THE IDEA OF A PROGRESSIVE MARKET ECONOMY

Until the late twentieth century, the prevailing view was one of skepticism that poor people would benefit much from economic growth in a capitalist economy. Well into the 1980s, it was common to hear in both popular and scholarly writings that economic growth was expected to largely bypass poor people in both rich and poor countries.

By one view, poverty is likely to persist in a growing economy because poverty is relative (Section 22.6). Strictly, poverty could still be eliminated when using a strongly relative poverty line set at a constant proportion of the mean with sufficient redistribution in favor of poor people. Growth in the mean will not eliminate poverty without a change in relative distribution. However, the past explicit acceptance of poverty among econ­omists and noneconomists alike does not appear to be the product of such a relativist view. In fact the latter is a modern idea, which appears to have emerged much later, in the 1970s (Section 22.6). Using absolute or weakly relative poverty measures, suffi­cient inequality-neutral growth will eliminate poverty.^[474]

But growth was not expected to be inequality-neutral. Most classical and Marxist eco­nomic thinkers saw little hope that even a growing capitalist economy would deliver rapid poverty reduction or even any poverty reduction. Although Smith was optimistic about the potential for a progressive, poverty-reducing, market economy, the prominent classical economists who followed, including Malthus and Ricardo, were more pessimis­tic about the prospects for higher real wages and (hence) less poverty, suggesting that they anticipated rising inequality from a growing capitalist economy. As discussed in Section 22.5, demographic responses to rising wages were expected to play a key role in attenuating the poverty impact of growth.

Distributional dynamics has long been a central theme of development economics. Poverty was a concern for the postcolonial governments of the newly independent coun­tries, but the earliest policy-oriented discussions were pessimistic about the prospects of economic growth bringing much benefit to poor people. It was widely believed that growth in low-income countries was bound to be inequitable, and that view is still heard today.

A foundation for this view was provided by Kuznets (1955), and came to be known as the “Inverted U Hypothesis,” whereby inequality first increases with economic growth in a poor country but falls after some critical income level is reached.^[475] Although there are other theoretical models in the literature that can generate such a relationship, in the Kuznets formulation, the economy is assumed to comprise a low-mean, low-inequality rural sector and a high-mean, high-inequality urban sector, and growth is assumed to occur through the migration of workers from the former environment to the latter. This growth is assumed to entail that a representative “slice” of the rural distribution is trans­formed into a representative slice of the urban distribution, preserving distributions within each sector. An inverted U can then be derived linking certain indices of inequal­ity and the population share ofthe urban sector (Anand and Kanbur, 1993; Robinson, 1976).

Some policy makers appear to have incorrectly inferred that this model also implied that economic growth in poor countries would bring little benefit to poor people. (This sometimes reflected a longstanding confusion between the ideas of “poverty” and “inequality” in development policy discussion.) It is easy to show that for all additive poverty measures, if poverty is initially higher in the rural sector, then aggregate poverty must fall under the Kuznets process of migration described above.

The economic history of today’s rich countries has often been seen as a source ofles- sons for the developing world. Contrary to the expectations of both the nineteenth­century supporters and critics of capitalism, Britain’s industrial revolution, which had started around 1760, almost certainly reduced poverty through rising real wage rates. But there was a long lag. Just how long depends on the position one takes in the debate about price indices. Clark’s (2005) discussion of builders’ real wage rates in England sug­gests that workers earned higher wages from about 1800, while Allen (2007, 2009) argues that the increase started closer to 1830. Either way the pessimists appear to have been right that for at least a few decades after the technical innovations, real wages did not increase.^[476] Real wages in Britain did start to rise in the nineteenth century despite con­tinuing population growth. Falling food prices in Europe due to refrigeration and lower freight transport costs also helped increase real wages later in that century (Williamson, 1998). And there is evidence that the gains in real wages for the working class from the mid-nineteenth century came hand-in-hand with improved nutrition.^[477]

The lag in the real wage rate response to the industrial revolution is suggestive of the model by Lewis (1954) in which a surplus of labor in the rural economy keeps wages at a low level until that surplus is absorbed by the economy’s modern (urban) sector, as this expands due to technical progress. Allen (2009) offers an alternative explanation whereby the extra demand for capital due to technical progress could only be met by savings from nonlabor income, under the assumption that workers were too poor to save. Then profits had to rise to finance the investments needed, and only when sufficient capital had accu­mulated did real wages rise.

