajectory. This information above could be re-cast into institutional and structural factors based on the definitions and distinctions made in Chapter 1. The structural reasons for arrested development included a created dependence on primary commodities for which the terms of trade over time were declining.24 The exhaustion of natural resources resulted from ecologically unwise commodity specializations forced on L/MICs to fulfill the needs of colonial powers. The distorted infrastructure did not encourage rational development due to distorted trade patterns that served the needs of colonial powers. A quick profit trading mentality was picked up from the mercantile colonial ethos by the elites. Thus, local elites avoided industrialization, which was a higher risk activity. Finally, there was a paucity of social investment (health, sanitation, water supply and education) evident from the comparative progress made in the post-colonial period.     The institutional drawbacks of colonialism that arrested future development included social dualism; with an elite that lived like the colonizers, spoke their language, was often cut off from their own mass culture and did not identify or serve their own country even in the post-colonial period. The concentration of economic and political power based on land grants was and still is detrimental to economic development because it precludes wider economic and political participation (Acemoglu and Robinson, 2012). Another drawback was inheriting an autocratic colonial bureaucracy that had instilled into it a culture of governing rather than serving.     Neocolonialism differs from colonialism in that the political administration is turned over to the colonies after they attain independence, but the power equation is still massively unequal. Scholars like Amsden (2008) argued that even though overt colonialism ended, HICs still politically and hence economically dominate the LICs via various mechanisms. Recall that the mechanism for the transfer of surplus under colonialism was force, manipulating the terms of trade, and taxes. These mechanisms had to be altered since the political administrations were no longer under colonial authority.     One mechanism is loans provided by banks in HICs that create indebtedness and an onerous burden of future interest and principal repayments. While the blame for this rests on LIC elites who misappropriate and badly use funds, future generations and the poor in L/LMICs suffer the collateral damage. Aid often comes in the form of tied loans which reduce the actual value of the loan since the country is not free to get the best price for the best quality of goods. A portion of the loan is a grant in that the interest rate may be below the market rate (LIBOR – London                                     Commonalities and differences among L/LMICs 59
86
87Interbank Offer Rate is used as a benchmark) and the repayment period may include a grace period and long repayment periods. In the parlance of international giving, soft loans are the ones with a high grant element. Soft loans can be used as a mechanism to cultivate political alliances, influence and secure donor country interests (Chapter 11).     Foreign direct investment is sought by L/LMICs for technology, capital and employment. However, ultimately how well L/LMICs can negotiate the terms determines the benefit-cost nexus. Multinational firms can transfer surplus out of L/LMICs via transfer pricing (over-pricing inputs and under-pricing outputs to siphon profits to low tax locales), excessive royalties, fees, and dividend and profit repatriation. Again, one could ask why political elites would want to bend over backward to attract FDI if its costs exceed its benefits. The political economy of development perspective is that the benefits accrued to the elites (kickbacks/ jobs) while the costs are borne by society at large (Chapter 12).     International trade can be another mechanism for the transfer of surplus. Once again, trade has many potential benefits by making markets available and enabling economies of scale, providing access to foreign exchange and technology, inducing higher productivity via competition, providing access to value chains that help in attaining quality control, marketing and managerial knowledge. Once again, the political economy of development perspective is that economic power determines the terms of trade (import and export prices) and this can result in surplus transfer. Monopoly and monopsony power mean high prices for imports and low prices for exports respectively for L/MICs (Chapter 8 and 9).     This approach also views the greater economic power of HICs as leading to greater political power which enables them to reinforce their economic dominance. Thus asymmetrical power in international organizations like the WB/ IMF may result in austere conditions for loans that make L/LMICs foreign capital friendly and susceptible to resource transfers. Limited power in the WTO may mean that they see trade rules including tariff structures confronting their exports that are detrimental to L/LMIC economic growth. HICs meanwhile get away with subsidies on their exports (HIC agricultural subsidies have been a big unresolved negotiating issue in the WTO), which were banned by the WTO Uruguay Round (Chapter 8).25     With the emergence of the BRICS (Brazil, Russia, India, China and South Africa), a power shift is under way. For example, the Chinese renminbi has been accepted by the IMF as a reserve currency. The IMF board approved more voting rights in the IMF for China, India, Brazil and Russia (now all among the top ten voting powers) in 2010, but the United States resisted approving this measure until 2016.26 China and the other BRICS countries launched a $100 billion New Development Bank (BRICS Bank) and the China-led Asian Infrastructure Investment Bank. The United States first opposed the BRICS Bank, but once the UK and EU signed up, the World Bank (led by the United States) agreed to work with these BRICS banks. These concessions acknowledged the shifting global economic and political power and the fact that power is acquired and not conceded.27 60 Background
88
89   Chang (2002) argued that currently advanced economies utilized policies and tools that global economic rules, such as those embodied in the WTO and policy dictates of the IMF and the World Bank, now routinely deny to L/LMICs. These include protectionism and the tools that engender local technological and industrial capacity development. Moreover, he argued that these are exactly the policies and tools that Japan and other East Asian countries used more recently to structurally transform their economies and attain catch-up growth.    The inspiration for these policies was Alexander Hamilton whose work was a revelation to Friedrich List, a 19th-century German economist, who conceptualized this mode of catch-up growth and coined the term “kicking away the ladder”. Both argued that while classical British political economists (Chapter 5) made the case for the US economy to specialize in primary commodities, it made more sense for them to industrialize to attain economic prosperity and that protection would be required to do so. Their view was that Britain was pushing free trade from a position of strength once it had industrialized and that its manufacturing advantage would prevent others from catching up without protection.28
90
91Limits to generalization While it may be possible to sensibly talk about development economics given the commonalities identified above, there are also major and important differences among L/LMICs that have a bearing on the growth strategy they may opt for. Countries vary a great deal by size and this can determine market size, the potential for attaining economies of scale and, consequently, MNC interest. For small countries it becomes more important to have an export-oriented growth strategy to attain economies of scale. While size brings economic advantages, it also brings administration challenges.     Countries can differ by the material resources they begin their development process with and this can have an important bearing on their growth trajectory. Resources can be a curse in that they could direct social energy into the appropriation of resource rents rather than on production. Apart from rent-seeking (corruption), a higher price for the resource on world markets can cause the local currency to appreciate as the higher demand for the resource translates into a higher demand for the local currency. This can hinder the ability of the country to diversify into other exports.29 Scandinavian countries as well as the United States and Australia are resource rich and thriving, and Norway in particular has used its oil wealth to form a rainy day sovereign wealth fund and hence demonstrated that if well used, resources are a blessing.     Early development economists pointed to cultural traits as militating toward or against nations prospering (Bauer, 1984). This has proved highly controversial as scholars have pointed out that the cause of late industrialization in nations such as Germany and Japan, relative to say Great Britain, was attributed by observers to laziness as a national trait. Of course once these nations prospered, efficiency was attributed to these nations as a cultural trait.30                                    Commonalities and differences among L/LMICs 61
92
93    The less homogeneous a country is the more potential there may be of them confronting political instability as a vicious circle. This is particularly the case when a geographic region coincides with an ethnic group, and there are regional development disparities in the nation, since this can promote separatism. The East Asian successes share cultural homogeneity as a commonality, though much else was important in explaining their catch-up growth as will be indicated in Chapter 9. Even so, the lack of ethnic strife is one less factor to worry about when trying to initiate catch-up growth.
94
95Comparative initial conditions Some of the earlier mainstream development economists suggested that LICs could follow the same economic development trajectory as current HICs.31 Apart from other analytical problems with this thesis, it ignored that L/LMICs started with a very different set of initial conditions compared to current HICs. The latter did not have to contend with the negative baggage of colonialism. In fact, being colonial countries, they had access to the labor, raw materials and markets of the colonies. They initiated agricultural revolutions that fed into their industrial revolutions. They were comparatively advanced in educational attainment, technological leaders, had political stability, and apart from these achievements had also experienced the socio-cultural and institutional changes complementing an industrial revolution.    These socio-cultural changes included the move to nuclear families and the emphasis on the legitimacy of personal striving and material incentives and rewards, and rational independent calculation which promoted labor mobility. Concomitantly, this reduced the emphasis on clan and ethnic group loyalties and placed more emphasis on efficiency and productivity and loyalty to the nation state. North America and Australia/New Zealand were important outlets for the emigration of those not content for religious or social reasons and also for economic migrants. Thus these lands provided a safety valve and also a resource due to the massive remittances. The movement of South/Southeast Asian labor to the Middle East is an example of this still being possible for LICs, but such opportunities are generally more limited now in that acquiring nationality is not an option.    Earlier development economists set much store by the importance of climate. A harsh hot climate can negatively impact productivity. All currently HIC countries are in the temperate zones. Tropical and sub-tropical climates bring to bear additional challenges of harsh weather and disease vectors.
96
97Current environment Gerschenkron (1962) pointed out that being a late developer could be an advantage. Perhaps it would be more accurate to characterize attempting catch-up growth in an environment that includes HICs as a mixed blessing. Current LICs have an 62 Background
98
99advantage of having a backlog of technology available for them to choose from. However, this technology is not free and evolved to respond to the very different needs of current HICs in a particular historical juncture. The premature use of labor-saving technology can compound L/LMIC unemployment problems.     Another related issue is that technology can only be viewed as disembodied from the society it evolved in a limited sense. If the social and physical infrastructure that the technology flourished in is different, the diffusion and maintenance of such technology in another context may be limited because of the lack of absorption capacity (Chapter 14).     Even so, at times there are clear advantages of adoption and adaptation. For example, the adoption of cell phones by farmers in Africa to cut out the middle person has been an unqualified benefit. Kenya adapted and used mobile technology for the transfer of funds and overcame the physical insecurity of financial transactions and the lack of banking infrastructure and access – low density of bank branches in the country and access for the poor to these branches. Tanzania is registering births using cell phones since distance and bureaucratic hurdles impede such registrations.     Leapfrogging is also possible and so mobile technology has made cable networks less relevant. Similarly, solar and other alternative energies may make an electrical grid irrelevant. Cheaper water cleaning technologies may make decentralized water supply possible. India adapted and made advances in cheaper medical technologies, like smaller X-Ray machines, that were subsequently imported by the west. China did the same for CT scanners.32 Rwanda gave approval for the first attempt to parachute medical supplies into remote areas via drones.     Public health technologies often exported with donor support have been a boon for L/LMICs. However, while this has had a beneficial impact on saving lives, it can also lead to a population explosion. Mortality rates declined much before a decline in fertility rates that organically accompanies higher education, incomes and cultural change. This partial demographic transition resulted in high population growth rates with a very high share of young cohorts in the population.33     Current conditions also do not easily permit a move to industrialization to attain catch-up growth as in the past. Current HICs dramatically added to the stock of greenhouse gases before knowledge of what they were doing was widely available and continued even after warnings of the consequences became widespread. Now the consequences of global climate change are upon the world and the impacts are being disproportionately felt by LICs whose ability to adapt is most limited. It is no longer possible to industrialize without restraint as a responsible member of the world community. However, clean production technologies have been developed by HICs and present a multi-win opportunity for green agriculture and green industrial development to L/LMICs (Chapters 14 and 15). Learning from the mistakes of current HICs is also a blessing in that some forms of resource degradation are not reversible.     Just as technology is often transplanted rather than an indigenous and organic evolution within L/LMICs, institutions evolved in the west can have                                     Commonalities and differences among L/LMICs 63
100
101side effects when transplanted. Welfare states evolved in response to ideological challenges after western societies had become more prosperous. L/LMICs that were becoming independent in a global environment in which the UN General Assembly had already approved an International Bill of Human Rights (1948) assuring humans of the rights to basic freedoms provided by a welfare state. Health care, education, progressive labor legislation, unions, occupational health and safety are all assured. The basic human needs/human development approach (Chapter 4) views social provision as paying dividends by making humans more productive. However, garnering resources for such provisions is a challenge.     The talented have historically often looked to enhance their skills wherever this may lead them. With more security and fewer restrictions, many of the talented in L/LMICs leave for HICs for higher education. While there is a potential gain, it can represent a loss not only from the productivity foregone but also due to the education subsidy that went into training them if the highly educated do not return. By the late 1980s, Africa had lost a third of its highly skilled labor force. Eighty percent of the highly skilled left Haiti (Khan, 2012).     Eventually the brain drain can be reversed and become a brain gain, such as has happened in China and India in the last decade, resulting in a virtuous circle when LICs eventually get engaged in catch-up growth. Such growth creates job opportunities, draws expatriate talent back home, which further boosts growth.34 A good example of this is how the growth of the information technology sector in Bangalore and its links with Silicon Valley was mutually reinforced by the flows of talent, information and services.35     Perhaps the starkest way in which the catch-up growth environment for current L/LMICs is different from what current HICs faced is the current context of economic globalization. Starting in the 1990s, economic globalization started receiving much attention in the development economics literature as judged by the books and articles published on globalization and economic development. Economic globalization can be thought of as economic interaction among nations, and such interaction, particularly between HICs and L/LMICs, can impact the economic development of the latter. The economic interactions viewed as economic globalization emanate from flows of resources (labor, capital, products and information) between nations. Viewed in this way, there is nothing new about economic globalization although there are time periods during which such flows have been more intense and unimpeded than at other times. As nations became independent, these flows could at least nominally be viewed as resulting from a “national volition” while during colonial times the colonized had little say.     The word nominally is used because national volition is difficult to realize. It would presuppose at least some form of representative government and at a bare minimum the ability to replace political parties representing constituents with elections. Second, it would presuppose the ability of governments to withstand foreign pressure that results from indebtedness or foreign investment and foreign trade dependency. The term neocolonial, as characterized above, represents the 64 Background
102
103lack of independence such that decisions do not serve L/LMIC nationals. The political alliances that are a mechanism for neocolonialism often mean the interests of national elites are concomitantly served with that of foreign governments at the expense of the broader L/LMIC populations (Baran, 1957).    The flows alluded to above include foreign aid and loans (capital), FDI (capital), foreign funds (portfolio or bank loans, capital), foreign trade (products), technology transfer (information) and migration (labor). There are forces opposing economic globalization in all countries based on the perception or reality of adverse consequences. In HICs, labor and environmentalists oppose it and populist politicians capitalize on such resentments. Labor opposes it because of the outsourcing of jobs or because of cheap imports that amount to the same. Environmentalists oppose it because the lack of environmental standards in L/ MICs can cause resource degradation, much of it non-reversible, which can have global repercussions such as on climate change from say carbon emissions or deforestation (degradation of a carbon sink). Labor in HICs demands a level playing field so that the L/LMICs products they compete with are subject to the same social and environmental standards. Some in L/LMICs view the imposition of such standards as another form of neocolonialism and argue that current HICs did not confront such standards at similar levels of economic development.36
104
105Summary and conclusions This chapter documents the commonalities among L/LMICs. These include poor social and economic indicators such as low literacy and mean education, high infant mortality, high maternal mortality, low life expectancy, high unemployment rates, high rural to urban migration and urban blight and economic dualism. They suffer from the triple deficits (balance of payments, fiscal and saving), high population dependency ratios, high levels of environmental/resource degradation, and regional and ethnic strife.     While all this is challenging enough, they are also attempting to attain catch-up growth in a global environment that is challenging. Even as they overcome the baggage of colonialism, they confront the challenges of neocolonialism resulting from asymmetric power between HICs and the rest. This power asymmetry results in the global economic rules for trade, capital flows and migration being framed in the interests of HICs and which represent barriers for L/LMICs.     L/LMICs face the additional challenge of attempting catch-up growth at a historical juncture in which the global commons (atmosphere, oceans) have already been severely degraded and have little capacity to sustain more degradation. International agreements that L/LMICs are signatory to require cleaner production standards in L/MICs as do MNC codes of conduct that L/LMIC partners agree to when soliciting FDI. Such pressures may result from buyers or shareholders in HICs pressuring MNCs for corporate responsibility or from international organizations like the WTO or the World Bank. In Chapters 14 and 15 we explore how L/LMICs might be able to own this process to their own advantage.                                    Commonalities and differences among L/LMICs 65
106
107    L/LMICs scholars might be right in pointing out that environmental standards are a domestic and not a trade issue. They might be right in pointing out that current standards in HICs were not the ones prevailing when current HICs were embarked on their catch-up growth. They might be right in pointing out that the lower standards and concomitantly the lower costs are part of their comparative advantage. However, all this is moot given the current state of the global environment. The silver lining of climate change is that it has brought to the fore the notion of global citizenship. While the unit of analysis in this textbook is the nation state, climate change highlights that the world is one entity and all global citizens will survive or sink together.     Social standards imposed on L/LMICs via trade rules and international organizations are not as clear cut. What are referred to as sweatshops in HICs are a source of livelihoods in L/LMICs. The alternative may be unemployment or hazardous occupations in the informal sector including becoming sex-workers or having no livelihood at all.     L/LMICs have rightly pointed to the baggage of colonialism and the challenge of neocolonialism in an economic globalization context. However, they may also need to look to their own managerial failings. The experience of NICs and BRICS suggests that nations have to take their destiny into their own hands. Economic and political power is zero-sum and is not conceded but has to be earned and taken. The BRICS have shown that this is possible but also that MICs should expect push back from HICs.     The challenge for L/LMICs is to attain middle income status so that they can meaningfully function on a global scale, but more importantly, to address internal problems of poverty and inequality based on their own resources. These problems represent the most critical commonality among L/LMICs. The alternative approaches to economic development have different perspectives on attaining catch-up growth to address these problems. Prior to the discussion of alternative approaches in Part II, the issues of poverty and inequality, as key economic development problems, are discussed in the next chapter.
