Is the Washington consensus dead?

• Author: Lal, Deepak

In the postwar years, most Third World countries turned inward partly in response to what they thought were the disastrous consequences of their 19th century integration into the world economy as the global economy collapsed in the Great Depression. The seeming success of Soviet central planning under Stalin also persuaded the leaders of these newly independent countries to substitute the plan for the market. Planning was all the rage, with the state seeking to control file commanding heights of the economy. Furthermore, the many theorists who created a seemingly "new development economics" provided the intellectual basis for the complex system of dirigiste controls on "anything that moved" (as one wit put it) in a market economy.

In the early 1980s, I wrote a small book, The Poverty of "Development Economics" (Lal 1983), which attempted to summarize the logical arguments and empirical evidence against what I called the "Dirigiste Dogma," which had done so much damage to the prospects of the Third World's poor. That book, which acquired some notoriety, if not franc, marked the neoclassical resurgence against the Dirigiste Dogma as many developing countries began to adopt (at least partly) the classical liberal policy package known as the Washington Consensus. (1) The reversal of dirigisme in the 1980s and early 1990s, particularly in China and India, but also in many other parts of the Third World, led to the surge in per capita incomes and the largest reduction in structural poverty in human history.

The Washington Consensus

The Washington Consensus is the standard classical liberal economic package, consisting of free trade, Gladstonian finance, and stable money. It also requires a good government that promotes Adam Smith's "popular opulence" through promoting natural liberty by establishing laws of justice that guarantee free exchange and peaceful competition. Egalitarianism is rejected as the norm for deriving principles of public policy, because of the contingent fact that there is no universal consensus on what a "just" or "fair" income should be, despite the efforts by moral philosophers to justify their particular prejudices as the dictates of reason. But classical liberals, from Smith to Friedman and Hayek, have also recognized that the state should seek to alleviate absolute poverty through targeted benefits to the indigent and disabled. For various merit goods--health, education, and possibly housing--these involve in-kind transfers through some form of voucher.

Over the past two decades, there has been a backlash against the Washington Consensus. Gerald Meier, a proponent of the new development economics, which draws from models of imperfect information, coordination failures, multiple equilibria, and poverty traps, claims that "with imperfect information and incomplete markets, the economy is constrained Pareto inefficient--that is, a set of taxes and subsidies exists that can make everyone better off" (Meier 2005: 119-20). (2) This echoes a similar claim made by Greenwald and Stiglitz (1986).

Meier (9.005: 124) commends Rodrik (1995) for emphasizing "coordination failures," and for demonstrating "how the South Korean and Taiwanese governments got interventions right." He claims that this viewpoint, and the Murphy, Shleifer, and Vishny (1989) modeling of the "Big Push," validate the old development economics of Nurkse (1953) and Rosenstein-Rodan (1943). He supports Krugman's (1993) belief that in its earlier incarnation it was not persuasive because its ideas were not formalized in mathematics. But as Stiglitz, (Krugman's discussant), rightly noted: "That we can write down a model of a phenomenon proves almost nothing. It does not make the idea right or wrong, important or unimportant." As regards Rodrik's views about smart dirigisme in Korea and Taiwan, Little (1994) convincingly showed that social rates of return to investment in these countries were inversely correlated with the degree of dirigisme.

In this article, I examine the basis of these "new" theoretical curiosa questioning the Washington Consensus, as they now form the basis of the advice given by the "new dirigistes" to alleviate Third World poverty. I also briefly examine the claim that the Chinese economic "miracle" was created not by the replacement of the plan by the market, but by various forms of dirigisme, which has led to a new "Beijing Consensus" to replace the defunct Washington Consensus.

Poverty Traps

I have a tremendous sense of deja vu at reading this "new" theoretical literature. The "'poverty trap" view, which now has wide currency among the young, is just a resurrection of the "vicious circle of poverty" arguments of the 1950s---except now it is attired in sophisticated mathematical garb. Peter Baner's ([1987] 2009: 173) castigation of this view still stands:

According to this notion, stagnation and poverty are necessarily self-perpetuating: poor people generally and poor countries or societies in particular are trapped in their poverty, and cannot generate sufficient savings to escape from the trap. This notion became a cornerstone of mainstream development economics. It was the signature tune of the advocates of foreign aid throughout the 1950s.... Yet it is in obvious conflict with simple reality. Throughout history, innumerable individuals, families, groups, societies, and countries--both in the West and the Third World--have moved from poverty to prosperity without external donations. All developed countries began as underdeveloped. If the notion of the vicious circle were valid, mankind would still be in the Stone Age at best. But validating the vicious circle is precisely what the purveyors of poverty traps seek to do.

The theoreticians have fiddled with the standard neoclassical growth model developed by Nobel laureate Robert Solow (1970). He showed that with a given per capita savings rate (as a fixed proportion of per capita income), and given rates of increase in the labor force (directly from population growth and indirectly through labor augmenting technical progress), a poor economy starting off with a low stock of capital per head (and hence low per capita income) will converge to a higher steady state capital per head and income per capita. In the steady state, both income and the capital stock will be growing at the "natural rate of growth" given by the sum of the rate of population growth and labor augmenting technical progress, with per capita income constant at its higher steady state level. But during the "traverse" to this steady state from the initial position, the economy's growth rate of per capita income and capital per head will be faster than this natural rate of growth. The poorer the country, the faster will be its rate of growth of per capita income as it reaches the steady state capital and income per head of the leaders in the world economy.

This convergence in per capita incomes, however, is conditional on the relevant countries having a common institutional and legal framework for economic activity, as Barro and Sala-i-Martin (1991, 1992) have shown. Regions within large economic units like the United States, Japan, and the European Union would meet this condition, and there is evidence of this "conditional convergence" in the growth rates of their regions. Clearly, within this theoretical framework there are no poverty traps.

Now enter the new theoreticians. They find that many countries, particularly in Africa, though poor, do not seem to be growing faster than their richer peers, as mainstream theory predicts. So, they argue the answer must lie in there being "multiple equilibria" where some countries are stuck in a steady state with a low income per capita and a low capital stock per capita, rather than moving smoothly on to the high income per capita steady state of the Solow framework.

[FIGURE 1 OMITTED]

So how do you theoretically generate these multiple equilibria? Simple, assume that instead of the per capita savings rate being a fixed proportion of income per capita, it is a lower share at low levels of income per capita, and then--after some threshold level of income per capita--it suddenly jumps to a higher proportion of income. Savings per capita as a function of per capita income then becomes steeply S-shaped with respect to the capital stock per worker. Savings is low at low levels of capital per worker, increases substantially at an intermediate level, and then levels off.

Figure 1 is a simple diagrammatic explication of the standard Solow growth model and the multiple equilibria grafted on to it by the purveyors of a savings "poverty trap." (3) The standard neoclassical production function (Oy) shows output (real income) per capita (y 5 Y/L) as a function of capital per capita (k 5 K/L). The ray nk shows the investment required to maintain any given capital-labor ratio constant, when the sum of the rate of population growth (p) and labor-augmenting technical progress (t) is equal to n (n 5 p 1 t). The savings function (sy) is a constant at a low rate until capital per capita is [y.sub.H], and then jumps to a constant higher rate when the capital per capita is k*. There will then be two stable steady state capital-labor ratios. If the country starts off with capital below ka, it will reach the low steady state capital-labor ratio [k.sub.L] and stay there. It is in a poverty trap as income per capita cannot...