The empirical foundations for the expectation that inequality would inevitably rise in growing developing countries were not particularly secure at the time the Kuznets hypothesis was influential. There was not much data to draw on. A debate in the early 1970s on the distribution of the gains from economic growth in Brazil left an appetite for better survey data for measuring poverty and inequality.^[478] As better evidence from household surveys accumulated, it was revealed that very few low-income countries have

developed over time in a manner consistent with the Kuznets hypothesis, as is shown by Bruno et al. (1998) and Fields (2001). We have learned that growth in developing coun­tries tends to be distribution-neutral on average, meaning that changes in inequality are roughly orthogonal to growth rates in the mean (Dollar and Kraay, 2002; Ferreira and Ravallion, 2009; Ravallion, 1995, 2001). Distribution-neutral growth implies that the changes in any standard measure of either absolute or weakly relative poverty will be neg­atively correlated with growth rates in the mean.

There is also evidence ofinequality convergence, whereby inequality tends to increase in low inequality countries and decrease in high inequality countries (Benabou, 1996; Ravallion, 2003). This is consistent with neoclassical growth theory, which shows that a fully competitive market economy contains forces for reducing inequality, as demon­strated by Stiglitz (1969) and Benabou (1996). As Ravallion (2003) argues, the evidence we see ofinequality convergence can also be explained by how economic policy conver­gence in the world during the 1990s interacted with prereform differences in the extent of inequality.

The periods of global trade openness fostered some progress toward convergence of living standards across countries. Although much attention has been given to the current globalization period, Williamson (1998) argues that the prior period of globalization, 1870—1914, fostered economic expansion and convergence within the “Atlantic economy.” This globalization almost certainly reduced poverty globally.

Post-Independence policies in most developing countries strived for economic growth, facilitated by government planning in relatively closed economies, although capabilities for effective implementation were often weak. India’s Second and Third Plans, as well as many other planning documents, aimed for growth through accelerated capital accumulation and industrialization. These plans were influenced by classical eco­nomics and the Harrod-Domar equation, although here, too, policy makers misinter­preted the implications of the model.^[479] The prioritization given to the capital-goods sector in India’s Second Plan was directly influenced by a two-sector growth model in Mahalanobis (1953), although there were dissenters at the time (including Vakil and Brahmanand, 1956), and subsequent research in growth economics did not find any robust implication to justify this prioritization. As Lipton (1977) points out, the plan­ners also ignored Adam Smith’s warning that the food supply would constrain urban growth in a closed economy.

China’s enormous progress against absolute poverty since around 1980, alongside ris­ing inequality, might superficially be seen as testimony to the idea that the country has been in the rising segment ofthe Kuznets inverted U. However, here, too, the model just does not fit the facts. For one thing, inequality is lower in urban China than rural China, unlike the case assumed by Kuznets (1955), although this is not necessary for an inverted U; see Robinson (1976). More importantly, neither analytic decompositions of the changes in poverty nor regression-based decompositions suggest that the Kuznets process of growth through modern sector enlargement was the main driver of growth and poverty reduction in China (Ravallion and Chen, 2007). One must look elsewhere, notably to the initial agrarian reforms—including the massive land reform when the land of the collec­tives was assigned to individual farmers—and market liberalization more broadly, for an explanation of China’s rapid poverty reduction in the 1980s.^[481] Manufacturing growth came to play an important role later though that success was based in part on favorable initial conditions, notably the legacy of investments in human development, including in rural areas. Unlike many developing countries, there was a large literate rural population to draw on as the workforce for China’s labor-intensive modern sector enlargement.

In thinking about policies for fighting poverty, the role played by the rural sector has been much debated. The sequence in China was roughly right: In the reform period from 1978, initial attention was given to the rural sector, and agrarian reforms to restore farmer incentives (in land allocation and prices) were crucial to ensuring a sustainably propoor development path, as had been the case elsewhere in East Asia.^[482] Few other countries got the sequence right, and China’s experience contains an important lesson for Africa today (Ravallion, 2009).