108
109Questions and exercises  1. Explain the distinction between economic growth and economic development.  2. Explain some of the costs of economic growth.  3. Explain economic globalization and its mechanisms from an LIC/MIC     perspective.  4. Explain some commonalities across L/MICs.  5. Explain why generalizing across L/LMICs could be problematic.  6. Explain the neoclassical theory of demand for children and how this and     an institutional story explain a decline in the population growth rate in     LICs.  7. How are current rates of urbanization in LICs different from historical     patterns? 66 Background
110
111 8. Explain the significance of the concepts of:      a. Missing children      b. Partial demographic transition      c. Demographic dividend.  9. Explain the concepts and inter-relationships of market segmentation and     dualism. 10. Elaborate on the concept of dualism by using conditions in any two markets     in LICs as examples. 11. How does dualism interface with rural-urban migration in the LIC context?     Identify and explain the main “pull” and “push” factors. 12. Explain rural-urban migration in L/LMICs paying particular attention to     private and social costs and benefits. 13. Explain why basing notions of economic development on the experience of     current HICs could be problematic. 14. Explain why being “late developers” could be a mixed blessing. 15. Explain the concept of “kicking away the ladder”. 16. Explain how the mechanisms of surplus transfer differ under colonialism and     neocolonialism. 17. Cite and explain the structural mechanisms by which colonialism is viewed     as having resulted in arrested or disarticulated development. 18. Cite and explain the institutional mechanisms by which colonialism is     viewed as having resulted in arrested or disarticulated development. 19. Explain how the current global conditions make catch-up economic growth     particularly challenging. 20. Use the World Development Indicators to construct your own Appendix     Table 3.1 by picking different variables. 21. Use the World Development Indicators to construct your own Appendix     Tables 3.3 by picking a different region.
112
113Notes  1 Since currencies can fluctuate a great deal when flexible, the World Bank uses a three-    year moving average (i.e. average of the current and preceding two years) to adjust for    such fluctuations.  2 Unless otherwise specified, the reference to $ in this textbook will be to US$.  3 More technically, a PPP ratio between say Kenya and the United States would be a    representative basket of commodities weighted by local prices in the numerator for    Kenya and denominator for the United States. The problem of course is that since    consumption patterns vary, finding a comparable basket of commodities is not straight-    forward and neither is finding accurate and relevant prices in an L/LMIC country    context. Setting these problems aside, dividing the per capita income converted by the    PPP exchange rate with the per capita income converted with the market exchange rate    provides an exchange rate deviation index. It measures the extent to which the market    exchange rate conversion understates the true purchasing power.  4 As indicated earlier in Chapter 1, endnote 29, since GDP in HICs is much larger than in    MICs or LMICs, in absolute terms industry and manufacturing are still much larger in    HICs than in MICs or LMICs.                                         Commonalities and differences among L/LMICs 67
114
115 5 This difference and the differential contribution by country income group to the stock    of greenhouse gases that trap heat in the atmosphere is why L/LMICs have argued that    HICs have a greater responsibility to address global climate change.  6 Data for the more populous South Asian countries are reported in Appendix Table    3.3. The other South Asian countries are Bhutan and the Maldives. Some lists include    Afghanistan as part of South Asia while others count it as part of Central Asia.  7 Unlike East Asian countries like South Korea and Taiwan, it was unable to sustain    the catch-up momentum and why that was the case should become clear from the    Chapter 9 discussion about what the East Asian countries did right compared to Latin    American, South Asian and African countries.  8 The debate about what accounted for India’s relative success is discussed in Chapter 8.  9 Bangladesh is considered one of the most climate vulnerable countries of the world    and it has already experienced sea level rise, salt water intrusion and extreme climate    events such as hurricanes. It will need the sustained high growth rates to adapt to    climate change. 10 One notable achievement for Pakistan might be containing its poverty rate to 4 percent    of the population based on World Bank estimates of poverty headcount ratio at $1.90    a day (2011 PPP). This could result from the computerized safety nets such as the    Benazir Income Support Program and the work of rural development NGOs. However,    based on the National Poverty Line which is more stringent, its performance in this    regard at 24.3 percent is like other comparator South Asian countries. 11 Given local knowledge, all the numbers tell a story. For example, rural electrification at    99 percent in Pakistan even exceeded that of Sri Lanka and is exceptional for an LMIC    which has an average of 68.1 percent rural electrification rate (Appendix Table 3.1).    Yet, the background story is one that involves Independent Power Producers, generous    government contracts and unsustainable fiscal deficits. 12 Recall the “rule of 72” from Chapter 1. A population growth rate of 3 percent would    mean a doubling of the population in less than a quarter of a century. 13 Population dependency ratios measure dependent population in a household i.e. those    less than 15 or greater than 64 relative to the working population between 16 and 64.    This is a crude measure, as is the case for most indicators, since children below 15    work and there is no concept of retirement for the poor so the elderly keep working    until they are not able to. 14 This may not be the case for the affluent who have servants for childcare and women do    not need to work, particularly in rural areas where working represents a loss of status.    Highly educated urban women from affluent backgrounds may choose to work as a    lifestyle choice. 15 Singapore experimented with such incentives. Positive monetary and other incentives    for having more children are also used in countries (e.g. France, Germany and Hungary)    where population growth rates have slowed and are expected to reduce the size of the    future labor force. 16 While our focus is on the economic dualism, there is a corresponding social dualism    reflected in different qualities of health care, education and social services in general    for the rich relative to the poor. 17 These are referred to as arbitrage opportunities where after accounting for transportation    and transaction costs, a trader may still earn a profit by buying in the market where the    price is low and selling where it is high. The implicit assumptions are that reasonable    communication, storage and transportation infrastructures are in place. 18 There is much more money to be made across the board with new projects than in    maintenance or the protection of existing social and physical infrastructure. Hence,    LICs suffer inordinately from the lack of maintenance. 19 These market failures created the opportunity for the introduction of microcredit    (Chapter 4). 20 Cypher (2014) is unique among development economics textbooks for devoting a whole    chapter to history including an excellent account of colonialism. Citing sources, he points 68 Background
116
117     out that at its peak in 1914, 84.4 percent of the global land mass had been colonized      and only Arabia (some parts), Turkey, Siam (Thailand), Persia (Iran), Mongolia, China,      Nepal and Tibet escaped overt colonialism. 21   In the political economy of development approach, societies reproduce themselves      economically (think of this as stable GDP). Anything beyond this economic      reproduction is surplus. 22   For more recent scholarship of the drain of resources from India, refer to Patnaik (2017). 23   For example, Dreze and Sen (2013, pp. 3–7) document this for India. Recent studies      on colonialism are nuanced and rigorously test the impact of various mechanisms      on the negative impact of the colonial legacy on future development outcomes. For      example, Angeles and Nianidis (2015) document a positive association of the degree      or density of European settlement in the colonies and corruption, Bruhn and Gallego      (2012) document an inverse association of future development to the nature of colonial      activities (good, bad, ugly) and Iyer (2010) relates the negative impact of colonialism      on social and physical infrastructure based on whether the rule was direct or indirect. 24   The simple definition of the external terms of trade is the price of exports relative      to the price of imports (using price indices). One of the most prominent debates in      development economics concerns whether the terms of trade for L/LMICs has been      declining over time (Chapter 7). 25   Rounds are periods of negotiations from one treaty to the next. The Doha Development      Round started being negotiated in 2001 and is still being negotiated. The Uruguay      Round, which was negotiated prior to the Doha Round, took the longest to negotiate      up to that point (eight years). One view is that L/LMICs signed away so much against      their interests in the Uruguay Round that they became resistant to HIC pressure for      more giveaways. In turn, HICs are turning to bi-lateral and regional trade agreements      to secure more trade liberalization. 26   http://thebricspost.com/imf-reforms-china-india-brazil-russia-get-greater-say/#.      WBXlMS0rK1s, consulted 1/20/2019. 27   However, power is not easily conceded and the United States initiating a trade war      against China in 2018 is emblematic of such geo-political tussle. 28   Interestingly Chang, originally from South Korea, pointed out that once South Korea      joined the OECD (Organization for Economic Cooperation and Development), the rich      country club, it even more aggressively opposed the policies it had used to catch-up      than other rich countries. 29   This phenomenon has been referred to as the Dutch Disease, a term coined by The      Economist magazine. Due to the natural gas finds in Holland in 1959, the Dutch guilder      appreciated and set deindustrialization into motion as the price of Dutch manufactured      exports rose in world markets. 30   Quotes suggestive of Japanese and later Korean laziness prior to their catch-up growth      were documented by Basu (2015, pp. 62–63). The point he was making is that norms      can be endogenous and hence can change. 31   The most prominent such thesis was due to Rostow (Chapter 5). 32   Frugal innovation has now become a well-known concept. Refer to http://      frugalinnovationhub.com/en/, consulted 1/20/2019. 33   Thus a full demographic transition is when both birth and fertility rates decline in tandem. 34   This phenomenon of virtuous circles (or cycles) has also been referred to by Myrdal (see      Chapter 7) as positive cumulative causation or positive path dependency. The obverse i.e.      negative cumulative causation or path dependency represents a vicious circle. 35   The timing has to be right for the virtuous circle to come into motion. Even so, countries      have sometimes successfully attempted to tap into its expatriate talent. Examples      include Taiwan in the 1960s and China in the 1990s. The United Nations Development      Program’s TOKTEN (Transfer of Knowledge through Expatriate Nationals) is a small      initiative for instituting a brain gain after the drain (Khan, 2012). 36   Basu (1999), using the case of child labor, showed how complex such issues are and      how policy should depend on context. APPENDIX TABLE 3.1 Selected defining characteristics by country income group
118
119Country income level                                                           LIC           LMIC      MIC        UMIC       HIC                                                              Economy – structure and indicators
120
121GDP per capita, constant 2010 US$, 2017                                       719.3          2,189.1   4,992.0    8,224.5    41,583.7 GDP per capita, PPP constant 2011, international $, 2017                      1,888.2        6,551.1   11,092.2   16,319.7   43,053.8
122
123Agriculture, forestry and fisheries value added per worker (constant, 2010    557.6          2,059.7   3,356.3    6,044.7    39,024.1 US$), 2016 Agriculture, forestry and fisheries value added, % of GDP, 2016               26.3           15.6      8.8        6.8        1.3 Industry, value added (including construction) per worker, 2016               3,675.1        7,543.5   16,690.8   23,736.1   90,990.1 Industry, value added (including construction), % of GDP, 2016                29.7           28.0      31.6       32.6       22.9 Manufacturing, value added, % of GDP, 2016                                    9.3            15.7      18.9       19.9       14.3 Services, value added, % of GDP, 2016                                         39.2           49.3      54.0       55.3       69.6 Services, value added per worker (constant, 2010 US$), 2016                   3,020.8        6,859.2   11,532.2   14,272.7   80,759.4 Access to electricity, % of rural population, 2016                            12.2           68.1      78.3       98.7       99.9 Access to electricity, urban (% of urban population), 2016                    67.6           94.9      97.6       99.8       99.9 Energy use, kg of oil equivalent per capita, 2014                                     na     646.3     1,395.8    2,204.2    4,637.5 Gross capital formation (I), % of GDP, 2016                                   23.5           26.7      30.6       31.8       21.1 Gross domestic savings (% of GDP), 2016                                       9.0            23.7      31.4       33.7       22.4 Cost of business start-up procedures, % of GNI per capita, 2017               70.1           24.0      22.7       21.6       4.6 Time required to start a business (days), 2017                                23.8           20.9      23.9       26.7       10.7
124
125                                                                                                                                   (Continued) APPENDIX TABLE 3.1 (Cont.)
126
127Country income level                                                          LIC            LMIC         MIC         UMIC        HIC External balance on goods and services, % of GDP, 2016                        −18.1         −3.6          −0.1        1.2         1.2
128
129                                                                   Technology indicators Mobile cellular subscriptions (per 100 people), 2017                          62.1          98.1          104.3       111.5       125.2 Individuals using the internet (% of population), 2016                        13.5          29.9          41.8        55.3        81.7 Fixed broadband subscriptions (per 100 people), 2017                          0.8           2.4           10.9        20.9        32.3 Scientific and technical journal articles, 2016*                              5,439.8       172,760       886,566.6   713,399.1   1,411,362 Patent applications, residents, 2016                                                  na    20,575        1,287,039   1,266,464   841,458                                                     Social investments and human development indicators Domestic general government health expenditure (% of GDP), 2015               1.2           1.3           2.8         3.3         7.7 Births attended by skilled health staff, % of total, 2014                     58.6          74.5          82.9        97.6        99.0 People using at least basic sanitation services (% of population), 2015       29.1          52.9          66.0        81.0        99.2 People using at least basic drinking water services (% of population), 2015   56.1          85.3          90.2        95.7        99.5 Life expectancy at birth, female, years, 2016                                 64.7          69.8          73.5        77.6        83.1 Life expectancy at birth, male, years, 2016                                   61.1          66.1          69.0        73.1        77.8 Fertility rates, births per woman, 2016                                       4.6           2.8           2.3         1.8         1.7 Maternal mortality ratio, per 100 000 live births, modeled estimate, 2015     479           257           180         41          13 Mortality rate, under-5, female (per 1,000), 2017                             64            46            34          13          5 Mortality rate, under-5, male (per 1,000), 2017                               74            50            38          15          6 Mortality rate, infant, female, per 1,000 live births, 2017                   44            34            26          11          4 Country income level                                                                 LIC               LMIC             MIC              UMIC             HIC Mortality rate, infant, male, per 1,000 live births, 2017                            53               39               30                13               5 Government expenditure on education, total (% of GDP), 2014                          3.8              4.3              4.5               4.5              5.2 Literacy rates, adult female, % for 15 years and above, 2016                         52.8             69.8             81.6              93.4             na Literacy rates, adult male, % for 15 years and above, 2016                           68.6             82.8             89.6              95.5             na School enrollment secondary (net), %, 2017                                           33.6             59.4             67.4              81.7             92.5 School enrollment tertiary (gross), %, 2017                                          8.8              24.4             35.6              52.1             77.1 School enrollment tertiary, gender parity index (female to male ratio), 2017         0.62             1.01             1.11              1.19             1.25                                                                                Pollution CO2 emissions (metric tons per capita), 2014                                         0.32             1.47             3.87              6.59             10.71
130
131Sources: World Development Bank development indicators on line, update 1/30/2019. Notes: The World Bank still reports data for LMIC, MIC and UMIC as reported in the table above. However, its 2017–2018 income classification has merged MIC and UMIC. Author attempts to get the cut-off between MIC and UMIC from the World Bank statistical department were not successful.     LIC = Low income countries (PCPPPUS$ = $995 or less)     LMIC = Lower middle income countries (PCPPPPUS$ = $996 to $4,125)     UMIC = Upper middle Income countries (PCPPPUS$ = $4,126 to $12,745)     HIC = High income countries (PCPPPUS$ = $12,746 or more)     na      = Data not available     *       = Classified by the Science Citation Index or the Social Science Citation Index APPENDIX TABLE 3.2 Selected defining characteristics of L/LMICs countries by region
132
133Region                                                                                   EA&P            LA&C            ME&NA             SA            SSA GDP per capita, constant 2010 US$, 2017                                                   6,137.9          8,889.0         4,301.2         1,779.7        1,651.4 GDP per capita growth (annual %), Average 2010–2017                                           6.6              1.1             1.0             5.5            1.1 Agriculture, forestry and fisheries value added per worker (constant, 2010                4,153.9          5,790.1         5,916.5         1,511.5        1,326.9 US$), 2016 Agriculture, forestry and fisheries value added, % of GDP, 2017                              8.7               4.7            9.5             16.0          16.0 Industry, value added (including construction) per worker, 2016                         19,905.1          23,507.2       22,129.5          5,635.6      10,920.0 Industry, value added (including construction), % of GDP, 2017                              39.6              23.3           34.1             25.5          25.2 Manufacturing, value added, % of GDP, 2017                                                  27.6              12.7           16.6             14.8          10.1 Services, value added per worker (constant, 2010 US$), 2016                              9,918.4          18,062.6       14,822.7          6,772.4       7,629.7 Services, value added, % of GDP, 2016                                                       51.2              61.6           52.4             49.7          52.2 Access to electricity, % of rural population, 2016                                          93.8              94.0           94.7             79.5          24.7 Access to electricity, urban (% of urban population), 2016                                  98.8              99.4           99.7             98.1          75.7 Energy use, kg of oil equivalent per capita, 2014                                        1,877.4           1,226.0        1,477.1            576.1         686.2 Gross capital formation (I), % of GDP, 2017                                                 40.3              18.6           29.8             29.4          20.4 Gross domestic savings (% of GDP), 2017                                                     43.6              18.3           24.8             26.8          18.7 Cost of business start-up procedures, % of GNI per capita, 2018                             21.8              49.5           30.9             11.0          45.1 Time required to start a business (days), 2018                                              30.1              34.5           24.6             13.8          23.2 External balance on goods and services, % of GDP, 2016                                       1.8              −0.7           −5.3             −4.3          −1.2 Human development index, 2017                                                                 .73               .76            .76              .64           .54
134
135Sources: World Development Bank development indicators on line, update 1/30/2019. For the human development index, UNDP, Human Development Indices and Indicators, 2018 Statistical Update, http://hdr.undp.org/sites/default/files/2018_human_ development_statistical_update.pdf, p. 49, consulted, 2/2/2019.
