Comments: Ravi Kanbur
- Ke-young Chu, Sanjeev Gupta, and Vito Tanzi
- Published Date:
- May 1999
I am going to focus my comments on the evolution of redistribution policy in Organization for Economic Cooperation and Development (OECD) countries. Aníbal Cavaco Silva’s entire paper is on the European Union, and half of Alberto Alesina’s paper is on OECD countries. So I am going to take off both my operational hat and my developing country hat and talk about frameworks and concepts as related to OECD countries, which I think will raise some interesting issues.
The framework put forward by James Mirrlees more than twenty-five years ago—in the paper that won him the Nobel Prize in Economics—is useful in thinking about the evolution of redistribution policy. Three components of his paper are useful for our discussion. First is the concept of inherent inequality in a society, based on, for example, skilled/unskilled wage differentials and asset inequality—the sorts of things that we have been discussing at this conference. If authorities let this inherent inequality play out, with no intervention whatsoever, the inherent inequality will be fully reflected in the outcomes. However, if policymakers’ egalitarian sentiments are aroused, and they want to intervene, they will find the second component of this analysis, namely, the degree of egalitarianism professed. And if authorities redistribute consistent with their egalitarian instinct, they will see the third component, the incentive effects. As policymakers try to redistribute resources away from those people at the top to those at the bottom of the distribution, as they take away and give resources, there will be incentive and disincentive effects.
We now have three components: (1) the inherent inequality in the system, given skilled/unskilled wage differentials and asset inequality; (2) the degree of egalitarianism espoused; and (3) the level of incentive and disincentive effects. Together, they produce the degree of redistribution that one sees in each system. We can use this framework to visualize the exercises that are being discussed in the papers presented at this conference.
Holding constant component (1), the degree of inherent inequality, and component (3), the level of incentive effects, as component (2), the degree of egalitarianism espoused, varies, we find the degree of attempted redistribution will differ among countries. That is what you see in Atkinson’s paper, when he discussed the degrees of intervention in OECD countries. Holding the other things constant, if disincentive effects become stronger, then the extent of redistribution that you would want to achieve for any given degree of egalitarianism will be less. Again, Atkinson pointed this out in his discussion: with globalization, policymakers may move toward less redistribution, if the disincentive effects of redistribution become greater.
But what I wish to emphasize today is that, as the degree of inherent inequality in the system changes, for any given degree of professed egalitarianism and any given degree of incentive effects, so will the amount of redistribution you will attempt to achieve. Thus, if the inherent inequality in a system decreases, so will society’s redistributive efforts; as the degree of inherent inequality increases, so will society’s redistributive efforts for any given degree of espoused egalitarianism and incentive effects.
This is where a paradox emerges. Looking at the evidence presented in the papers at this conference, we could roughly divide the OECD’s postwar history into two phases. The first ran from 1945 until about 1980, when the degree of inherent inequality in the system decreased because of reductions in skilled/unskilled wage differentials and asset inequality. The second phase has occurred since 1980, when the degree of inherent inequality in the system reversed course and increased, owing to, for example, technical change and trade.
My earlier analysis might suggest that in the 35 years following the Second World War (1945–80), redistributive efforts should have decreased in OECD countries because inherent inequality decreased, and that since 1980, redistributive efforts should have increased because inherent inequality has started to increase. However, from the papers of Alesina, Cavaco Silva, and Atkinson, we see that the pattern has been exactly the opposite: from 1945 to 1980, a period in which inherent inequality fell, the degree of redistribution rose; and since 1980, while inherent inequality has increased, at least in the United Kingdom and the United States—although the OECD averages may tell a different story—redistributive efforts have waned.
Alesina’s explanation for this, at least for the period from 1945 to 1980, is the political economy. But I think the political economy interacts tightly with the technocratic formulation. How? Well, let’s go back to the original framework but with more information. Suppose we are told that poor people live in larger families with more children, have higher unemployment rates, have housing problems, and are older than affluent people. Technocratically, we can design more efficient redistribution mechanisms by using this categorical information. And that, of course, is exactly what was done in 1945, in the Beveridge plan for the U.K. welfare state. Using this categorical information, we designed a more efficient redistributive mechanism. But now we come to the point that Alesina made. Over the following 35 years, the political economy kicked in, these groups became interest groups, and sociodemographic changes led to huge changes in the size of these groups. Thus, if policymakers had a clean slate now, they would not design the system the way we did in 1945, although that was appropriate then. And, as I said, since 1980, we have had increases in inherent inequality, yet we have moved away from—and now experience almost a revulsion for—what we consider the excesses of those 35 years.
Where do we stand now? Where do OECD members’ electorates stand? If we look at Cavaco Silva’s paper, we see an example of this problem in the Portuguese experience. He says that, on the one hand, electorates want politicians to say that they will improve equity; but when the politicians are in power, the electorates are not so sure they want redistributive measures introduced. On the other hand, the electorates do seem to care about specific groups—for example, minorities and excluded groups—although they are not too sure about cash transfers to these groups. Even with regard to noncash transfers, the electorates worry about things like scroungers on welfare. Thus, it seems to me that electorates in OECD countries are conflicted and schizophrenic. They are fearful of excessive redistribution in a period in which—the original Mirrlees framework suggests—the new insecurities and inequalities actually require more redistribution. In Alesina’s paper, we have econometric confirmation of this: the blocking coalitions of public employment groups, public employees, pensioners, and other interest groups, on the one hand, and electoral acceptance of governments acquiescing to these coalitions, on the other.
Something odd is going on. We are caught moving away from the old system because by 1980 it was too redistributive since inherent inequality had been falling for 35 years. Yet since 1980 inherent inequality has been rising. This is the central paradox in our discussions.
Where do we go from here? I will answer with two end notes. You often hear the statement that we need to step up education and retraining in burgeoning fields like computer programming, to raise income, not transfer payments. To illustrate why that strategy will not be fully satisfactory, consider an unemployed, 50-year-old male steelworker in France, the United Kingdom, or the United States, or an unemployed 40-year-old female mill worker in a New England town. The notion of turning these workers into computer programmers by retraining them is problematic—yet they are going to live for another 30 years.
Society must face the fact that we will need some fundamental, bottom-line transfers, the nature of which will be determined by social circumstances. We can discuss many ways to accomplish this: the efficiency of different methods, the social framework, what constitutes a socially dignified form of transfer, and so on. But the bottom line is that for the next 30 years, there will be a group of people who will have to be supported by the computer programmers in one way or another.
My second note on redistribution concerns education. Education needs to be equally accessible to all, but the bottom line is that somebody, somewhere, must pay for it. The resources will have to come from somewhere. Fundamental distributional questions are not avoided simply by invoking “education” rather than transfers.
Where is the political economy going? I have given you technocratic fundamentals, so your guess is as good as mine. My forecast is that we will muddle through, as usual. We will do all the things that have been discussed in the papers presented here: we will provide more education here, more employment subsidies there, and so on—and maybe those are the right things to do. But unless some form of strong leadership comes forward and suggests a comprehensive review of these issues, the alternative is change through a social crisis.