英文版发展经济学Development Economics By Debraj Ray

Development Economics By Debraj Ray, New York University March 2007. Prepared for theNew Palgrave Dictionary of Economics, edited by Lawrence Blume and Steven Durlauf. 1 Introduction What we know as the developing world is approximately the group of countries classified by the World Bank as having “low” and “middle” income. An exact description is unnecessary and not too revealing; suffice it to observe that these countries make up over 5 billion of world population, leaving out the approximately one billion who are part of the “high” income developed world. Together, the low and middle income countries generate approximately 6 trillion (2001) dollars of national income, to be contrasted with the 25 trillion generated by high income countries. An index of income that controls for purchasing power would place these latter numbers far closer together (approximately 20 trillion and 26 trillion, according to the World Development Report (2003)) but the per-capita disparities are large and obvious, and to those encoutering them for the first time, still extraordinary. Development Economics, a subject that studies the economics of the developing world, has made excellent use of economic theory, econometric methods, sociology, anthropology, political science, biology and demography and has burgeoned into one of the liveliest areas of research in all the social sciences. My limited approach in this brief article is one of deliberate selection of a few conceptual points that I consider to be central to our thinking about the subject. The reader interested in a more comprehensive overview is advised to look elsewhere (for example, at Dasgupta (1993), Hoff, Braverman and Stiglitz (1993), Ray (1998), Bardhan and Udry (1999), Mookherjee and Ray (2001), and Sen (1999)). I begin with a traditional framework of development, one defined by conventional growth theory. This approach develops the hypothesis that given certain parameters, say sav- ings or fertility rates, economies inevitably move towards some steady state. If these parameters are the same across economies, then in the long run all economies converge to one another. If in reality we see utter lack of such convergence — which we do (see, e.g., Quah (1996) and Pritchett (1997)) — then such an absence must be traced to a presumption that the parameters in question are not the same. To the extent that his- tory plays any role at all in this view, it does so by affecting these parameters — savings, demographics, government interventionism, “corruption” or “culture”. This view is problematic for reasons that I attempt to clarify below. Indeed, the bulk of my essay is organized around the opposite presumption: that two societies with the same fundamentals can evolve along very different lines — going forward — depending on past expectations, aspirations or actual history. Now, after a point, the distinction between evolution and parameter is a semantic one. By throwing enough state variables (“parameters”) into the mix, one might argue that there is no difference at all between the two approaches. Formally, that would be correct, but then “parameters” would have to be interpreted broadly enough so as to be of little explanatory value. Ahistorical convergence and historically conditioned divergence express two fundamentally different world views, and there is little that semantic jugglery can do to bring them together. 2 Development From The Viewpoint of Convergence Why are some countries poor while others are rich? What explains the success stories of economic development, and how can we learn from the failures? How do we make sense of the enormous inequalities that we see, both within and across questions? These, among others, are the “big questions” of economic development. It is fair to say that the model of econonomic growth pioneered by Robert Solow (1956) has had a fundamental impact on “big-question” development economics. For theory, calibration and empirical exercises that begin from this starting point, see, e.g., Lucas (1990), Mankiw, Romer and Weil (1992), Barro (1991), Parente and Prescott (2000) and Banerjee and Duflo (2005). Solow’s pathbreaking work introduced the notion of convergence: countries with a low endowment of capital relative to labor will have a high rate of return to capital (by the “law” of diminishing returns). Consequently, a given addition to the capital stock will have a larger impact on per-capita income. It follows that, controlling for parameters such as savings rates and population growth rates, poorer countries will tend to grow faster and hence will catch up, converge to the levels of well-being enjoyed by their richer counterparts. Under this view, development is largely a matter of getting some economic and demographic parameters right and then settling down to wait. To be sure, savings and demography are not the only factors that qualify the argument. Anything that systematically affects the marginal addition to per-capita income must be controlled for, including variables such as investment in “human capital” or harder- to-quantify factors such as “political climate” or “corruption”. A failure to observe convergence must be traced to one or another of these “parameters”. Convergence relies on diminishing returns to “capital”. If this is our assumed starting 2 point, the share of capital in national income does give us rough estimates of the concavity of production in capital. The problem is that the resulting concavity understates observed variation in cross-country income by orders of magnitude. For instance, Parente and Prescott (2000) calibrate a basic Cobb-Douglas production function by using reasonable estimates of the share of capital income (0.25), but then huge variations in the savings rate do not change world income by much. For instance, doubling the savings rate leads to a change in steady state income by a factor of 1.25, which is inadequate to explain an observed range of around 20:1 (PPP). Indeed, as Lucas (1990) observes, the discrepancy actually appears in a more primitive way, at the level of the production function. For the same simple production function to fit the data on per-capita income differences, a poor country would have to have enormously higher rates of return to capital; say, 60 times higher if it is one-fifteenth as rich. This is implausible. And so begins the hunt for other factors that might explain the difference. What did we not control for, but should have? This describes the methodological approach. The convergence benchmark must be pitted against the empirical evidence on