136
137    Notes: na   =   Data not available     EA&P        =   East Asia and the Pacific     LA&C        =   Latin America and the Caribbean     ME&NA       =   Middle East and North Africa     SA          =   South Asia     SSA         =   Sub Saharan Africa APPENDIX TABLE 3.3 Selected defining characteristics of South Asian countries
138
139Countries                                                               Bangladesh   India     Nepal     Pakistan   Sri Lanka GDP per capita, constant 2010 US$, 2017                                 1,093.0      1,964.6    732.4    1,222.5     3,849.5 GDP per capita growth (annual %), average 2010–2017                         5.3          6.0       3.2       2.0         5.0 Agriculture, forestry and fisheries value added per worker (constant,     990.7      1,678.8    549.7    1,699.8     2,532.8 2010 US$), 2017 Agriculture, forestry and fisheries value added, % of GDP, 2017            13.4        15.5      26.2      22.9          7.7 Industry, value added (including construction) per worker, 2017         4,067.6      6,386.8   2,131.0   2,904.6    10,340.1 Industry, value added (including construction), % of GDP, 2017             27.8        26.3      13.4      17.9         27.4 Manufacturing, value added, % of GDP, 2017                                 17.3        15.1        5.1     12.0         15.9 Services, value added per worker (constant, 2010 US$), 2017             3,674.5      7,729.8   3,118.3   5,707.9    11,726.3 Services, value added, % of GDP, 2017                                      53.5        48.7      51.6      53.1         55.7 Access to electricity, % of rural population, 2016                         68.8        77.6      85.2      98.8         94.6 Access to electricity, urban (% of urban population), 2016                 94.0        98.4      94.5      99.7       100 Energy use, kg of oil equivalent per capita, 2014                         222.2       637.4     412.7     484.4       515.7 Gross capital formation (I), % of GDP, 2017                                30.5        30.6      45.7      16.0         36.7 Gross domestic savings (% of GDP), 2017                                    25.3        29.5       11.9       6.8        29.5 Cost of business start-up procedures, % of GNI per capita, 2018            21.2        14.4      22.2        6.8         9.4 Time required to start a business (days), 2018                             19.5        16.5      16.5      16.5          9
140
141                                                                                                                      (Continued) APPENDIX TABLE 3.3 (Cont.)
142
143Countries                                                                               Bangladesh         India             Nepal              Pakistan        Sri Lanka External balance on goods and services, % of GDP, 2016                                      −5.2              −2.9             −33.8               −9.3              −7.8 Poverty headcount ratio at national poverty lines (% of population)                         24.3              21.9               25.2              24.3              4.1                                                                                           (2016)            (2012)             (2010)            (2015)            (2016) Poverty headcount ratio at $1.90 a day (2011 PPP) (% of population),                        14.8              21.2               15.0               4.0              0.7 World Bank estimate                                                                       (2016)            (2012)             (2016)            (2015)            (2016) Human development index, 2017                                                                 .61               .64                .57               .56              .77
144
145Sources: World Development Bank development indicators on line, update 1/30/2019. Notes: For the human development index, UNDP, Human Development Indices and Indicators, 2018 Statistical Update, http://hdr.undp.org/sites/default/files/2018_human_ development_statistical_update.pdf, p. 49, consulted, 2/2/2019. For poverty estimates, parentheses contain the latest year the data were available. The data used for each selected variable in all the appendix tables are the latest available in the online World Bank World Development Indicators. In Appendix Table 3.1, the income variable is used to classify countries into LICs, LMICs, MICs and HIC to explore commonalities by income classification. The main finding is that economic and social indicators vary positively with income classification. Thus HIC do much better on all indicators than LICs as expected but there is also a linear pattern by income grouping across all variables.                                        Commonalities and differences among L/LMICs 75
146
147References Acemoglu, D. and J. A. Robinson. 2012. The Origins of Power, Prosperity, and Poverty:    Why Nations Fail (New York: Crown Business). Alam, M. S. 1999. “Does Sovereignty Matter for Economic Growth? An Analysis of    Growth Rates between 1870-1950,” in: J. Adams and F. Pigliaru (eds.), Economic    Growth and Change (Cheltenham, UK: Edward Elgar) 46-70. Amsden, A. H. 2008. Escape from Empire: The Developing World’s Journey through    Heaven and Hell (Cambridge, MA: The MIT Press). Angeles, L. and K. C. Nianidis. 2015. “The Persistent Effects of Colonialism on    Corruption,” Economica, 82(326), 319–349. Baran, P. 1957. The Political Economy of Growth (New York: Monthly Review Press). Basu, K. 1999. “Child Labor: Cause, Consequence, and Cure, with Remarks on International    Labor Standards,” Journal of Economic Literature, 37(3), 1083–1119. Basu, K. 2015. An Economist in the Real World: The Art of Policy Making in India    (Cambridge, MA: The MIT Press). Bauer, P. T. 1984. Reality and Rhetoric: Studies in the Economics of Development    (Cambridge, MA: Harvard University Press). Becker, G. S. and H. G. Lewis. 1973. “Interaction between Quantity and Quality of    Children,” Journal of Political Economy, 81(2), S279–S288. Bruhn, M. and F. A. Gallego. 2012. “Good, Bad, and Ugly Colonial Activities: Do They    Matter for Economic Development?” The Review of Economics and Statistics, 94(2),    433–461. Caldwell, J. C., Barkat-e-Khuda, B. Caldwell, I. Pieris and P. Caldwell. 1999. “The    Bangladesh Fertility Decline,” Population and Development Review, 25(3), 67–84. Chang, H.-J. 2002. Kicking Away the Ladder: Development Strategy in Historical Perspective    (London: Anthem Press). Cleland, J., J. F. Phillips, S. Amin and G. M. Kamal. 1994. The Determinants of Reproductive    Change in Bangladesh: Success in a Challenging Environment (Washington, DC: World    Bank). Cypher, J. M. 2014. The Process of Economic Development 4th ed. (London: Routledge). Dreze, J. and A. Sen. 2013. An Uncertain Glory: India and Its Contradictions (Princeton,    NJ: Princeton University Press). Ebenstein, A. 2010. “The ‘Missing Girls’ of China and the Unintended Consequences of    the One Child Policy,” Journal of Human Resources, 45(1), 87–115. Gerschenkron, A. 1962. Economic Backwardness in Historical Perspective (Cambridge,    MA: Harvard University Press). Gunatilleke, G. 2000. “Sri Lanka’s Social Achievements and Challenges,” in: D. Ghai    (ed.), Social Development and Public Policy: A Study of Some Successful Experiences    (New York: St. Martin’s Press/MacMillan Press) 139-189. Iyer, L. 2010. “Direct Verses Indirect Colonial Rule in India: Long-Term Consequences,”    The Review of Economics and Statistics, 92(4), 693. Khan, S. R. 2011. “Growth Diagnostics: Explaining Pakistan’s Lagging Economic    Growth,” Global Economy Journal, 11(4), 1–17. Khan, S. R. 2012. “Highly Educated Labor Flows from Low to High Income Countries:    Mutual Security,” in: K. Khory (ed.), Global Migration: Challenges in the Twenty-    First Century (New York: Palgrave MacMillan) 87-112. Kravis, I. B. 1986. “The Three Faces of the International Comparison Project,” The World    Bank Research Observer, 1(1), 3–26. 76 Background
148
149La Porta, R. and A. Shleifer. 2014. “Informality and Development,” Journal of Economic    Perspectives, 28(3), 109–126. Lipton, M. 1977. Why Poor People Stay Poor: Urban Bias in World Development    (Cambridge, MA: Harvard University Press). Papanek, G. F. 1967. Pakistan’s Development: Social Goals and Private Incentives    (Cambridge, MA: Harvard University Press). Patnaik, U. 2017. “Revisiting the ‘Drain’, or Transfer from India to Britain in the Context    of Global Diffusion of Capitalism,” in: S. Chakrabarti and U. Patnaik (eds.), Agrarian    and Other Histories: Essays for Binay Bhushan Chaudhuri (New Delhi: Tulika Books)    278-317. Preston, S. H. 1979. “Urban Growth in Developing Countries: A Demographic Reappraisal,”    Population and Development Review, 5(2), 195–215. Rhee, Y. W. 1990. “The Catalyst Model of Development: Lessons from Bangladesh’s    Success with Garment Exports,” World Development, 18(2), 333–346. Rostow, W. W. 1960. The Stages of Economic Growth: A Non-Communist Manifesto.    Cambridge: Cambridge University Press. Sen, A. K. 1990. “More Than 100 Million Women Are Missing,” New York Review of    Books, December 20. Taylor, L. 1994. “Gap Models,” Journal of Development Economics, 45(1), 17–34. 4 POVERTY, INEQUALITY AND SOME PROPOSED SOLUTIONS
150
151Introduction This chapter addresses the two key problems of L/LMICs that the approaches reviewed in Part II address directly or indirectly. The neoliberal and developmentalist approaches for the most part seek to mitigate these problems by initiating catch-up growth. The political economy and basic human needs/human development approaches differ in directly addressing these problems and hence they get more play in this chapter.     The current state of knowledge on absolute poverty is investigated first. Various definitions of poverty such as income, basic needs, risk and vulnerability poverty are explained. The findings of research that characterize poverty in L/LMICs are presented next, following which inequality is investigated, including policy debates on whether inequality should be addressed. Finally, policy initiatives proffered to address various forms of poverty are discussed and since these are targeted at the poor, social inequality could also be simultaneously addressed by them.
152
153Absolute poverty National poverty lines The concept of absolute poverty pertains to how many people in a given geographic entity such as a nation state or regions within a nation state are in a state of poverty. To be able to make such a statement, poverty first needs to be defined. Based on a consensus definition, a national poverty line can then be identified. Following that, everyone that falls below the line could be considered to be in absolute poverty in a particular nation. This is why this way of measuring poverty is also known as headcount poverty since it entails identifying the percentage of the people in a given population who are in a state of poverty based on a given definition of poverty.1     The concept of a national poverty line is based on the amount of income needed to meet a minimum caloric intake for a specified period. For example, one estimate for a minimum caloric intake specified by nutritionists for an adult male (greater than 14 years of age) for a given day is 2,250. Nutritional needs of other 78    Background
154
155individuals are determined based on adult equivalents (women 2/3 and children 1/2) of adult male caloric needs.     The next step is to convert a typical basket of commodities that would provide the specified calories using nutritionist conversion chart (food units to calories). The typical basket is then converted to local income equivalents in local currency. Finally, based on survey data, the numbers of people that fall below the income needed for specified calories for a specified time period (say one year) are considered to be in poverty. These estimates are possible for various categories such as region, or urban/rural and gender bifurcations.     Recall the data problems identified in Chapter 2. In addition to those, there are problems specific to constructing national poverty lines. Survey information is usually collected at the household level and then converted to the individual level based on assumptions about intra-household food distribution and adult equivalency. However, because that information on within household inequality is generally not available, estimates of poverty may be underestimated.     Since women often work both within and outside the home, researchers challenge the validity of the sex and adult equivalency assumptions specifying lower requirements for women than men. Beyond that, caloric needs vary greatly based for example on metabolism, body size and activity levels. Thus, averages in poverty research are more likely than usual to mask important variations. Again, at given levels of expenditure, caloric intake could vary significantly due to household consumption patterns. While calories are an important energy measure, critics have pointed out that other nutrient needs such as proteins, minerals, vitamins and micro nutrients also need to be accounted for.     The required information on consumption is collected by asking respondents to recall their consumption over a specified time period. The longer the time period for recall required of respondents, the greater the likely inaccuracy in the data. Poverty research suggests that a one-month recall systematically understated poverty compared to surveys that use a one-week recall.     There may be rural-urban and regional differences not captured by the numbers. For example, there may be a larger component of non-market goods and services in rural transactions that are not captured by survey data and hence rural poverty in this regard could be overstated. This bias could be reinforced by not taking into account that a culture of sharing is more predominant in rural areas.     Headcount poverty estimates per se do not take into account either the duration of poverty or the poverty gap. The duration refers to the time an average household spends below the poverty line. Policy makers have to be concerned about the stability (relative to mobility) below the poverty line. Mobility is desired because it suggests opportunities for movement above the poverty threshold. Such concerns have resulted in distinctions such as transient vs. chronic (long duration) poverty.     The poverty gap is the ratio by which the mean income of the poor falls below the poverty line. The head count index does not take into account how far below the poverty line the mean income of the poor falls. Thus even if policy is successful in ameliorating the condition of the “poorest of the poor” (those in                                    Poverty, inequality and some proposed solutions 79
156
157extreme poverty), this may not show up in the poverty statistics if there is a big gap between the poverty line and the mean income of the poorest group. Hence a smaller poverty gap for a specified poverty line is preferable. Statistics are now reported when possible on both duration and the poverty gap.    Appendix Table 4.1 shows the state of poverty in LICs based on the national poverty line and the international poverty lines. The World Bank uses international poverty lines for global estimates of extreme poverty (headcount ratio at $1.90 per day, 2011, PPP, percentage population) and moderate poverty (headcount ratio at $3.20 per day, 2011, PPP, percentage population). Also reported are poverty gaps based on the national and international poverty lines.    As expected, poverty varies a great deal within the LIC group. Even so, 78 percent of the population in Madagascar was in extreme poverty by the World Bank definition. In 20 out of the 29 countries that data were reported for in various years, about a third of the population or more was in extreme poverty by the World Bank definition. Using the National Poverty Lines, in 20 out of the 28 countries that data were reported for in various years about two-fifths or more of the population was in extreme poverty, four-fifths in South Sudan. The average percentage of the population in LICs in extreme poverty using the World Bank and National Poverty Lines were 47.1 and 43 percent respectively. The averages for LICs were much higher than the averages for LMICs at 28 percent and 14.6 percent respectively. Thus moving to higher income status is associated with a decline in average poverty rates.    The average poverty gaps for both the World Bank and National Poverty lines for LICs were 16.2 percent, again much higher than LMIC. The size of these gaps suggests that in most cases a major effort would be required to move populations out of extreme poverty. While these numbers suggest a dire situation, the method of estimating poverty lines, particularly the international poverty lines, are highly controversial, however.2 For this reason, scholars have turned to other methods to explore the reality of extreme poverty confronted by populations in L/LMICs.
158
159Basic needs poverty The basic needs approach mentioned in Chapter 1 found income poverty to be grossly inadequate as a measure of the state of deprivation and in keeping with its approach to development advocated a more holistic method of characterizing and measuring poverty. According to this approach, income poverty does not directly take into account other dimensions of well-being such as health, educational attainment and access to shelter and clothing. One way to estimate basic needs poverty is therefore to find the minimum amount of income that would provide for the basic needs of an average household.    The multi-dimensional poverty index proposed by Alkire et al. (2011) quantifies basic human needs poverty. The index measures deprivation by the unit of analysis, say an individual or household, on various sets of indicators of well- being such as education, health and living standards. The indicators are weighted 80    Background
160
161and individuals or households are deemed to be in multi-dimensional poverty if the deprivation scores fall below a cross-dimensional poverty cut-off.