There were efforts to reprioritize development policy in the 1970s and 1980s. World Bank President Robert McNamara’s 1973 “Nairobi speech” signaled such an effort from the international development institutions. In development thinking, “urban bias” was increasingly recognized as bad for growth as well as for poverty reduction, though it reflected political structures in much ofthe developing world (Lipton, 1968, 1977). However, the temptation to industrialize rapidly—“run before you have walked”—was strong. Com­bined with huge inequities in access to finance and human development, the subsequent growth paths were disappointing, both in growth and (especially) poverty reduction.

The debt crises of the 1980s brought a wave of structural adjustment programs sup­ported by the international financial institutions (IFIs) that attempted to restore macroeco­nomic balances and promote economic growth. Given that the World Bank had produced Redistribution with Growth 10 years earlier (Chenery et al., 1974), it is surprising that its own adjustment programs in the early and mid-1980s gave little serious attention to the impacts on poor people though this neglect was consistent with the broader 1980s backlash in the Anglo-Saxon world against the distributional focus ofthe 1960s and 1970s. TheBankand Fund programs were much criticized for their neglect of distributional impacts, and the criticisms stuck. A progressive recovery in thinking within the IFIs was underway by the late 1980s, and add-on programs to “compensate the losers from adjustment” were soon common. Today, it is widely recognized that poverty and inequality mitigation has to be designed into economy-wide reform programs from the outset.

By the turn of the twenty-first century, enough evidence had accumulated for econ­omists to be confident that higher growth rates tended to yield more rapid rates of abso­lute poverty reduction.^[483] A more poverty-reducing process of global economic growth emerged after 2000, and not just because of China’s growth. The trend rate of decline in the “$1.25-per-day” poverty rate for the developing world outside China rose from 0.4% points per year from 1980 to 2000 to 1.0% points per year after that period (Ravallion, 2013).

The poverty impact of a given rate of growth depends in part on the initial distribu- tion.^[484] Intuitively, when inequality is high, poor people will tend to have a lower share of the gains from growth. Ravallion (1997a, 2007) confirmed this using household survey data over time.^[485] Easterly (2009) conjectured that the initial poverty rate is likely to be the better predictor of the elasticity than initial inequality though no evidence was provided. Ravallion (2012b) provided that evidence, and it compellingly shows that it is not high initial inequality that impedes the pace of poverty reduction at a given rate of growth, but high poverty.

Saying that growth typically reduces poverty does not, of course, mean that any growth-promoting policy will do so or that everyone will benefit. That depends on the distribution—horizontally as well as vertically—of the gains and losses from that pol­icy. There may be vertical inequalities—between people at different levels of mean income—generated in the process that mitigate the gains to poor people from growth. And there can be horizontal inequities, whereby people at the same initial levels of income fare very differently, and some poor people may well lose from a policy that reduces poverty in the aggregate. (Recall that Harrington (1962) emphasized this point in describing the new “minority poverty” in the “other America.”)

This point has been clearest in the literature on external trade and poverty. A number of studies have found support for the view that trade openness—typically measured by trade volume as a share of GDP—promotes economic growth.^[486] It is unclear that trade volume can be treated as exogenous in these cross-country regressions; higher trade vol­ume may be a response to growth rather than a cause. The policy implications are also unclear since trade volume is not a policy variable; see the discussion in Rodrik (1994) and Rodriguez and Rodrik (2001). But, putting this issue to one side, what about the distributional effects? A number of studies have combined survey-based measures of income inequality at country-level with data on trade and other control variables to assess the distributional impacts of trade openness, as reviewed in Winters et al. (2004). The evidence is mixed. Dollar and Kraay (2004) find little or no effect of trade volume on inequality. Other studies have reported adverse effects. Lundberg and Squire (2003) find evidence that higher trade volume tends to increase inequality. On balance, Ravallion (2006) reports little or no correlation between greater trade openness and the pace of pov­erty reduction in developing countries.