world income distributions, savings rates, or rates of return to capital. The two will usually fail to agree. Then we look for the parametric differences that will bridge the model to the data. “Human capital” is often used as a first port of call: might differences here account for observed cross-country variation? The easiest way to slip differences in human capital into the Solow equations is to renormalize labor. Usually, this exercise does not take us very far. Depending on whether we conduct the Lucas exercise or the Prescott- Parente variant, we would still be predicting that the rate of return to capital is far higher in India than in the U.S., or that per-capita income differences are only around half as much (or less) as they truly are. The rest must be attributed to that familiar black box: “technological differences”. That slot can be filled in a variety of ways: externalities arising from human capital, incomplete diffusion of technology, excessive government intervention, within-country misallocation of resources, . . . . All of these — and more — are interesting candidates, but by now we have wandered far from the original convergence model, and if at all that model still continues to illuminate, it is by way of occasional return to the recalibration exercise, after choosing plausible specifications for each of these potential explanations. This model serves as a quick and ready fix on the world, and it organizes a search for possible explanations. Taken with the appropriate quantity of salt, and viewed as a first pass, such an exercise can be immensely useful. Yet playing this game too seriously reveals a particular world-view. It suggests a fundamental belief that the world economy is ultimately a great leveller, and that if the levelling is not taking place we must search for that explanation in parameters that are somehow structurally rooted in a society. 3 To be sure, the parameters identified in these calibration exercises do go hand in hand with underdevelopment. So do bad nutrition, high mortality rates, or lack of access to sanitation, safe water and housing. Yet there is no ultimate causal chain: many of these features go hand in hand with low income in self-reinforcing interplay. By the same token, corruption, culture, procreation and politics are all up for serious cross-examination: just because “cultural factors” (for instance) seems more weighty an “explanation” does not permit us to assign it the status of a truly exogenous variable. In other words, the convergence predicted by technologically diminishing returns to in- puts should not blind us to the possibility of nonconvergent behavior when all variables are treated as they should be — as variables that potentially make for underdevelopment, but also as variables that are profoundly affected by the development process. 3 Development from The Viewpoint of Nonconvergence This leads to a different way of asking the big questions, one that is not grounded in any presumption of convergence. The starting point is that two economies with the same fundamentals can move apart along very different paths. Some of the best- known economists writing on development in the first half of the twentieth century were instinctively drawn to this view: Young (1928), Nurkse (1953), Leibenstein (1957) and Myrdal (1957) among them. Historical legacies need not be limited to a nation’s inheritance of capital stock or GDP from its ancestors. Factors as diverse as the distribution of economic or political power, legal structure, traditions, group reputations, colonial heritage and specific institutional settings may serve as initial conditions — with a long reach. Even the accumulated baggage of unfulfilled aspirations or depressed expectations may echo into the future. Factors that have received special attention in the literature include historical inequal- ities, the nature of colonial settlement, the character of early industry and agriculture, and early political institutions. 3.1 Expectations and Development Consider the role of expectations. Rosenstein-Rodan (1943) and Hirschman (1958) (and several others following them) argued that economic development could be thought of as a massive coordination failure, in which several investments do not occur simply because other complementary investments are similarly depressed in the same bootstrapped way. Thus one might conceive of two (or more) equilibria under the very same fundamental conditions, “ranked” by different levels of investment. 4 Such “ranked equilibria” reply on the presence of a complementarity: a particular form of externality in which the taking of an action by an agent increases the marginal benefit to other agents from taking the a similar action. In the argument above, sector-specific investments lie at the heart of the complementarity: more investment in one sector raises the return to investment in some related sector. Once complementarities — and their implications for equilibrium multiplicity — enter our way of thinking, they seem to pop up everywhere. Complementarities play a role in explaining how technological inefficiencies persist (David (1985), Arthur (1994)), why financial depth is low (and growth volatile) in developing countries (Acemoglu and Zili- botti (1997)), how investments in physical and human capital may be depressed (Romer (1986), Lucas (1988)), why corruption may be self-sustaining (Kingston (2005), Emer- son (2006)), the growth of cities (Henderson (1988), Krugman (1991)), the suddenness of currency crises (Obstfeld (1994)), or the fertility transition (Munshi and Myaux (2006)); I could easily go on. Even the traditional Rosenstein-Rodan view of demand comple- mentarities has been formally resurrected (Murphy, Shleifer and Vishny (1989)). An important problem with theories of multiple equilibrium is that they carry an un- clear burden of history. Suppose, for instance, that an economy has been in a low-level investment trap for decades. Nothing in the theory prevents that very same economy from abruptly shooting into the high-level equilibrium today. There is a literature that studies how the past might weigh on the present when a multiple equilibrium model is embedded in real time (see, e.g., Adsera` and Ray (1998) and Frankel and Pauzner (2000)). When we have a better knowledge of such models we will be able to make more sense of some classical issues, such as the debate on balanced versus unbalanced growth. Rosenstein-Rodan argued that a “big push” — a large, balanced infusion of funds — is ideal for catapulting an economy away from a low-level equilibrium trap. Hirschman argued, in