162
163Risk and vulnerability poverty The critique on data collection methods explored in Chapter 2 applies to measuring poverty and well-being. Recall that sample surveys are an instrument for estimating income poverty and that such data were roundly criticized by those with little faith in survey methods. As critiques developed on the drawbacks of sample surveys, alternative qualitative methods for gathering information emerged as explained. One particular alternative technique was the use of participatory approaches that enabled the poor to define poverty themselves and identify those who by their own criteria were in poverty.    The World Bank Research Department picked up on this initiative, as they often do with innovations emerging from academia. This can add value because, given their ample resources, they can often push such innovations much further than the innovators. In this case, they conducted a large number of participatory poverty appraisals in a number of L/LMICs. Two volumes were published in 1999 (40,000 responses in Vol. 1 and 20,000 in Vol. 2 for 23 countries) with the effort led by Narayan et al. (2000) who edited the responses in volumes titled Voices of the Poor.    Some findings that emerged from these studies were that the poor were very preoccupied with their political exclusion and lack of “voice”.3 Further, there was a preoccupation with the risk and vulnerability they confronted in their daily life. Such risk and vulnerability could be associated with for example natural disasters, political violence, feudal oppression and crime. The lack of ability to cope with shocks and covariate risks4 is another way of viewing risk and vulnerability poverty.    Another finding of this study was that coping mechanisms could perpetuate poverty. The poor sensibly diversify their activities as a form of insurance. However, growing both food and cash crops and engaging in informal sector activities provides insurance but may not maximize income. Cash crops may provide the highest income, but because of possible crop failure, this may also be a high-risk strategy.    When confronting shocks, coping mechanisms might increase long-run poverty. For example, if short-term household survival entails cutting back on education and health expenditure, it perpetuates inter-generational poverty. Another form of coping is loans from within the village (shopkeepers, friends, moneylenders), but shocks often hit the village across the board and so in that case this coping mechanism is nullified.
164
165Environmental degradation poverty Observation rather than deep research is enough to fathom that the poor are the most vulnerable to environmental depredations. Pedestrians and cyclists are the most exposed to auto-emission and other toxic fumes. Rich motorists can leave behind the smog of inner cities as they drive to relatively clean residential areas or                                     Poverty, inequality and some proposed solutions 81
166
167they can roll up the windows and turn on the air-conditioners if caught in traffic jams for some relief.     Again, low-income neighborhoods mushroom around industrial areas as property values fall and because of jobs, and they are hence the most exposed to dirty air, bad odors and effluents. It should thus not be surprising that poor children have a higher incidence of asthma, other respiratory diseases and lead poisoning.5 Thus those who are the most disadvantaged are made more so. The probability of escaping poverty or enhancing the range of choices for children is thus reduced.     The poor, particularly women, are also more exposed to indoor pollution due to the use of biomass (biological material) for cooking and indoor heating (Ezzati and Kammen, 2005). Again, at work they spend long hours in polluted environments. The poor are also the most exposed to diseases such as diarrhea and dysentery due to the lack of access to clean water and sanitation (Seguin and Nino- Zarazua, 2011). Low nutritional intake also means less resistance to such diseases.     Poor uneducated farmers are more exposed to the detrimental effects of chemical agricultural inputs. Excess use, drifting sprays, leaky applicators and the lack of knowledge of handling equipment and dangerous substances enhances the risk of ill health. Similarly, women and children engaged for example in cotton picking, the cash crop for which pesticide use is the greatest, are exposed to numerous ailments. These include carcinogenic diseases, enzyme imbalances, skin and allergic reactions, lung diseases, sterility, cataracts, memory loss, change in the central nervous system and damage to the immune system(Devine and Furlong, 2007; chapter 13). For all these reasons, a clean environment needs to be viewed as part of an anti-poverty strategy.
168
169Other conceptualizations of poverty Mullainathan and Shafir (2013) and Mani et al. (2013) originated the concept of bandwidth poverty. Scarcity results in the poor obsessing about what they lack and if this is food, much of their thinking bandwidth is occupied by hunger and food. This negatively impacts cognitive functioning that would enable them to devise strategies to overcome the scarcity, another example of a vicious circle whereby say indebtedness leads to further indebtedness. The authors tested this hypothesis on data collected for sugarcane farmers in India using a natural experiment whereby they were able to compare the cognitive functioning of the same individuals who confronted scarcity and relative plenty. The cognitive functioning of farmers not having to contend with scarcity during harvest periods improved relative to their cognitive functions in times of scarcity.    For mainstream economists, opportunity costs drive decision making for individuals. Very often, hourly wages foregone would be the relevant cost of putting one hour of time toward another activity. This may understate the true opportunity cost of time for the poor relative to those who have adequate income for basic needs. Due to diminishing marginal utility, the marginal utility from income from one hour of work foregone for the poor is much greater than the same amount of income for 82    Background
170
171the more comfortable. This could explain why social organizers find it so difficult to get participation from the poor in poverty alleviation initiatives. For those existing at or near a poverty threshold, the opportunity cost of their time is much higher than those who are more comfortable. This would be true even if the time is not spent in gainful employment but hustling to find ways to make ends meet.
172
173The incidence of poverty Poverty research in L/LMICs over decades has identified the incidence of poverty and the results are as one might expect. It is disproportionately present in rural areas and the urban bias identified in Chapter 3 has much to do with this. Within urban areas, the bulk of poverty is within shanty towns. Within households, poverty varies positively with the size of household, and females and children are more susceptible than adult men to poverty-related illnesses.     Standard mainstream household economics assumes that the head of the household represents the interests of the whole family in a utility maximization framework (Becker, 1981). Sen (1990) challenged this assumption and argued that there could be both conflict and cooperation within a household. This focus on household decision making revealed that there can be a systematic bias in the allocation of household goods (such as food) and resources (such as health and education expenditures) in favor of males and this is reflected in a higher incidence of female poverty.     Research also shows a higher incidence of poverty in female headed households. On average they are less educated so they get the lowest paying jobs. However, they also suffer from discrimination and so for the same education, skills and experience they get paid less, and this discrimination in L/LMICs is more intense than in HICs. In agriculture in particular, women also have less access to resources like land and capital (Chapter 13). There is also an asymmetry in the introduction of technology and its diffusion since most technology is introduced to ease the burden of the more highly paid male tasks such as tractors for plowing. If the women headed households belong to an ethnic minority the exclusion and intensity of poverty is even greater (Chapter 13).
174
175Inequality Causes A great deal of attention is paid to inequality since it may have more to do with social instability than poverty. The idea is that people react more to their deprivation if others are prosperous or living in luxury. Inequality is also viewed as self-reinforcing as the rich parley their economic power to get political power that further consolidates their economic power in an ongoing circle.    Due to the social and political instability, as explained below, it can adversely impact economic growth and hence breed more inequality as the rich are more capable of surviving in difficult circumstances.                                         Poverty, inequality and some proposed solutions 83
176
177    Inequality needs to be distinguished from inequity. The latter is suggestive of unfairness as judged by some theory of justice. Inequality may be based on inequity (social exploitation) but it need not be. Also, promoting equality may result in inequity – taking people’s due share for redistribution if the initial distribution was just.6     Neoclassical economic analysis takes the existing distribution of income as a given and implicitly defends it by defending existing property rights. Heterodox approaches like the political economy of development or neo-Marxist approaches start their analysis by questioning the justness of the existing distribution of income and often advocate for redistribution if the existing distribution represents accumulation based on social injustice perpetrated either by colonialism or predation based on a concentration of political power.     Neoclassical analysis of the distribution of income is based on marginal productivity theory as reviewed in introductory economics. As indicated above, this theory takes the existing distribution of income/assets as a given and then shows it to be just, given certain assumptions including perfect competition or constant returns to scale technology and full employment. Based on these assumptions, factors of production get a share of total income according to their contribution to this income. Sensible employers will hire factors equal to the contribution of these factors on the margin to income.     Assume two factors (labor and capital) to enable graphical representation.7 W is the market wage determined by the market demand and supply curve for labor in Figure 4.1 below.     In a competitive labor market, each firm has to pay the market determined wage. Given this wage and a downward sloping marginal productivity curve (diminishing returns given capital is fixed) each firm hires labor up to the point where W = MPL (marginal product of labor). The area under the demand for labor curve, up to where labor (L1) is hired, represents the total product for the firm in Figure 4.1a.8 Labor gets W*L1 (wage bill) and the rest goes to capital.     The implication of this model is that distribution resulting from free market production is completely fair because all factors get what they contribute to production and the sum of factor contributions exhausts total product (adding
178
179                             a: Firm                         b: Market           W                                      W
180
181                                                                     SL
182
183          W                                      W
184
185                 W*L1                                        MPL                                                                      DL                         L1                   L                                 L FIGURES 4.1a and b   Marginal productivity theory 84    Background
186
187up theorem). This theory has been critiqued on several grounds by heterodox scholars including by Moseley (2012).     Much else goes into the production process other than the factors of production. For example, companies like Apple that represent the epitome of the success of market capitalism benefitted from a great deal of government support of research. Thus, not only does this theory take the existing distribution of income as a given, it also takes the existing institutional structure as a given and does not acknowledge its contribution to productivity and production.     Consider a student from Nigeria who earns an electronic engineering degree at the Massachusetts Institute of Technology as does her US roommate and friend from San Francisco. Suppose also that her GPA in graduate school is exactly the same as that of her US friend. However, while she returns to Nigeria her friend gets a job in Silicon Valley at a starting salary that is ten times higher than what she is offered.9 Marginal productivity theory would suggest that the student from the United States is ten times more productive. However, we already know that is not the case and so why the salary differential? The answer is that complementary factors like research facilities, access to university facilities, a critical mass of colleagues and connectivity account greatly for the productivity differential. Thus, productivity is not just inherent to the person but also to the environment in which they work and much of that is provided by the state and the community of the country in question.     Recall that for the theory to hold the markets must be competitive and full employment must hold. If for example, employers have disproportionate labor market power, they can pay workers less than their marginal product and in introductory micro theory this is referred to as monopsony power. If there is unemployment, this would also give firms more leverage.     The theory also assumes that capital (K) and labor (L) are abstract entities and can be sensibly added up for aggregate analysis. Labor is heterogeneous (unskilled, semi-skilled, skilled, professional, technical, managerial) so aggregation is difficult. In fact, much of the social inequality is due to income differences across different categories of labor and so this theory assumes away the problem of inequality. The aggregation of capital is even more problematic and for heterodox scholars, without such aggregation, it is therefore not possible to compute marginal products.10     Also recall the assumption of constant returns to scale technology. If the technology results in increasing returns to scale, the efficient firms will necessary grow over time since expansion reduces costs and hence the firm will acquire market power. This violates the perfect competition assumption and hence the presumption that returns to the different factors exhaust the total product. Even in neoclassical analysis, acquiring market power on the selling side (monopoly/ oligopoly) or the buying side (monopsony) can result in unfair outcomes i.e. the exploitation of the consumer or of labor respectively.     Marx (Chapter 5) viewed the exploitation of labor by employers as the normal state of play in the capitalist system. Capitalists are viewed as having leverage over workers by virtue of owning capital or the means of production (factories/workplaces). This enables capital to hire workers and set the remuneration below labor’s contribution                                     Poverty, inequality and some proposed solutions 85
188
189if there is high unemployment.11 High unemployment was assured by what he called the reserve army of the unemployed and one capitalist strategy was to ensure the refurbishing of this reserve army by moving to ever more capital intensive techniques of production12 and moving female and child labor into the workforce.     In such labor market conditions (with monopsony power) capital could get away with merely paying a subsistence wage i.e. one considered enough to enable labor to maintain and reproduce labor as a class. The remaining contribution of labor is then appropriated by capital.13 For example if workers work eight hours (16–18 in Marx’s day) and they produce enough to reproduce the household (subsistence wage) with four hours work, capital will hold on to the other four hours. As the unemployment rate falls, capital’s leverage over labor declines.14     Two other reasons why the adding up theorem may not hold is intermediate products and joint production. The contribution of intermediate product to total output is not accounted for when assuming only factors of production contribute to output and holding intermediate inputs constant to compute the marginal product of a particular factor is not logical. Joint products result when a set of input leads to more than one output due to commercial byproducts such as wood chips in the production of furniture.
190
191Measurement Measuring inequality within a population requires knowledge of the size distribution of income within that population. This means a frequency distribution of the percentages of population that fit into different income categories. A hypothetical frequency or size distribution is shown in Table 4.1 using a fictitious currency.    Based on the fictitious distribution of income in Table 4.1, it is possible to get a graphical account of the state of income inequality for a particular population. For
192
193      TABLE 4.1 Hypothetical discrete size distribution of income
194
195       Income category                       Percentage of population in income        (Galoses (Gs) per month)              group        <10,000                               0.400        10,001–20,000                         0.200        20,001–30,000                         0.150        30,001–40,000                         0.100        40,001–50,000                         0.050        50,001–60,000                         0.040        60,001–70,000                         0.035        70,001–80,000                         0.015        80,001–90,000                         0.007        90,001–100,000                        0.002        >100,000                              0.001 86    Background
196
197example, we know that 90 percent of the population earns less than 50,000 Gs per month while 10 percent earns greater than 50,000 Gs. Another way to look at it is that only .01 percent of the population earns greater than 100,000 Gs. Either way, the data tell us that there is a high level of income inequality in the population in question. However, it is possible to go further than this in characterizing the nature of inequality in a given country.     The size distribution of income above is called discrete since the population distribution is based on discrete income groups rather than continuous income. Imagine even more granular data so that the income distribution is continuous such that we know exactly what percentage of the income is garnered by each percentage of the population. Such a continuous size distribution of income, admittedly highly demanding of data, can be used to construct measures of income distribution and inequality.     In 1905, the statistician Max O. Lorenz proposed a way to represent inequality in a two-dimensional graph by charting the percentage of population on one axis and the percentage of the variable of interest on the other axis and the distribution that results provides a visual account of the inequality for the variable in question. This tool could be utilized to represent for example inequality in income, wealth or a component of wealth such as land holdings if data are available.     To represent income inequality, we put the percentage of the population on the horizontal axis and the percentage of income on the vertical axis. Perfect equality in a population would be represented by a 45 degree line i.e. 50 percent of the population getting 50 percentage of income as shown in Figure 4.2. Inequality would be present if 50 percent of the population earned less than 50 percent of the total income. The greater the distance below the 45 degree line, the greater the income inequality for every given population point. Joining all such points, the curve that results is named after Lorenz and it is a graphical representation of the extent of income inequality in a population.15 Thus the way to view inequality visually is to look at the depth of the arc the income line or Lorenz curve makes relative to the 45 degree line or the line of equality. The deeper the arc, the greater is the income inequality in a population being studied, as shown in Figure 4.2.     It is possible to go still further in characterizing inequality. Instead of relying on a visual observation, the Lorenz curve can be used to compute a summary statistic that has been developed as a measure of inequality. Since the arc (A) in Figure 4.2 represents the extent of inequality, that divided by the total area under the triangle (A + B) is a summary statistic representing the extent of income inequality referred to as the Gini coefficient. The Gini coefficient varies from 0 to 1; the coefficient is 0 when there is no arc (perfect equality, A = 0 i.e. 0/B = 0) and 1 when there is perfect inequality (i.e. B is zero; A/A = 1). Perfect inequality means that in the limit one individual gets all the income and everyone else gets none.     The great advantage of the Gini coefficient is that it is conceptually straightforward and can in a simple way represent income and other forms of inequality in a population and also makes possible comparisons across populations. However, if the relevant Lorenz curves intersect in a comparison of two countries                                         Poverty, inequality and some proposed solutions 87
198
199          100%
200
201        Income
202
203                        45 degree line of equality
204
205                                                     A
206
207                                                          B
208
209                                                         Population                                                                       100 %
210
211FIGURE 4.2   Lorenz curve
212
213          100%
214
215        Income
216
217                        45 degree line of equality
218
219                                                     1
220
221                                                          2
222
223                                                         Population                                                                       100 %
224
225FIGURE 4.3   Intersecting Lorenz curves
226
227or of one country over time, it is then possible to get the same Gini coefficients with very different income distributions, as shown in Figure 4.3.    Figure 4.3 shows two Lorenz curves for country 1 and country 2 (or time periods 1 and 2 within one country). Assume that they yield the same Gini coefficient for two countries. If there is more interest in what is happening to the lower half of the population, then distribution 1 would be preferable even though the Gini for the two distributions are the same. Another related problem is that the Gini coefficient can be scale insensitive. Small changes at the lower end of the distribution may mean much for reducing absolute poverty (push many people close to or above the poverty line), but these changes may not register much of a change in the Gini coefficient.    A more technical problem is that the Gini coefficient cannot be decomposed. Thus the inequality within and between groups should, but does not, add up to the Gini. Several more sophisticated measures such as the Atkinson or Theil index do 88    Background
228
229not have this shortcoming but they have other shortcomings and as such there is no perfect measure. Despite the caveats mentioned above, the Gini remains the most widely used due to its simplicity.     Simpler suggestive measures of inequality can be computed with data on total income by population groupings like deciles or quintiles. For example, one can calculate the ratio of what the top 1 or 10 percent claim relative to the bottom 20 or 40 percent. The ratios chosen are arbitrary, but nonetheless indicative and the specific ratios chosen can vary depending on the point the researcher wants to make.16     LICs are for the most part hybrid market systems with a mix of developmentalist interventionism and neoliberal economic philosophy that became dominant starting in the 1980s. The inequality results reported below have emerged from this mix. To provide context, ratios of selected HICs representing different models of capitalism are also reported. Sweden is selected to represent a social democratic model with a focus on social solidarity and hence high marginal tax rates and generous social safety nets. Korea represents East Asian corporate capitalism based on an interventionist state.17 Notwithstanding the focus on the business sector and economic growth, buy in from the public depended on building such a model on a foundation of equity, as was the case in Japan, which it emulated (Chapter 9). Both these HICs can therefore be expected to have moderate income inequality.     The United States represents the Anglo-Saxon model of market-driven capitalism with think tanks supporting the Republican Party pushing for individual liberty and entrepreneurship based on low tax rates and low regulation. This economic philosophy is also the basis of neoliberalism, which pushes for vibrant economic growth and considers the prosperity created by the capitalists as likely to trickle down to the rest of the population.18 In this case, one would expect higher social inequality than in the social democrat or corporatist capitalist models.     There is no pure model of capitalism and political pulls and pushes constantly change leanings. For example, Sweden has become notably more conservative, and while the Democratic Party in the United States favors the market system and entrepreneurship, there is much more emphasis on state supported social justice and equity when it is in power. Appendix Table 4.2 reports Gini indices and income share ratios for LICs and for context those of LMICs, Sweden, Korea and the United States.     The average Gini coefficient for LICs of 40 is considered modest and only a couple of countries have a Gini above 50, which is considered to be high inequality. The average of 37 for LMIC is lower but not substantially so. However, the Gini index in LICs is almost twice that of Sweden and 10 points higher than South Korea. As reflected by the Gini index, the average state of income inequality in LICs, which has been under the tutelage of World Bank/IMF-led neo liberalism since the early 1980s, mirrors that of the United States (41.5).     The income share ratio (income share held by top 10 percent to bottom 20 percent) is high at 5.6 but mostly this is pushed up by a few outliers (Benin, Guinea-Bissau, Mozambique and South Sudan). Once again, it is twice that of Sweden and much higher than Korea (3.3) but this time below that of the United                                     Poverty, inequality and some proposed solutions 89
230
231States (6.1). Barring the same outliers, in most cases the income share of the top 10 percent is about the same as that held by the bottom 40 percent.     Given the evidence reported above regarding the high levels of social inequality and some evidence suggesting that this can disrupt economic growth (see below), there is an instrumental economic reason to be concerned assuming that a social philosophy of justice and equity is not enough to drive policy. The policy debates on addressing inequality in development economics are as old as the field itself and ongoing.