However, there can be winners and losers at all levels of living, even when a standard measure of inequality or poverty is unchanged. There are many sources of heterogeneity, yielding horizontal impacts of reform. Geographic disparities in access to human and physical infrastructure affect prospects for participating in the opportunities created by greater openness to external trade. Differences in household demographic composition influence consumption behavior and hence the welfare impact of the shifts in relative prices associated with trade openness. Ravallion (2006) reports on two case studies of this heterogeneity in the welfare impacts of liberalizing trade reform, for China and for Morocco. The results indicate a sizable, and at least partly explicable, variance in impacts across households with different characteristics—differences that influenced their net trading positions in the relevant markets.

Where does all this leave us? The antitrade policies (on quotas, tariffs, and exchange rates) of the post-Independence development policy regimes were unlikely to bring much benefit to poor people, the bulk of whom produced tradable goods from primarily nontradable inputs. Although this remains a plausible generalization, there is likely to be considerable heterogeneity across countries in such effects, and one might be skeptical of basing policy advice for any specific country on generalizations from either standard Stolper-Samuelson arguments or cross-country regressions (Ravallion, 2006). For exam­ple, some studies have found evidence that higher trade volume increases inequality in poor countries but that the reverse holds true at a higher mean income (Milanovic, 2005; Ravallion, 2001). The macro perspective, focusing on impacts on an aggregate measure of poverty or inequality, hides potentially important horizontal impacts with implications for other areas of policy, notably social protection efforts that may well be needed to complement the growth-promoting reforms. (Section 22.9 discusses these policies further.)

Trade policies have also played a role in social protection, though this, too, has been much debated. Governments of food-exporting but famine-affected areas have often implemented food export bans in the hope of protecting vulnerable citizens. Classical economists were influential in arguing against such policies in favor of free trade. For example, Aykroyd (1974) describes how the Governor of Bombay in the early nine­teenth century quoted Smith’s The Wealth of Nations when defending his policy stance against any form of trade intervention during the famines that afflicted the region. Various “Famine Commissions” set up by the British Raj argued against the trade interventions that were being called for to help protect vulnerable populations. Similarly, Woodham­Smith (1962) describes the influence that Smith and other classical economists had on British policy responses to the severe famines in Ireland in the mid-nineteenth century. In modern times, free trade has been advocated as a means of stabilizing domestic food consumption in the presence of output shocks (World Bank, 1986). Other economists have been less supportive. Sen (1981a) and Ravallion (1987) pointed to the possibility that real income declines in the famine-affected areas can generate food exports while people starve.^[487] Regulated trade through taxes or even export bans may then be a defen­sible policy response to help vulnerable groups relative to feasible alternatives (Ravallion, 1997b).

Critics of trade intervention for the purpose of protection from external price shocks (such as in the period from 2007 to 2011) have pointed out that such a policy can exac­erbate the problem of price volatility (Martin and Anderson, 2012). However, in the absence of better options for aggregate intertemporal smoothing, the optimal nontrade protection policy would entail transfers between net food producers and net consumers, to coinsure. And this, too, would exacerbate the price volatility, as shown by Do et al.

(2013). So one cannot simply argue that external trade intervention is an inferior form of social protection; any such protection would have a similar feature. Trade interventions will probably entail some price distortions, which must be evaluated against the distor­tions generated by alternative schemes. There are situations in which trade insulation dominates feasible options for protection (Do et al., 2013).

The key point here is to avoid sweeping generalizations about policies. To take another example (possibly even more contentious than trade policy) consider active industrial policies—the effort to encourage selected promising sectors or firms using tar­iffs, subsidies, or tax breaks.^[488] Advocates point to the successes of some East Asian coun­tries with these policies, though sometimes downplaying the failures of other countries with similar policies. Instead of arguing for or against such policies in the abstract, the focus should be on understanding under what conditions these, or other interventions, work.

Possibly any country will have a good chance of success with a reasonably wide range of policies in a context of macroeconomic stability and a capable public administration that can pragmatically choose sensible interventions and minimize the damage from mis­taken ones. But will that be enough? The next section turns to another set of potentially important initial conditions related to the distribution of wealth and income.

22.8.