contrast, that certain “leading sectors” should be given all the attention, the resulting imbalance in the economy provoking salubrious cycles of private investment in the complementary sectors. To my knowledge, we still lack good theories to examine such debates in a satisfactory way. 3.2 Aspirations, Mindsets and Development The aspirations of a society are conditioned by its circumstances and history, but they also determine its future. There is scope, then, for a self-sustaining failure of aspirations and economic outcomes, just as there is for ever-progressive growth in them (Appadurai (2004), Ray (2006)). Typically, the aspirations of an individual are generated and conditioned by the experi- ences of others in her “cognitive neighborhood”. There may be several reasons for this: 5 the use of role models, the importance of relative income, the transmission of informa- tion, or peer-determined setting of internal standards and goals. Such conditioning will affect numerous important socio-economic outcomes: the rate of savings, the decision to migrate, fertility choices, technology adoption, the adherence to norms, the choice of ethnic or religious identity, the work ethic, or the strength of mutual insurance motives. As an illustration, consider the notion of an aspirations gap. In a relatively narrow economic context (though there is no need to restrict oneself to this) such a gap is simply the difference between the standard of living that’s aspired to and the standard of living that one already has. The former isn’t exogenous; it will depend on the ambient standards of living among peers or near-peers, or perhaps other communities. The aspirations gap may be filled — or neglected — by deliberate action. Investments in education, health, or income-generating activities are obvious examples. Does history, via the creation of aspirations gaps, harden existing inequalities and generate poverty traps? Or does the existence of a gap spur individuals on ever harder to narrow the distance? As I have argued in Ray (1998, Sections 3.3.2 and 7.2.4) and Ray (2006), the effect could go either way. A small gap may encourage investments, a large gap stifle it. This leads not only to history-dependence, but also a potential theory of the connections between income inequality and the rate of growth. These remarks are related to Duflo’s (2006) more general (but less structured) hypoth- esis that “being poor almost certainly affects the way people think and decide”. This “mindset effect” can manifest itself in many ways (an aspirations gap being just one of them), and can lead to poverty traps. For instance, Duflo and Udry (2004) find that cer- tain within-family insurance opportunities seem to be inexplicably foregone. In broadly similar vein, Udry (1996) finds that men and women in the same household farm land in a way that is not Pareto-efficient (gains in efficiency are to be had by simply realocating inputs to the women’s plots). These observations suggest a theory of the poor household in which different sources of income are treated differently by members of the household, perhaps in the fear that this will affect threat points in some intrahousehold bargaining game. This in itself is perhaps not unusual, but the evidence suggests that poverty itself heightens the salience of such a framework. 3.3 Markets and History-Dependence I now move on to other pathways for history-dependence, beginning with the central role of inequality. According to this view, historic inequalities persist (or widen) because each individual entity — dynasty, region, country — is swept along in a self-perpetuating path of occupational choice, income, consumption, and accumulation. The relatively poor may be limited in their ability to invest productively, both in themselves and in their children. 6 Such investments might include both physical projects such as starting a business, or “human projects” such as nutrition, health and education. Or the poor may have ideas that they cannot profitably implement, because implementation requires startup funds that they do not have. Yet, faced with a different level of initial inequality, or jolted by a one-time redistribution, the very same economy may perform very differently. Now investment opportunities are available widely through the population, and a new outcome emerges with not just lower inequality, but higher aggregate income. These are different steady states, and they could well be driven by distant histories (see, e.g., Dasgupta and Ray (1986), Banerjee and Newman (1993), Galor and Zeira (1993), Ljungqvist (1993), Ray and Streufert (1993), Piketty (1997) or Matsuyama (2000)). The intelligent layperson would be unimpressed by the originality of this argument. That the past systematically preys on the present is hardly rocket science. Yet theories based on convergence would rule such obvious arguments out. Under convergence, the very fact that the poor have limited capital relative to labor allows them to grow faster and (ultimately) to catch up. Economists are so used to the convergence mechanism that they sometimes do not appreciate just how unintuitive it is. That said, it is time now to cross-examine our intelligent layperson. For instance, if all individuals have access to a well-functioning capital market, they should be able to make an efficient economic choice with no heed to their starting position, and the shadows cast by past inequalities must disappear (or at least dramatically shrink). For past wealth to alter current investments, imperfections in capital or insurance markets must play a central role. At the same time, such imperfections aren’t sufficient: the concavity of investment re- turns would still guarantee convergence. A first response is that such “production func- tions” are simply not concave. A variety of investment activities have substantial fixed costs: business startups, nutritional or health investments, educational choices, migration decisions, crop adoptions. Indeed, it is hard to see how the presence of such noncon- vexities could not be salient for the ultrapoor. Coupled with missing capital markets, it is easy to see that steady state traps, in which poverty breeds poverty, are a natural outcome (se, e.g., Majumdar and Mitra (1982), Galor and Zeira (1993)). Surveys of the econonomic conditions of the poor (Banerjee and Duflo (2007), Fields (1980)) are eminently consistent with this point of view. A related source of nonconvexity arises from limited liability. A highly indebted economic agent may have little incentive to invest. Similarly, poor agents may enter into contracts with