232
233Debates on policy to address inequality National advocacy for policy change can be strengthened using comparative global data. PPP/PCGNP is one component of UNDP’s (United Nations Development Program) Human Development Index (HDI – the other two components being education and health). The HDI could be made inequality sensitive by using Gini coefficients as weights for PPP/PCGNP. The way to do this is to subtract the Gini coefficient from 1 and use the residual as a deflator of PPP/PCGNP. The higher the income inequality, the higher the deflation of the PPP/PCGNP and the lower its inequality adjusted ranking. Such a hypothetical exercise is shown in Table 4.2.     As indicated, the higher the Gini (income inequality), the higher the deflation of PCGNP for global ranking. Country 1 dominates country 2 in a straight comparison of PPP/PCGNP, but once inequality is taken into account, the adjusted PPP/PCGNP in Table 4.2 shows country 2 delivering a higher level of well-being to its population.     The size distribution of income can also be used to deconstruct economic growth rates for policy advocacy.19 The simple hypothetical exercises below show that PCGDP growth is an even worse measure of well-being than commonly assumed. As normally measured, PCGDP implicitly weights the income of the more prosperous by more, simply because the higher income group has a higher income share. Consider the hypothetical example constructed in Table 4.3 for an illustration of this point.     Suppose the upper 20 percent gets 50 percent of the income and the second quintile gets 30 percent. Suppose the income growth rate of the upper quintile is 10 percent and that of the second quintile 3 percent, while it is 0 percent for the remaining three quintiles. The PCGDP growth rate (G1) as usually calculated would be the weighted average G1 = .10 × .50 + .03 × .3 = 5.09 percent (i.e. growth rates by income group weighted by the income share in the groups). An alternative would be to use population weights based on the justification that each person should have equivalent weight (neutral). In this case each income group would have the same weight of .2 and so the weighted average growth rate would be G2 = .10 × .20 + .03 × .20 = 2.06 percent. Attaching high normative weights to the poorest quintile would produce a third PCGDP growth rate estimate. Positive high growth is recorded above despite the 0 growth of the bottom 60 percent. If poverty was important as a social objective, one could, for example, have weights as indicated in the last column of Table 4.3. In that case the PCGDP weighted average growth estimate would be G3 = .10 × .05 + .03 × .10 = .08 percent. 90          Background
234
235                     TABLE 4.2 Deflating PPP/PCGNP for inequality
236
237                         Country   PPP/GNP      Gini     Inequality adjusted                                                          PPP/PCGNP                          A         4,010        .57      .43 × 4,010 = 1,724                          B         3,760        .42      .58 × 3,760 = 2,180
238
239     TABLE 4.3 Hypothetical weighted per capita GDP growth rates
240
241      Income groups                    Growth          Income     Population   Poverty                                        rates           weights    weights      weights       I                                0.00            0.02       0.2          0.40       II                               0.00            0.05       0.2          0.30       III                              0.00            0.13       0.2          0.15       IV                               0.03            0.30       0.2          0.10       V                                0.10            0.50       0.2          0.05       Weighted growth rate (%)                         5.09       2.06         0.08
242
243    The “let the market work” perspective of mainstream neoclassical economists is to focus on achieving economic growth, and the presumption is that this will address the poverty and inequality problems as prosperity starts to trickle down with more jobs created by a growing economy. The empirical case for this rested on showing that there is a trade-off between growth and inequality. In other words, if policy makers made addressing inequality a priority, they might jeopardize economic growth in which case such a policy would backfire.     Kuznets, (1955) was the pioneer in exploring the association of economic growth and inequality, although he did so at a time when data were very limited. Based on some simulations and the economic history of the United States, Germany and England, he hypothesized that the relation between growth and inequality was an inverted-U as shown below.     The inverted-U in Figure 4.4 shows that as PCGDP increases due to economic growth on the horizontal axis, inequality represented by the Gini first increases and then decreases as society continues to prosper. Kuznets strongly qualified his findings and conceded in his paper that they were “perhaps 5 percent empirical information and 95 percent speculation, some of it possibly tainted with wishful thinking”.     Nonetheless, he suggested as a possible stylized fact of economic development that inequality initially increases with economic growth, but as growth continues inequality eventually declines. Kuznets himself was concerned with identifying a possible empirical regularity based on long-term time series data for the United States, Britain and Germany rather than with establishing a case for trickle down. This was an inference from his findings that other researchers drew; scholars who were much less careful than he was.                                     Poverty, inequality and some proposed solutions 91
244
245          Gini
246
247                                                                      PCGDP
248
249FIGURE 4.4   Kuznets curve
250
251    Kuznets proposed a few causal mechanisms for the inverted-U based on market and policy mechanisms. At the beginning of catch-up growth for current HICs, the immigration of workers from low income rural areas to higher income urban ones could initially increase inequality since less prosperous people come into a prosperous area. As such immigration continues, this effect could reverse itself as urban areas prosper and inequality decreases. Further, at the early stage of economic growth, dependency ratios (Chapter 3) are high and rising among the poor and this could contribute to the inequality. With more prosperity, this process reverses itself as the demographic transition kicks in (i.e. as income and education rise and fertility rates decline).     Following neoclassical factor proportions theory, initially migration enhances the return to capital in urban areas because capital is scarce and labor abundant. However, with continuing economic growth the labor force is absorbed and capital becomes relatively abundant and labor relatively scarce. Following the logic of this theory, the returns to capital fall while the productivity and hence remuneration of labor, which has more capital on average to work with, rises. Thus, the economic growth process diffuses prosperity and reduces social inequality.     Policy mechanisms such as the provision of public education reinforce this trend. As the education and skill level of the labor force rises, the consequent higher productivity also results in higher wages. Fertility rates decline with higher education because the opportunity cost of having children increases for women who go into the workforce (Chapter 3). Also, as societies become more prosperous, they can afford to provide subsidies and reduce social inequalities that persist.     Kuznets’ findings promoted a spate of research and controversy that is continuing to this day. Using some measure of inequality, like the Gini, as the left-hand side variable, researchers showed that PCGNP had a strong positive association with inequality and that the association followed an inverted-U shape as suggested by Kuznet i.e. that inequality first increased with catch-up growth and then started to decline after some threshold level of per capita GDP was attained.20 92    Background
252
253    Most of these studies were based on cross-country regressions with country as the unit of analysis.21 Thus these studies implicitly assumed that the countries in the sample were structurally similar (same mechanisms operated in similar ways). Heterodox economists challenged this implicit assumption by arguing that the countries in the sample were obviously structurally dissimilar given that the samples included a diversity of countries to establish the existence of the Kuznets curve. Others provided counter evidence while using the cross-country method. Thus many researchers doubt that the Kuznets curve has the status of an empirical regularity. There are now more observations for time series analysis but the evidence once again is not conclusive.     Heterodox economists who use the inductive case study method argue that there really are no commonalities and that case studies can demonstrate many different patterns and also that policy can change patterns. For example, Brazil represented a classic case of high growth-high inequality. However, under the leadership of Luiz Inácio Lula da Silva (2003–2011), through social programs in education, health, nutrition and other welfare programs, Brazil started reversing inequality and brought the income share ratio of the top 10 percent to bottom 20 percent down from to 17.5 when he took office in 2003 to 12.6 in 2011 when he left office and the average economic growth in per capita GDP during this period was 3.3 percent.22     Other counter examples include the East Asian Economies (Japan/Korea/Taiwan) all of which demonstrated that high growth and high equality are possible. Land reforms and high wage policies contributed to this. Based on socialist redistributive mechanisms, China is another counter example. In the early 1980s, its social and employment policies ensured high equality even as it started to experience rapid economic growth. A discontinuation of social policies caused inequality to rise with catch-up growth. The Gini in 1981 was 28.8 and subsequently rose to 42.2 in 2012.23     Some studies suggested a reverse causality and explored instead the impact of inequality on growth. Thus a cross-country study by Alesina and Perotti (1996) showed more equality could result in more political stability and hence more investment and economic growth.     To sum up, the justification of trickle down in the neoliberal approach is that the rich will save and facilitate capital accumulation and hence redistribution should be avoided. However, the rich in L/MICs do not necessarily save. Luxury imports, foreign travel, real estate investments and capital flight are all possible leakages from savings.     Alternatively, according to the basic human needs/human development approach, investment in education and health of the poor would make them more productive and hence enhance equality and economic growth. Further, since the poor spend more on the local product market this would enhance the market size and enable the attainment of economies of scale.     The upshot is that theory casts doubt on the mainstream trickle-down approach and the data also show many different patterns are evident based on the actual economic experience of different countries. The critical issue is the policy or redistributive mechanisms adopted by the state; something that ironically Kuznets himself recognized as important.24 Various poverty alleviation policy initiatives                                     Poverty, inequality and some proposed solutions 93
254
255are associated with the various definitions of poverty, and solutions proposed by the various approaches are reviewed in the next section.
256
257Poverty alleviation initiatives The different definitions of poverty have resulted in different policy initiatives associated with the different approaches to be investigated in detail in Part II. In keeping with the Kuznets curve, the mainstream/neoliberal approach is straightforward in that it endorses policy initiatives that it views as economic growth enhancing. While organizations like the World Bank and IMF, which enforce neoliberal prescriptions, have recently been more open to initiatives associated with the other approaches, economic growth is still center stage as the trickle- down mechanism to deliver widespread prosperity. A number of initiatives have been used in the late 20th and 21st centuries to address basic human needs/human development and risk and vulnerability poverty, and these are reviewed below.     The methodology to access the success of these initiatives is not straightforward since while costs are easy to monitor and document, the benefits are more elusive and difficult to quantify. For example, if it is a social investment initiative such as enhanced health or education, mainstream economists may try to quantify the benefits in the form of discounting to the present value the enhanced lifetime earnings. However, most recognize this is difficult to do accurately, particularly when the quality of the intervention is poor and may not enhance productivity, and labor markets are not perfect and much else may explain enhanced earnings. The human freedom approach in any case would place a much higher premium on the intrinsic value of the social investment rather than the instrumental value of higher earnings.     A focus on the intrinsic value provides a resolution to the difficulties in trying to estimate the economic rates of return to social investments. If policy makers choose what the social investment is to be, the cost effectiveness method quantifies the most efficient way of attaining that objective. This is less rigorous than a social benefit cost approach, but also more realistic. Various poverty alleviation initiatives have been devised and implemented over the decades and the important ones are reviewed below.
258
259Basic needs/human development The basic needs approach, as explained in Chapter 1, was a reaction to trickle- down economics. The argument was that it would take too long if ever for trickle-down economics to deliver well-being to the poor and that, in any case, basic human needs like education, health and water supply must be delivered by the state as a constitutional right. The World Bank under Robert McNamara’s leadership adopted this approach. This could well have been motivated by the intense ideological rivalry during the Cold War and the fear that the excluded would be targeted by and be susceptible to radical thought. 94    Background
260
261    Mahbub ul Haq, who served at the World Bank as chief economic advisor to Robert McNamara, championed the basic human needs approach. When this approach fell out of fashion after the Cold War ended it was reconstituted by Haq at the UNDP, with advice from A. K. Sen, as the Human Development approach. This approach gained traction partly due to the annual release of the Human Development Report that attempts to measure and rank countries based on the human development attained by its citizens in terms of health, education, sanitation, water supply and other basic human needs. Over time, the definition of development has been broadened to address gender inequality, regional inequality, urban/rural inequality and human freedoms, and metrics accordingly devised.     Another hallmark of the global success of the human development approach was in the framing of the Millennium Development Goals (MDGs) that all 189 member states of the UN signed on to at the turn of the century and hoped to achieve by 2015 with the help of the donor community. The goals included eradicating extreme poverty and hunger, achieving universal primary education, promoting gender equality, reducing child mortality, improving maternal health, combatting HIV/AIDS, malaria and other diseases, ensuring environment sustainability and developing a global partnership for development. Each objective had several targets and critics charged that overall the MDGs became too diffused and cumbersome (Fukuda-Parr, Yamin, and Greenstein, 2015).     The achievements were mixed and the MDGs morphed into Sustainable Development Goals (SDGs) as the objective for the next 15 years until 2030. Environmental sustainability became the central issue and all countries were included in the SDGs rather than just the L/MICs. In the transition from HDGs to SDGs, the need for quality in setting goals, targets (169 in all) and indicators (what is actually measured) was acknowledged.     In practice the rich have better opportunities since for example the best education is accessible to them and the poor are deprived of educational opportunities or attend poor quality government schools. Quality deteriorates when the rich leave public schooling and there is no voice for improvement. Poor public schooling deprives the poor of a key mechanism for social mobility and non-public services (private and NGO) have a very low bar to cross to be better. The same applies to other public services and the result has been the mushrooming of poor quality private sector services as a substitute for public services. This has been encouraged by the World Bank, even with subsidies, as the state readily abandons its key constitutional functions of providing basic human needs. This education and health dualism reflects the broader social dualism in society.
262
263Integrated rural development Based on the work of development economists like Lewis and Harris and Todaro (Chapter 5), it was acknowledged that the bulk of the labor force in L/LMICs resided in rural areas. Further, the high incidence of rural poverty and that a lack of rural development resulted from urban bias, which in turn led to rural to urban migration                                     Poverty, inequality and some proposed solutions 95
264
265with its attendant problems of shanty towns and the associated intensification of crime and urban poverty. A prescription following on from this research in the 1970s was for the provision of a comprehensive package of infrastructure, training, agricultural inputs and credit directed at the rural poor. This was a state-centered and delivered approach and failed for the most part because the landed elite appropriated all the subsidies (elite capture). One notable exception was Korea’s Saemaul Undong (cooperative) movement and the success has been attributed to the land reforms that pre-dated the rural development initiative (Park, 2009).     Land concentration leads to a concentration of political and hence economic power and elite capture and exclusion in poverty alleviation initiatives. Thus land distribution in Korea prior to the rural development initiative diffused economic power and negated elite capture. However, as expected, there is a great deal of political resistance to land redistribution and mainstream economists are wedded to property rights and reluctant to push this issue.     Since the 1980s rural development NGOs have inherited the mantle of rural development and generally engage in an overtly non-political fashion. The initiatives, referred to as participatory, for the most part rely on organizing and mobilizing grassroots communities to engage in their own development via collective action with much of the resources provided by the international donor community. This became a way for the latter to bypass the state, which is deemed to be ineffective and corrupt, and in providing service delivery directly to the poor.
266
267Microcredit One key instrument adopted by rural development NGOs as a poverty alleviation initiative is microcredit. Mohammed Yunus started the Grameen Bank to deliver microcredit on a small scale with his own money in 1983 when he was unable to persuade public sector banks to lend to the poor. Yunus was trained as an economist in the United States and recognized the lack of rural banking serving the poor as a market failure. The nature of the market failure is that rural private or public sector banks confront high information costs (asymmetrical information or lack of knowledge of the poor as credit risks), high transaction costs (too many small loans mean high administrative costs), lack of access (poor rural infrastructure) and no or low collateral that the poor can offer.     The lack of collateral was the most important and Yunus overcame this by relying on peer group pressure/social capital such that in small groups the poor, based on mutual contacts, networks and trust vouched for each other and were willing to incur the penalty of no loans if peers defaulted. The poor therefore became the mechanism to ensure there was a repository of accurate information and reduced transactions costs via self-policing.     The focus soon shifted to women who were the most marginalized and better credit risk, and repayment rates were as high as 98 percent.25 This basic model has been adopted and adapted by about 60 countries. Some banks turn a profit while most need to be subsidized. However, this subsidy, as a form of public policy to 96    Background
268
269reduce risk and vulnerability poverty, may be well justified if it compares well to the cost of other poverty alleviation initiatives.     There is evidence from numerous earlier impact assessment studies suggesting that microcredit enabled households to engage in consumption smoothing (transfer from surplus to lean periods), engage in high risk–high return options, improve household health and education, and empower women (Pitt, Khandker, and Cartwright, 2006; Khandker and Samad, 2014). Other research based on RCTs was more skeptical and a special issue of the American Economic Review, Applied Economics (2015, 7(1), p. 3) contained six randomized evaluations and the summary suggested a “lack of evidence of transformative effects [of microcredit] on average borrowers”.     In the 1990s, the World Bank led a shifting of focus from the original poverty alleviation of borrowers to the sustainability of the organization. For-profit, including publicly traded companies, were encouraged to enter this new market. Yunus was so distressed at this trend that he equated it with loan sharking. A spate of suicides in India due to the harsh outstanding loan collection tactics of “for-profit” microcredit led to a Parliamentary inquiry. However, it appears that neoliberalism has won this battle and “for-profit” microfinance persisted. Apart from microcredit, microfinance includes new products including consumer credit, saving accounts, insurance, pensions and remittance services.     While once viewed as a panacea, many critiques of microfinance emerged with the feminist critique as the most telling (Goetz and Gupta, 1996; Parmar, 2003). Feminists critiqued this market-friendly initiative as one that is exploitative of women because the latter are used as a front to secure loans for males and confront abuse from husbands if they fail to secure loans. Moreover, they confront repayment pressure from abusive loan officers and social pressure from peers including interference in their personal life such as with regards to their consumption pattern. Instead of building social capital, such pressure undermines and destroys social capital. Women also confront a higher work burden since husbands do not help with household chores even if women’s workload increases due to the credit they receive.     Feminists define social empowerment as real when associated with transformation in the power configuration in gender relations in households and gender equality in society. Instead, advocates of microcredit define female empowerment functionally as greater mobility, access to funds or say in the use of funds. Another feminist critique is that via WB/IMF structural adjustment (Chapter 8), the burden of care has been pushed even more onto individuals with the state abdicating its role. Microcredit is touted as a stand-in for state services and hence implicitly facilitates structural adjustment and hence is very much part of the neoliberal world view and package of reforms (Rankin, 2001, 2002).     Other critics have shown that microcredit programs bypass the poorest of the poor who are not considered bankable by peers and rural banking staff. They criticize the loan treadmill that microcredit puts the poor on (new loans paying for old ones), especially now that the organizations view borrowers as profit units. Due to organizational saturation, there is often loan recycling such that lenders borrow from one organization to pay another.                                    Poverty, inequality and some proposed solutions 97
270
271   Critics also argue that the claimed success of microcredit is subject to a “fallacy of composition” in that there is a limit to how many crafts of a particular kind can be sold in a particular market i.e. there is market saturation. Thus what might work for one borrower will not work for all in a community. Thus borrowers are locked in a zero-sum game where the success of one borrower is premised on the failure of another.26    For critics like Chang (2011) microcredit and the self-employment it promotes simply cannot provide the needed employment opportunities, the best form of poverty alleviation, or the kind of technological progress industrialization can deliver. Hence they view privileging microcredit as an economically destructive diversion of scarce funds from economic activities with higher social and economic rates of return.
272
273Employment Guarantee Schemes An Employment Guarantee Scheme (EGS) was initiated as a rural sector poverty alleviation initiative by the Government of Maharashtra in India in the 1970s. The basic idea was “self-revelation”. The wage for temporary work was kept a bit below the minimum market wage so that only the poorest would reveal themselves for such jobs. This would limit leakages to those not entitled and reduce the administrative expense of targeting. Also, the EGS projects could help to develop valuable rural infrastructure to raise productivity.    The challenge, as with all public projects, is keeping down administrative costs and limiting corruption. The Congress government in India in 2006 scaled up the EGS to the national level and guaranteed 100 days employment to all the rural poor. There is controversy regarding its effectiveness but early findings in the development economics literature suggests it is a qualified success (Jha et al., 2015; Breitkreuz et al., 2017).
274
275Conditional Cash Transfer27 Conditional Cash Transfer (CCT) initiatives started in the 1990s with Bolsa Familia in Brazil and Progresa (now Oportunidades) in Mexico among the most famous of the almost two dozen such initiatives in L/MICs. The idea is to provide cash to households conditioned on the household taking responsibility for enhancing social indicators within the household with the assistance provided. For example, women may need to agree to participate in health care workshops or nutritional assessment programs (child weight and height measurements) or ensure that household children attend school.     Feminist critics charge that CCTs place additional burdens on women’s time that is already overloaded with responsibilities. Heterodox scholars view this initiative as yet another in the neoliberal social policy bag of tricks to ameliorate the harsh outcomes of the austerity conditions associated with economic reforms and a forced abdication of state responsibility for its citizens. 98    Background
276
277Unconditional cash transfers Some neoclassical economists have suggested that cash without conditions would be a better way to enable the households since relaxing the budget constraint and providing choice enhances utility in a utility maximization framework and that households know what is best for them.     Others argue for unconditional assistance based on a rights-based approach i.e. that basic human needs are a fundamental human right. Unconditional cash transfers (UCT), in the form of a Universal Basic Income, has also garnered the attention of Silicon Valley libertarians on the right and social democrats on the left as a form of a social safety net in response to the threat to unskilled jobs due to automation in HICs (Chapter 14).     Silicon Valley entrepreneurs have funded multi-year (12) RCTs in Kenya to explore if individuals become lazy and less motivated or if instead they make good use of the funds provided. Social democrats are less concerned about incentive effects and more about the right to a decent life. Initial research by Haushofer and Shapiro (2017) suggests a positive impact on psychological well-being and economic outcomes. Finland and Canada have also trialed pilot projects to test outcomes of UCT.
278
279BRAC graduation initiative This is a multi-dimensional initiative devised by Building Resources Across Communities (BRAC)28 that views poverty as multi-dimensional and hence draws on other initiatives for a multi-dimensional solution. The concept is to provide a “big push” to the household much as developmentalists argued that nations needed a big push (Chapter 5) to build a momentum that they could subsequently sustain.29 A multi-dimensional intervention directed at people to get them sustainably out of poverty includes an asset to make a living, such as livestock or an informal sector store, training on how to manage the asset, basic food or cash support to reduce the need to sell the asset in an emergency, frequent (usually weekly) coaching visits to reinforce skills, confidence building and help with handling challenges. In addition, health education or access to healthcare is provided to enable participants to stay healthy and not lose work time and a savings account to help them put away money to invest or use in a future emergency.    Interventions are tailored to the particular circumstances of the L/LMIC in question. Banerjee et al. (2015) conducted an RCT covering Ethiopia, Ghana, Honduras, India, Pakistan and Peru with a large sample including 21,063 adults in 10,495 rural households. The initiative was shown to be cost effective, with positive returns in five of the six countries they studied. The returns ranged from 133 percent in Ghana to 433 percent in India. Thus for every dollar spent on the program in India, the ultra-poor households got $4.33 in long-term benefits.    In terms of specifics, they reported that over a three-year period, there was a 5 percent increase in per capita income, 8 percent increase in food consumption, 15                                    Poverty, inequality and some proposed solutions 99
280
281percent increase in assets and a 96 percent increase in savings. While expensive and personnel intensive, the idea is that the poor can graduate from poverty with this initiative and hence in that regard the intervention is viewed as more cost effective than others that do not provide a sustained pathway out of poverty.
282
283Sustainable development Human development focuses on extending the range of choices for the poor. This includes sustainable livelihoods to ensure the capacity for making choices and human capital (education and health) to make sound choices over a longer period. The focus on the environment as central to human development has two underpinnings. First, as explained above, a poor environment is associated with negative health impacts to which the poor are most vulnerable. Second, a healthy environment ensures the health of not just current but also future generations.     Initially the concern was with sustainable human development, but this was broadened to sustainable development in recognition that the concern should be for all species. Fortunately, protecting humans by protecting the environment also protects other species, although a special focus is needed on endangered species.     Initially the focus was on the attainment of inter-generational justice. However, a concern with inter-generational justice brought to the fore the importance of intra-generational justice and hence the integration of poverty alleviation into the concept of sustainable development. Sustainable development is an umbrella concept and as such subsumes the initiatives discussed above. Recall that sustainable development has now been adopted by the UN as its key initiative and SDGs replaced MDGs in January 2016. Specific sustainable development initiatives for agriculture and industry are discussed in Chapters 13 and 14.
284
285Cross-cutting mechanisms in poverty alleviation initiatives Based on the operational experience of the initiatives above a number of cross- cutting mechanisms can be identified. Participation emerged as explained in Chapter 2 from the view that the poor should be empowered via the process of social mobilization, creation of grassroots organizations by tapping social capital and being involved in all stages of projects intended to benefit them (a move away from top-down mechanisms). Rural grassroots organizations of, by and for the poor would also enable them to engage in collective action. This would make use of the information that only the poor have i.e. what would benefit them the most and how best to execute the project to attain those benefits.    It was argued that if the poor are involved in all stages of the project from design to execution, they will be more engaged and likely to contribute to it and maintain it.30 The big challenge is conceiving such projects in a way that avoids elite capture. Rural grassroots NGOs do the social mobilization, often with donor support, although there are cases of spontaneous evolution of grassroots 100    Background
286
287organizations of the rural poor that engage in collective action (Khan, Kazmi, and Rifaqat, 2007, chapter 2).    The perceived success of this mechanism has also filtered into service delivery by the public sector. One mechanism for this has been public-rural NGO partnerships or the state even contracting out the service delivery to development NGOs. While the ideal is that the development NGOs would merely be catalysts in sparking social organizations of the poor, in practice, like all organizations, they have a stake in survival and do so as long as the funding is available. Some of these organizations remain true to their ideals while others are criticized for putting the organizational agenda, particularly that of the NGO management, center stage, and this is reflected in high salaries and perks, detachment from the grassroots and distrust by those they are supposed to serve.    Targeting is another mechanism for distribution of services to the poorest in rural areas. Recall that the challenge is to avert the power structure to avoid elite capture. In practice, the state administration works closely with the rural elites and so leakages are viewed as inevitable. Even if there were dedicated sections of line department (service delivery) bureaucracy, intent of effective service delivery, resources and administrative capacity are limited.    Evidence shows that targeting women leads to the greatest gains as they are most likely to invest in the nutrition, health and education of children. A study by Rubalcava, Teruel and Thomas (2009) shows that when the balance of power within the household changes in favor of women due to resource transfers, they are more likely than men in a similar situation to invest in the future of the household.    Finally, operational experience with poverty alleviation initiatives shows that there are inter-relationships between different kinds of provisions. For example, better education can ensure the ability to secure job opportunities. It can also ensure better health outcomes because the ability to process information improves and educated mothers ensure better nutrition and health of their children. Healthier children perform better at school and hence ensure the ability to secure livelihood opportunities later on. This could break the vicious circle of poverty.    Better health for adults can make them more productive and also lead to fewer working days lost and less medical expenses. Safety nets, say in the form of an EGS, enable individuals to pursue high risk–high profit options. Political freedom can ensure economic freedom as politicians respond to constituents. Based on these interactions, a holistic approach to service delivery, as for example with the BRAC graduation initiative, suggests the whole is much greater than the sum of the parts.
288
289Summary and conclusions This chapter reviews the concepts and methods of measuring absolute poverty and inequality. Income poverty, being below a defined poverty line in income terms, is the most common concept, though this has been challenged by broader multi- dimensional poverty concepts. Basic human needs/human development poverty and risk and vulnerability poverty are two such broader concepts reviewed.                                   Poverty, inequality and some proposed solutions 101
290
291    There are limitations to the mainstream competitive market approach in examining the issue of inequality. The Marxian or political economy of development approach attributes it to the leverage capital (business) has over workers by virtue of ownership of the means of production and hence of the ability of capital to capture surplus that is rightly due to labor. Neoclassical economics, when not assuming perfect competition and full employment, explains this concept in terms of the monopsony power businesses can exercise over workers.     There have been many initiatives to address basic human needs/human development and risk and vulnerability poverty. These include state provision of basic human needs as a right (rights-based approach), integrated rural development delivered by the state or more recently rural development NGOs, a state-led EGS, state-provided CCTs, Unconditional Cash Transfers and a development NGO- initiated graduation from poverty initiative.     The microcredit initiative, also mostly delivered by development NGOs, is viewed by critics as captured by the IMF/WB market-led development approach. This perspective views microcredit as a Trojan horse that pushes back the role of the state in poverty alleviation and makes the poor responsible for themselves.     More recently, the IMF/WB have recognized that the capital friendly structural adjustment they unleash on borrowing countries intensifies inequality and poverty and have welcomed the poverty alleviation initiatives identified above as safety nets.     These poverty alleviation initiatives can make some difference as a palliative. However, all these approaches are silent about the structural causes of poverty such as the skewed ownership of land and the leverage capital has over labor. In the political economy of development approach, real structural reform, as opposed to neoliberal-led structural adjustment, requires asset distribution, particularly land redistribution, so that the power structure is altered. Other key initiatives are improved tax administration to implement wealth and progressive income taxes and public sector reform to deliver quality public sector provision.     Another structural transformation reform would be devolution of service delivery to the local level once the concentration of power has been defused with land redistribution. Institutional changes like devolution could make the implementation of the rule of law more likely and corruption and elite capture less likely. From a political economy of development perspective, without fundamental structural reforms, the various initiatives discussed above may simply be a Band-Aid.     The human development approach captured the imagination of the international donor and development community and they encapsulated it in the MDGs embarked on for a 15-year period, 2000–2015. Recognizing that the poor are the most vulnerable to environmental degradation and therefore of the need for both intra- and inter-generational justice, the UN replaced the MDGs with the SDGs, 2016–2030. As an umbrella approach, the SDGs subsume all the poverty alleviation initiatives discussed in this chapter in addition to addressing the environmental challenges faced by the planet. However, as argued in Chapter 15, a more parsimonious approach to structural transformation that has greening industry and agriculture at the core might be more effective. APPENDIX TABLE 4.1 Poverty indicators in low income countries
292
293 Country                       Poverty headcount   Poverty headcount    Poverty headcount    Poverty gap at   Poverty gap at Poverty gap at                                ratio at national   ratio at $1.90 per   ratio at $3.20 per   national poverty $1.90 per day $3.20 per day                                poverty lines (%    day (2011, PPP, %    day (2011, PPP, %    lines (%)        (2011, PPP, %) (2011, PPP, %)                                of population)      population)          population)  Afghanistan                   35.8 (2011)         na                   na                   8.4 (2011)      na              na  Benin                         35.2 (2010)         49.6 (2015)          72.6 (2015)          9.8 (2011)      22.4 (2014)     39.5 (2015)  Burkina Faso                  40.1 (2015)         43.7 (2014)          76.4 (2014)          9.7 (2014)      11.1 (2013)     32.2 (2014)  Burundi                       64.9 (2014)         71.7 (2013)          89.2 (2013)          25.1 (2014)     30.3 (2013)     51.5 (2013)  Central African Republic      na                  na                   na                   na              na              na  Chad                          46.7 (2011)         38.4 (2011)          66.5 (2011)          19.7 (2011)     15.3 (2011)     30.8 (2011)  Comoros                       42.0 (2014)         18.1 (2013)          38.1 (2013)          na              6.3 (2013)      15.4 (2013)  Congo, Democratic Republic    63.9 (2012)         77.1 (2012)          91.3 (2012)          na              39.2 (2012)     58 (2012)  Eritrea                       na                  na                   na                   na              na              na  Ethiopia                      23.5 (2015)         26.7 (2015)          61.4 (2015)          7.8 (2010)      7.7 (2015)      22.7 (2015)  Gambia, The                   48.6 (2015)         10.1 (2015)          37.8 (2015)          27.9 (2010)     2.2 (2015)      10.8 (2015)  Guinea                        55.2 (2012)         35.3 (2012)          70.3 (2012)          18.4 (2012)     10.3 (2012)     28.4 (2012)  Guinea-Bissau                 69.3 (2010)         67.1 (2010)          84.5 (2010)          na              30.5 (2010)     49.6 (2010)  Haiti                         58.5 (2012)         23.5 (2012)          48.3 (2012)          24.4 (2012)     7.5 (2012)      19.2 (2012)  Korea, Democratic Peoples     na                  na                   na                   na              na              na  Republic  Liberia                       50.9 (2015)         38.6 (2014)          73.8 (2014)          na              11.7 (2014)     30.7 (2014)  Country                               Poverty headcount       Poverty headcount        Poverty headcount        Poverty gap at   Poverty gap at Poverty gap at                                        ratio at national       ratio at $1.90 per       ratio at $3.20 per       national poverty $1.90 per day $3.20 per day                                        poverty lines (%        day (2011, PPP, %        day (2011, PPP, %        lines (%)        (2011, PPP, %) (2011, PPP, %)                                        of population)          population)              population)  Madagascar                            70.7 (2012)             77.7 (2012)              91.0 (2012)              na                    39.0 (2012)         57.9 (2012)  Malawi                                51.5 (2016)             71.4 (2010)              88.8 (2010)              18.9 (2010)           33.6 (2010)         53.2 (2010)  Mali                                  41.1 (2009)             49.7 (2009)              79.4 (2009)              na                    15.5 (2009)         36.3 (2009)  Mozambique                            46.1 (2014)             62.9 (2014)              81.9 (2014)              na                    27.9 (2014)         46.6 (2014)  Nepal                                 25.2 (2010)             15.0 (2010)              50.8 (2010)              5.4 (2010)            3.1 (2010)          15.8 (2010)  Niger                                 44.5 (2014)             44.5 (2014)              76.9 (2014)              19.6 (2011)           13.5 (2014)         33.6 (2014)  Rwanda                                39.1 (2013)             56.0 (2013)              80.8 (2013)              14.8 (2010)           20.2 (2013)         40.7 (2013)  Senegal                               46.7 (2011)             38.0 (2011)              67.5 (2011)              14.5 (2010)           12.8 (2011)         29.6 (2011)  Sierra-Leone                          52.9 (2011)             52.2 (2011)              81.3 (2011)              16.1 (2011)           16.7 (2011)         38.0 (2011)  Somalia                               na                      na                       na                       na                    na                  na  South Sudan                           82.3 (2016)             42.7 (2009)              64.8 (2009)              23.7 (2009)           18.9 (2009)         33.3 (2009)  Syrian Arab Republic                  na                      na                       na                       na                    na                  na  Tajikistan                            31.3 (2015)             4.8 (2015)               20.3 (2015)              na                    1.0 (2015)          5.3 (2015)  Tanzania                              28.2 (2011)             49.1 (2011)              79.0 (2011)              6.7 (2011)            15.4 (2011)         36.3 (2011)  Togo                                  55.1 (2015)             42.9 (2015)              73.2 (2015)              24.4 (2011)           19.9 (2015)         37.1 (2015)  Uganda                                21.4 (2016)             41.6 (2016)              69.8 (2016)              5.2 (2012)            13.2 (2016)         31.2 (2016)  Yemen, Republic                       48.6 (2014)             18.8 (2014)              52.2 (2014)              na                    4.5 (2014)          17.3 (2014)  Zimbabwe                              na                      21.4 (2011)              42.7 (2011)              na                    5.2 (2011)          17.3 (2011)  Average LICs                          47.1 (14.9)             42.5 (20.6)              66.2 (18.2)              16.2 (7.4)            16.2 (11.0)         32.8 (14.2)  Average LMICs                         27.7 (15.8)             14.6 (15.6)              33.8 (24.5)              10.5 (10.2)           4.8 (6.7)           31.9 (23.7) Sources: For country classifications https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world bank-country-and-lending-groups, consulted 1/15/2019. For data, World Development Bank development indicators on line update 1/30/2019. Notes: The classification of LIC is for the 2019 fiscal year (July 1, 2018 to June 30, 2019).Parentheses contain the latest year the data were available for going back ten years.For the averages, the parentheses contain standard deviations.na = Not available 104        Background
294
295APPENDIX TABLE 4.2 Inequality indicators in LICs and selected HICs
296
297 Country                   Gini index (World   Ratio of income       Ratio of income                            Bank estimate)      share held by         share held by                                                the top 10% to        the top 10% to                                                bottom 20%            bottom 40%
298
299 Afghanistan               na                  na                    na  Benin (2015)              47.8                11.7                  1.6  Burkina Faso (2014)       35.3                3.6                   1.0  Burundi (2013)            38.6                4.5                   1.1  Central African           na                  na                    na  Republic  Chad (2011)               43.3                6.6                   1.2  Comoros (2013)            45.3                7.5                   1.3  Congo, Dem. Republic      42.1                5.8                   1.2  (2012)  Eritrea                   na                  na                    na  Ethiopia (2015)           39.1                4.8                   1.2  Gambia, The (2015)        35.9                3.9                   1.0  Guinea (2012)             33.7                3.5                   0.9  Guinea-Bissau (2010)      50.7                9.3                   1.8  Haiti (2012)              41.1                5.7                   1.1  Korea, Dem. Peoples       na                  na                    na  Republic  Liberia (2014)            33.2                3.4                   0.9  Madagascar (2012)         42.6                5.9                   1.3  Malawi (2010)             45.5                6.7                   1.4  Mali (2009)               33.0                3.2                   0.8  Mozambique (2014)         54.0                10.8                  2.1  Nepal (2010)              32.8                3.2                   0.9  Niger (2014)              34.3                3.5                   0.9  Rwanda (2013)             45.1                6.3                   1.5  Senegal (2011)            40.3                5.1                   1.1  Sierra-Leone (2011)       34.0                3.4                   0.9  Somalia                   na                  na                    na  South Sudan (2009)        46.3                8.5                   1.2  Syrian Arab Republic      na                  na                    na  Tajikistan (2015)         34.0                3.6                   0.9  Tanzania (2011)           37.8                4.2                   1.1  Togo (2015)               43.1                6.3                   1.2                                          Poverty, inequality and some proposed solutions 105
300
301 Uganda (2016)                    42.8                    5.6                    1.3  Yemen, Republic (2014)           36.7                    4.0                    1.0  Zimbabwe (2011)                  43.2                    5.8                    1.3  Average LICs                     40.4  (5.6)                            5.6  (2.2)                            1.2  (0.3)  Average LMICs                    37.2                    4.9                    1.0  Sweden (2015)                    22.9                    2.8                    0.7  South Korea (2012)               31.6                    3.3                    0.8  United States (2016)             41.5                    6.1                    1.1
302
303Sources: For country classifications https://datahelpdesk.worldbank.org/knowledgebase/articles/906519- world-bank-country-and-lending-groups, consulted 1/15/2019. For data, World Development Bank development indicators on line, update 1/30/2019. Columns 2 and 3 are author manipulations. Notes: The classification of LIC is for the 2019 fiscal year (July 1, 2018 to June 30, 2019).Parentheses contain the latest year the data were available for going back ten years.na = Not available
304
305Appendix 4.1: Environment Kuznets curve In the 1990s, the Kuznets association was applied to the study of environmental degradation. To illustrate an environment Kuznets curve (EKC), different pollutants could be measured on the vertical axis and once again, PCGDP on the horizontal axis. Some research has shown that the association once again is an inverted-U shape as with the Kuznets curve as shown in Appendix Figure 4.1.31    One mechanism in this case is that beyond some threshold level of income, YT as shown in Appendix Figure 4.1, the demand for a cleaner environment as a luxury good will increase and this would induce policy to work toward a clean environment. Thus, as with the income Kuznets curve, there is an element of automaticity built into the association suggesting that policy is endogenous and that no premature action is called for. Another mechanism suggested by Grossman and Krueger (1991) is that as nations prosper, the scale of production is polluting but this can be offset by the composition effect (move to cleaner services) and cleaner technology. Before the benign turn of the Kuznets curve, governments may have to confront very angry citizenry as evident from the experiences of China and India.32 In any case, given the evidence of the rapid increase in the pace of climate change, waiting for benign turns is no longer an option (Chapter 9, endnote 31).    However, for the sake of argument, even if the Kuznets association does hold for some pollutants and for some countries using time series data, ecologists urge caution. They argue that much biodiversity and ecosystem loss is 106    Background
306
307       Pollution        indicator(s)
308
309                                          YT                           PCGDP
310
311       APPENDIX FIGURE 4.1   Environment Kuznets curve
312
313non-reversible and so getting it wrong has a very high cost. Further, knowledge about the non-reversible threshold is limited and these thresholds can change and complex interactions within ecological systems are difficult to fathom. This complexity also suggests that even if the EKC held for one pollutant, it may not do so more generally. Most important, these non-reversibility thresholds may in any case be below the identified income threshold per capita GDP levels YT as indicated in Appendix Figure 4.1.    Ecologists therefore urge “the precautionary principle” and suggest preemptive action because the cost of reversing environmental damage is beyond the remit of L/LMICs. As with the income Kuznets curve, there is little faith in automatic improvements in the environment with rising per capita GDPs. They argue that in any case the environment improvements witnessed for HICs are at the expense of L/MICs since they outsource dirty production onto them. Thus, it is important to address the problem globally.    From a development perspective, the poor are disproportionately hurt by environmental degradation as discussed in this chapter and so addressing environment degradation is a key part of poverty alleviation. In a nutshell, instead of moving along the inverted-U, they argued that L/MICs path should be altered and they should be assisted to “burrow through the curve” in a straight line as shown in Appendix Figure 4.2.    The straight line with an arrow indicates a lower level of pollution even at low levels of per capita GDP. This alternative path can ideally come from green industry and agriculture policies as identified in Chapters 13 and 14. Such domestic policies can be facilitated by a grand global bargain in which L/ LMICs focus on clean production that is facilitated by technology transfer from HICs. A modest global bargain was struck in the form of the Paris Agreement which went into effect in November 2016 with 195 countries participating.                                      Poverty, inequality and some proposed solutions 107
314
315         Pollution          indicator(s)
316
317                                                                          PCGDP
318
319         APPENDIX FIGURE 4.2   Burrowing through the environment Kuznets curve
320
321All participating countries pledged to cut emissions to keep global warming to well below 2 degrees centigrade warming compared to pre-industrial levels.33 HICs pledged to provide L/LMIC with $100 billion per year for clean energy technology starting in 2020.
322
323Questions and exercises 1.  Explain how one moves from GNP in current prices to per capita GNP in     constant PPP$. 2. Fully explain in Q1 the adjustments and possible problems at each stage in     an L/LMIC context. 3. Explain the logic of conversion to PPP$ for comparison of per capita GNIs     across countries. 4. Explain why PCGDP in PPP$ may overstate L/LMIC prosperity. 5. Explain how the HDI is computed and adjusted for income, gender or regional     inequalities if data are available? 6. Explain the headcount index as a measure of poverty and explain the     shortcomings of this approach. 7. Explain the distinction between the functional distribution of income and the     size distribution of income. 8. Explain criticisms of the theory underlying the functional distribution of     income. 9. Discuss neoclassical and Marx’s views on exploitation that call the fairness     of the functional distribution of income into question. 10. Explain how the following are conceptually different:      •      Income distribution vs. income redistribution      •      Inequity vs. inequality. 108        Background
324
32511. Explain how the size distribution of income is utilized for computing a     measure of inequality (Gini coefficients) and explain the criticisms of this     approach. 12. Explain the distinction between head-count and basic human needs poverty. 13. Explain the characteristics of poverty (incidence). 14. Explain how different definitions of poverty brought about different policy     initiatives for poverty alleviation. 15. Discuss some of the cross-cutting mechanisms used to make poverty     alleviation policy options more effective. 16. Explain the case for “trickle-down economics” in the development economics     literature. 17. What were the mechanisms proffered for “trickle-down economics”? 18. What is the problem with using cross-country data analysis to estimate the     Kuznets curve and why do country case studies cast doubt on trickle down? 19. Evaluate the application of the Kuznets curve to the environment. 20. Why is investment in the environment viewed as an investment in the poor? 21. Go to the online World Bank, World Development Indicators and pick an income     variable (real PCGDP in constant $). Go to the online UNDP HDI database and     pick the HDI for the same group of countries. Review and then correlate the     data. Comment in the context of a discussion on growth and development. 22. Go to the online World Bank, World Development Indicators and pick an     income (real PCGDP in constant $s) and emission variable for all countries     the data are available for. Review and then chart the data. Do you see a     pattern? Comment in the context of the EKC. 23. Pick any poverty alleviation initiative for a country of your choice (you can     pick one from the list discussed in this chapter).       a.    Identify the conceptual underpinnings of the initiative (e.g. sustainable             livelihoods, basic human needs, etc.).       b.    Evaluate the success of the initiatives.       c.    What are your recommendations for improvements in the implementation             of the initiative?
326
327Notes 1 The relative poverty threshold is defined as the mean or medium income and the   proportion of individuals falling below that are viewed to be in relative poverty (Farina,   2016, p. 126). 2 Refer to the debate between Sanjay Reddy and Thomas Pooge vs. Martin Ravillion on   international poverty lines in (eds.) Anand, Segal and Stiglitz (2010). 3 This term was popularized in the development literature by Hirschman (1970). 4 The covariate risk refers to shocks occurring simultaneously such as disease vectors   associated with droughts and earthquakes. The cholera outbreak that occurred after the   Haiti earthquake in 2016 is an example. 5 The detrimental impact of lead poisoning on children’s learning abilities has been   widely documented (Aizer et al., 2016).                                       Poverty, inequality and some proposed solutions 109
328
3296 As explained in Chapter 15, catch-up growth can be initiated on a foundation of equity. 7 The result carries over to more than two factors using algebra. 8 For those comfortable with calculus, the total product results from integrating marginal    product. 9 Gender discrimination is not being taken into account in this story. 10 For the aggregation problem debate refer to Burmeister (2000) and Cohen and Harcourt    (2003). 11 While capital hiring labor is the most common approach to how production is organized,    there are alternative production models such as producer cooperatives in which labor hires    capital. In this case, labor bears the risk of production and is the claimant of the residual    once all other factors are paid. The Mondrogon producer cooperative in Spain is the most    famous example of such a producer cooperative. Refer to Campbell (2011) for details. 12 Heterodox economists view such labor displacement by automation as an ongoing    issue in capitalist market economies. 13 Marxian theory of exploitation is elaborated on in Chapter 5. 14 For example, the living wage movement gained strength in the United States as    unemployment fell after the 2007–2009 financial and economic crises. While recent    explanations of growing inequality in HICs eschew Marxian analysis, such as that put    forward by Piketty (2014), the crux of the explanation still amounts to capital’s ability    to garner a higher share of total income relative to labor. 15 The Lorenz curve starts at the origin because zero percentage of the population can only    hold 0 percent of total income. Similarly 100 percent of the population holds all of total    income, the point at which the curve ends. These therefore are the two points at which    the Lorenz curve coincides with the 45 degree line, the starting and end point. 16 The “Occupy Movement” in the United States, which protested Wall Street shenanigans    that led to the 2007–2009 financial and economic crisis, used such indices very    effectively. It argued that the top 1 percent of the population drew a large part of total    income and wealth relative to the other 99 percent. For example, according to a 2014    US Congressional Budget Office Report (Distribution of Household Income and    Federal Taxes, table 3), the before tax annual income ratio of the top 1 percent relative    to the bottom 20 percent was 59 in 2011. 17 It is also interesting and relevant since it launched itself into HIC status only in the    1990s after three decades of sustained catch-up growth. 18 In the face of evidence to the contrary, or at least the likelihood of long trickle-down    lags, the World Bank and IMF now push for social safety nets as part of the reform    programs they advocate for L/LMICs that solicit their assistance (Chapter 8). 19 Such exercises were common during the heyday of national centralized planning during the    1960s and 1970s but fell out of favor during the neoliberal era starting in the early 1980s. 20 Following Kuznets’ identification of mechanisms, other variables in the regressions    included education, urbanization, share of agriculture in GDP, rural workers as a    percentage of the total labor force, population growth and various policy interventions. 21 For a brief review of the literature, refer to Kiefer and Khan (2008). 22 Data are drawn from the World Bank, World Development Indicators. 23 The Gini index for 1981 for China was procured from an earlier release of the World    Development Indicators and is not reported in the latest release (issued 1/30/2019) at    the time of writing. 24 The Kuznets curve has been applied to the environment. Given that the environment is    one of the key themes in this textbook, this association is explored in Appendix 4.1. 25 This was shown by Morduch (1999) to be an overstatement based on the accounting    methods adopted (historic rather than current portfolio in the denominator), but much    better than the 50 percent or so repayment rates for public sector rural banking. 26 Refer to Bateman and Chang (2012) for this view. Khan and Ansari (2018) found that in    most competitive informal sector markets in Pakistan, borrowers who had entrepreneurial    drive were finding their niche and making repayments based on sales revenue. 110     Background
330
33127 For a comprehensive review of this social policy refer to (eds.) Adato and Hoddinott    (2010). For a comprehensive meta analysis (study of studies), refer to García and    Saavedra (2017). They review 94 studies from 47 education-related CCTs across L/    LMICs to ascertain how educational outcomes and cost effectiveness are impacted by    design features and compare their results to the earlier meta analyses. 28 BRAC was founded in 1972 as Bangladesh Rural Advancement Committee. It is a large    development NGO that has diversified and now has a presence in 13 countries across    Asia, Africa and Latin America. 29 The analogy often used was that of a gear shift car needing to be pushed until the engine    fired up. 30 Newly introduced development initiatives and mechanisms have a tendency to become    a fad with a bandwagon effect and this may have been the case with participation as a    mechanism. For reservations see da Cunha and Pena (1997). 31 For alternative views, refer to Cole (2003) and Kearsley and Riddel (2010). 32 www.economist.com/china/2017/03/02/chinas-citizens-are-complaining-more-    loudly-about-polluted-air, consulted 1/6/2019 and www.livemint.com/Politics/    cJeH7pAmyAFnHuFa2ilwYP/Delhi-pollution-fuels-anger-in-citizens-as-toxic-air-    burns-e.html, consulted 1/6/2019. 33 Some argue that at this stage even 2 degrees centigrade is too much and would not avert    the climate tipping points that would result in more frequent and more severe disasters    unfolding. Various populations on the globe have already witnessed a foreshadowing of    such events including stronger hurricanes and sea level rises.
332
333References Adato, M. and J. Hoddinott. (eds.). 2010. Conditional Cash Transfers in Latin America    (Baltimore: Johns Hopkins University Press). Aizer, A., J. Currie, P. Simon and P. Vivier. 2016. “Do Low Levels of Blood Lead Reduce    Children’s Future Test Scores?” National Bureau of Economic Research Working    Papers 22558, Boston, MA. Alesina, A. and R. Perotti. 1996. “Income Distribution, Political Instability, and    Investment,” European Economic Review, 40(6), 1203–1228. Alkire, S., J. M. Roche, M. E. Santos and S. Seth. 2011. “Multidimensional Poverty Index    2011: Brief Methodological Note,” OPHI (Oxford Poverty and Human Development    Initiative) Briefing 07, University of Oxford. Anand, S., P. Segal and J. E. Stiglitz. (eds.), 2010. Debates on the Measurement of Global    Poverty (New York: Oxford University Press). Banerjee, A. et al. 2015. “A Multifaceted Program Causes Lasting Progress for the Very Poor:    Evidence from Six Countries,” Science, 348(6236). DOI:10.1126/science.1260799. Bateman, M. and H.-J. Chang. 2012. “Microfinance and the Illusion of Development: From    Hubris to Nemesis in Thirty Years,” World Economic Review, 1(September 6), 13–36. Becker, G. 1981. A Treatise on the Family (Cambridge, MA: Harvard University Press). Breitkreuz, R., C. Stanton, N. Brady, J. Pattison-Williams, E. D. King and C. Mishra.    2017. “The Mahatma Gandhi National Rural Employment Guarantee Scheme: A    Policy Solution to Rural Poverty in India?,” Development Policy Review, 35(3),    397–417. Burmeister, E. 2000. “The Capital Theory Controversy,” in: H. D. Kurz. (ed.), Critical    Essays on Piero Sraffa’s Legacy in Economics (Cambridge: Cambridge University    Press). 305-314 Campbell, A. 2011. “The Role of Workers in Management: The Case of Mondragon,”    Review of Radical Political Economics, 43(3), 328–333.                                    Poverty, inequality and some proposed solutions 111
334
335Chang, H.-J. 2011. “Hamlet without the Prince of Denmark: How Development Has    Disappeared from Today’s ‘DEVELOPMENT’ Discourse,” in: S. R. Khan and J.    Christensen. (eds.), Towards New Developmentalism: Market as Means Rather than    Master (New York: Routledge). 47-58 Cohen, A. J. and G. C. Harcourt. 2003. “Retrospectives: Whatever Happened to the    Cambridge Capital Theory Controversy,” The Journal of Economic Perspectives, 17(1),    199–214. Cole, M. A. 2003. “Development, Trade and the Environment: How Robust Is the    Environment Kuznets Curve?,” Environment and Development Economics, 8(4),    557–580. da Cunha, P. V. and M. V. J. Pena. 1997. “The Limits and Merits of Participation,” The    World Bank, Policy Research Working Paper Series No.1838, Washington, DC. Devine, G. J. and M. J. Furlong. 2007. “Insecticide Use: Contexts and Ecological    Consequences,” Agriculture and Human Values, 24(3), 281–306. Ezzati, M. and D. Kammen. 2005. “The Health Impacts of Exposure to Indoor Air    Pollution from Solid Fuels in Developing Countries: Knowledge, Gaps, and Data    Needs.” Resources For the Future, Discussion Paper 02-24, Washington, DC Farina, F. 2016. “Development Theory and Poverty,” in: C. Sunna and G. Davide. (eds.),    Development Economics in the Twenty-First Century (New York: Routledge). 122-142 Fukuda-Parr, S., A. E. Yamin and J. Greenstein. 2015. “The Power of Numbers: A Critical    Review of Millennium Development Goal Targets for Human Development and Human    Rights,” Journal of Human Development and Capabilities, 15(2–3), 105–117. Goetz, A. and R. S. Gupta. 1996. “Who Takes the Credit?” Gender, Power, and Control over    Loan Use in Rural Credit Programs in Bangladesh,” World Development, 24(1), 45–63. Grossman, G. M. and A. B. Krueger. 1991 “Environmental Impacts of a North American    Free Trade Agreement,” National Bureau of Economic Research Working Paper 3914,    Cambridge, MA. Haushofer, J. and J. Shapiro. 2017. “The Short-term Impact of Unconditional Cash    Transfers to the Poor: Experimental Evidence from Kenya,” The Quarterly Journal of    Economics, 132(4), 2057–2060. Hirschman, A. O. 1970. Exit, Voice, and Loyalty; Responses to Decline in Firms,    Organizations, and States (Cambridge, MA: Harvard University Press). Jha, R., R. Gaiha, K. Manoj, M. K. Pandeya and S. Shankarb. 2015. “Determinants    and Persistence of Benefits from the National Rural Employment Guarantee Scheme –    Panel Data Analysis for Rajasthan, India,” European Journal of Development Research,    27(2), 308–329. Kearsley, A. and M. Riddel. 2010. “A Further Inquiry into the Pollution Heaven Inquiry    and the Environment Kuznets Curve,” Ecological Economics, 69(4), 905–919. Khan, S. R. and N. Ansari. 2018. A Microcredit Alternative in South Asia: Akhuwat’s    Experiment (London: Routeledge). Khan, S. R., S. Kazmi and Z. Rifaqat. 2007. Harnessing and Guiding Social Capital for    Rural Development (New York: Palgrave/Macmillan). Khandker, S. R. and H. A. Samad. 2014. “Dynamic Effects of Microcredit in Bangladesh,”    Policy Research Working Paper No 6821, World Bank Group, Washington, DC. Kiefer, D. and S. R. Khan. 2008. “Revealed Social Preferences for Equality and Growth,”    Journal of Income Distribution, 17(1), 21–33. Kuznets, S. 1955. “Economic Growth and Income Inequality,” American Economic    Review, 45(1), 1–28. Mani, A., S. Mullainathan, E. Shafir and J. Zhao. 2013. “Poverty Impedes Cognitive    Function,” Science, 341(976), 976–980. 112     Background
336
337Morduch, J. 1999. “The Microfinance Promise,” Journal of Economic Literature, 37(4),     1569–1614. Moseley, F. 2012. “A Critique of the Marginal Productivity Theory of the Price of Capital,”     Real-World Economics Review, issue no. 59, www.paecon.net/PAEReview/issue59/     Moseley59.pdf, consulted 1/28/2019. Mullainathan, S. and E. Shafir. 2013. “Scarcity: Why Having Too Little Means so Much,”     Science News, 184(8), 34–35. Narayan, D., Patel, R, Schafft, K. Rademacher, A. and Koch-Schulte, S. (2000)Can Anyone     Hear Us?: Voices of the Poor (World Bank, Washington, D. C.). Park, S. 2009. “Analysis of Saemaul Undong: A Korean Rural Development Programme in     the 1970s,” Asia-Pacific Development Journal, 16(2), 113–140. Parmar, A. 2003. “Micro-Credit, Empowerment, and Agency: Re-evaluating the     Discourse,” Canadian Journal of Development Studies, 24(3), 461–476. Piketty, T. 2014. Capital in the Twenty-First Century (Cambridge, MA: Belknap Press of     Harvard University Press). Pitt, M., S. R. Khandker and J. Cartwright. 2006. “Empowering Women with Micro     Finance: Evidence from Bangladesh,” Economic Development and Cultural Change,     54(4), 791–831. Rankin, K. N. 2001. “Governing Development: Neoliberalism, Microcredit, and Rational     Economic Woman,” Economy and Society, 30(1), 8–37. Rankin, K. N. 2002. “Social Capital, Microfinance, and the Politics of Development,”     Feminist Economics, 8(1), 1–24. Rubalcava, L., G. Teruel and D. Thomas. 2009. “Symposium: Impacts of the Oportunidades     Program: Investments, Time Preferences, and Public Transfers Paid to Women,”     Economic Development and Cultural Change, 57(3), 507–538. Sandra García, Juan Saavedra. Issued in July 2017. Educational Impacts and Cost-     Effectiveness of Conditional Cash Transfer Programs in Developing Countries: A     Meta-Analysis, NBER Working Paper No. 23594, NBER Program(s):Development     Economics, Economics of Education. Seguin, M. and M. Nino- Zarazua 2011. “What Do We Know about Non-clinical     Interventions for Preventable and Treatable Childhood Diseases in Developing     Countries?” World Institute for Development Economic Research (UNU-WIDER)     Working Paper WP2013/087, Helsinki. Sen, A. K. 1990. “Gender and Co-Operative Conflicts,” in: I. Tinker. (ed.), Persistent     Inequalities: Women and World Development (New York: Oxford University Press).     123-149 PART II
338
339Key approaches to economic development and the middle income trap 5 CLASSICAL AND RADICAL ANTECEDENTS OF DEVELOPMENT ECONOMICS
340
341Introduction Classical political economists focus on the causes of the wealth (income) of nations and the distribution of that wealth across classes. Thus the units of analysis are the collective (economy) and groups or classes. Class is generally not mentioned in mainstream economics textbooks but was introduced as a central unit of analysis by Marx. While Marx is referred to as a classical political economist, he introduced his own conceptual framework and terminology and his writings inspired radical thought in many social science disciplines and fields including development economics.     The transition from classical political economy to economics in the late 19th century shifted the unit of analysis to the individual or the firm and the focus to the efficient allocation of resources.1 This transition also marked the dropping of the word political and so “principles” (key works) of dominant scholars like Alfred Marshall during the onset of the neoclassical era became principles of economics rather than principles of political economy.     The pioneers of development economics (Chapter 6), trained in both classical and neoclassical economics, recognized that it was the classical political economists who grappled with issues, such as the progress of nations (economic growth), that were more pertinent to newly independent colonies in the 1940s onwards. They also recognized that economic conditions in these newly independent countries were structurally and institutionally different from those in the dominant nations, and their implicit or explicit economic models portrayed this difference and drew on classical assumptions and economics as they saw fit.     In addition to theories of progress, a second theme in development economics pertains to theories of underdevelopment i.e. why do nations not progress? Pertinent to this question are the theories of colonialism and imperialism from which the heterodox strain of development economics emerged. While Marx mostly focused on the critique of capitalism, his writings inspired the radical strain in development economics thought.     The review of Marx in this chapter, like the review of classical political economists in general, is selective. Just as the review of classical political economists focuses on ideas and concepts that inspired the pioneers of development economics, the 116    Economic development, middle income trap
342
343review of Marx focuses on presenting ideas and views that inspired later thinkers on colonialism and imperialism.    Also, Marx’s critique of classical political economy and the capitalist system made class and distribution its centerpiece, and the neo-Marxist and political economy of development approaches in development economics drew on these constructs. This chapter turns to antecedents of the radical strain in development economics after discussing classical political economic thinkers.
344
345Classical political economists Smith (1908) drew on prior thought but is widely agreed to have been the first to synthesize earlier ideas and present a systematic framework of an economy. The power of a country for Smith was in proportion to the value of its annual produce (GNP) and hence his central focus was on production. He recognized that the wealth of nations was production and welfare generated from the consumption of these products. This may seem obvious now, but it was not so in his time. Mercantilist thinking of his time associated maximizing ownership of specie – gold and silver – with the wealth of nations.2    In a nutshell, for Smith the wealth of nations or size of the “fund” (GNP) was principally determined by the skill and dexterity of the workforce and the judgment with which it was applied to production and by the ratio of active and non-active productive labor (participation rate). The prosperity of nations was greater the higher the ratio of productive to unproductive labor (see below). Progress or economic growth was determined principally by capital accumulation and innovations that raised the skill and dexterity or productivity of labor.    His famous observation of a pin factory, with which he started Book I of The Wealth of Nations, showed how the division of labor and subsequent specialization unleashed the productive potential of society. Thus, he observed and noted that output increased from a scarce 20 per person per day to 4,800 per person per day with specialization when the job was broken down into 18 tasks. However, machinery (capital accumulation) was necessary for such specialization and division of labor to increase labor productivity, and hence capital accumulation played a primary role. The extent of specialization depended both on the size of the market and the availability of capital. Having an unrestricted exchange economy was thus critical for progress. This was another of his tremendous insights.    Manufacturing lent itself to such division of labor and specialization much more than agriculture and hence its vital role in enhancing the wealth of nations. Because agriculture was by nature limited in the extent to which tasks could be sub-divided, he argued that labor productivity in manufacturing necessarily exceeds that in agriculture. Due to the greater potential for division of labor, and the ingenuity of machine makers whose inventions make tasks “easier and readier” (p. 8), time saving is also more likely in manufacturing.    The savings of the entrepreneur were special because they went into investment and boosting labor productivity whereas landlords frittered away their income.                                                   Classical and radical antecedents 117
346
347The hallmark of a more advanced society was one in which the stock of capital per capita was higher and this still has resonance. Given that he assumed only capitalists saved, he regarded a distribution of income in their favor as positive for society.3 Smith argued that public prodigality can impoverish great nations by eating into the stock that can be used to hire productive labor. By the same token, nations could flourish as a result of the parsimony or frugality of individuals. Frugality is critical for increasing the fund that hires productive labor and it, rather than industry (hard work), is the immediate cause of the increase in capital.    Smith distinguished between two types of labor as mentioned earlier. The [relative] progress of nations is determined by the part of labor that “adds value to the subject it is bestowed on” such as machinery and materials in manufacturing and he considered this component of labor as productive. The other part of labor is not productive and perishes in the moment of performing a service. In this category Smith included the sovereign and associated splendid court services, the ecclesiastic establishments and the military. He criticized “great fleets and armies, who in time of peace produce nothing, and in time of war acquire nothing which can compensate for the expense of maintaining them, even while the war lasts” (p. 254).4    Unlike the political economists who succeeded him, Smith was quite optimistic. He envisaged opulence that would raise the general standard of living because of capital accumulation and the increasing division of labor and specialization and skill development that this facilitated. Smith’s optimistic evaluation was that in great nations, natural progress is maintained by frugality which offsets public extravagance. He asserted that the driving spirit behind this noble conduct was the “uniform, constant and uninterrupted effort of every man to better his condition” (p. 264). Since these motivations were universal, including in primitive societies, the other supportive conditions such as the collective institutional framework needed to come into play.    These other supporting determinants of prosperity and progress included natural resources such as soil, climate, laws, able administration,5 stable money (because this facilitated transactions) and navigable rivers (that facilitated transportation and enhanced internal trade and growth).6 Nations needed tolerable security, otherwise there is an incentive to bury capital for contingencies rather than consume it or employ it for profit. Most current mainstream development economists would probably agree that this list of determinants of economic growth is still very relevant as is the centrality of capital accumulation.7    The determinants of prosperity and progress identified above resulted for Smith in “the full complement of riches” which enabled society to be in a progressive state. Eventually, he saw society as moving to a stationary state (growth reproducing society at the level it has attained) but anticipated that this state would not come about for a couple of hundred years for Great Britain.    Smith argued that the great commerce of any civilized nation is carried on between town and country to the mutual benefit of both. Trade within a nation had the advantage of greater control over and security of capital, but the principle of mutual benefit from trade extended across borders. Smith originated modern trade theory, which represents a key complement to development economics. 118    Economic development, middle income trap