Economists, Policy Reform, and Political Economy
by John Williamson, The World Bank
Keynote address to a conference at the Rajiv Gandhi Foundation
December 6, 1998
© Peterson Institute for International Economics
This speech was given while Mr. Williamson was the Chief Economist for the South Asia Region at the World Bank.
Most economists tend to assume that most market-oriented reforms will benefit most of the people, including most of the poor, for most of the time. We read last year that the Finance Ministry estimates that, of the 13% of GDP spent in subsidies, less than one-third can be justified as what it termed "merit subsidies", meaning subsidies that either promote some good with positive externalities or that redistribute income progressively, and we wonder how the politicians can square their conscienses with continuing to run vast fiscal deficits rather than eliminate the rest of the subsidies. We have concluded that the inflation tax is regressive, and so we are puzzled that inflation has so often (though happily not in India) been readily tolerated, especially by those supposedly of the political left. The most elementary economic theory teaches us that restricting entry will give rise to rents for the few, which normally come at the expense of the many, and so we wonder how the Urban Land Ceiling Act and small-scale industry reservation and sundry other impediments to entry manage to survive. Theory tells us, and empirical analysis has confirmed, that import liberalization accompanied by a competitive exchange rate will lead to faster and more labour-intensive growth, and therefore a more equal income distribution, than import-substituting industrialization, yet the trend to reduce protection in India seems to have been interrupted long before that could be rationalized by optimal tariff theory or the need to retain a bargaining counter. Cheap loans always go to the most privileged of those who satisfy whatever criterion has been chosen to define eligibility, for the very good reason that these offer the least risk and trouble in shoveling out the required sum of money; if you want the poor to get a fair chance at access to credit, you must let them bid for it. Privatization sometimes involves selling a loss-making industry used overwhelmingly by the rich (an airline, say) for a positive price and subsequently receiving tax income as private incentives turn performance around, and getting private investment to replace the need for public funding of capital investment, yet even such transactions, which involve a quadruple benefit to the public finances, are sometimes denounced as a betrayal of the national interest.
The case for such reforms often seems so overwhelming to us economists, on both efficiency and equity grounds, that we have trouble in comprehending why we need to argue the case rather than just identify what needs to be done. Yet the case does need arguing, as all familiar with the somewhat hesitant pace of Indian reform will agree. Let me discuss the political economy reasons why reform may seem less compelling to politicians than it does to most economists.
The Traditional Hypothesis.
It is usually argued that economic reform is politically difficult because its costs are immediate, concentrated, and obvious, while its benefits are delayed, widely dispersed, and many of its beneficiaries do not even know that they will benefit.
The hypothesis suggests, in the first place, that rational politicians (defined as the type who want to be re-elected) will undertake economic reforms only when they believe there is a sufficient time interval between the implementation of the reforms and the next election to enable the public to perceive the benefits of the reform programme. We accordingly tell politicians they ought to make good use of their honeymoon. For example, if George Bush had done what needed to be done in raising taxes in 1989, when the economy could have taken a tax rise in its stride, instead of coming into office mortgaged by electorally superfluous pledges on "no new taxes", there is little doubt that he would have sailed back to power in 1992. And, since most people do not get passionately concerned about balance of payments statistics, a deflationary programme instituted to fix the payments situation is never a short-run vote-winner, but the return of growth once the deficit has been fixed certainly yields political dividends.
In the second place, the hypothesis suggests that policy reform may often be intensely unpopular with a minority, who may accordingly be more active politically than the majority who can expect to benefit. Thus privatization may well jeopardize the interests of the workers and/or managers who expect to lose their jobs, however great the benefits may be to the public finances, the consumers, and even those workers and managers who keep their jobs. Labour market liberalization is likely to harm the interests of the industrial elite whose privileges it undermines, even though it may open up jobs for many more. Higher prices for electric power and greater cost recovery in water are strongly resisted by their immediate beneficiaries, although in these cases it may be possible to persuade people that higher prices are worthwhile if the link can be established in the public mind to improved service quality. I presume that dereservation is strongly resisted by those who would have to face new competition from larger firms, and that repeal of the Urban Land Ceiling Act is strongly resisted by those who hold urban real estate to which the Act does not apply. It is not at all surprising that such minorities should resist the loss of their privileges, however damaging it may be for society if they succeed; indeed, it provides confirmation for the postulate of "economic man". But it does pose a problem to politicians, who can only profit politically from reform if they succeed in making the general public understand where their true interests lie. Moreover, where the minorities who lose are among the underprivileged, their political appeal may be buttressed by concerns for fairness that will resonate with potential gainers unless they receive some form of compensation.
In the third place, the hypothesis suggests that the losers will know what they stand to lose but the gainers will have little idea of what they can expect to win. I once heard Anne Krueger give a graphic description of this phenomenon, and she has supplied me with the following account:
The survey [about which I had enquired] was done in the late l970s in connection with KDI's joint celebration with Harvard of 25 years of Korean growth. I spent, I think, a summer and several other visits at KDI, and they arranged interviews for me. The book was published by Harvard Press in l980, but the interviews were more informal and used as background and I don't recall telling the story in the book. The story was about the man who had been the wealthiest businessman in Korea in the l950s and had built the tallest office building. He had been able to look at all of Seoul...When I visited him almost 20 years later, his office was dark, because he was surrounded by taller office buildings and was "only" president of Korean Exporters Association (or some such). At any event, he had strenuously opposed the reforms...He made much more money after the reforms (and had closed down his "no-good" cement plant to build one to "modern" standards and was exporting that, among other things) and was richer, but others had gained even more than he. He said that he (and many others that I interviewed) had been wrong to oppose reforms; it had been good for him, although he wasn't the richest man in Korea any more! Many other exporters told the story of opposing reforms and then switching (or at least concentrating) lines of activity.
Those who are not aware that they stand to benefit do not provide a political constituency to support reform. To the extent that economists can help illuminate those benefits, we may be able to help reduce this bias in the political system against reform.
Exaggerated Political Pessimism
Having said all that, I also have to say that I believe there is some reason to believe that politicians in general tend to have an exaggerated concern as to how politically dangerous it is to be identified as a strong reformer. There are some reforms that seem to be very popular with the public, even in the short run. Stopping inflation brings an immediate electoral reward (the classic case being Brazil in 1994). So, it seems, does import liberalization: people actually value being able to buy imported goods that were formerly denied them. This may not be very surprising, except that we usually assume that the only people who will notice import liberalization are the firms and workers in the import-substituting industries who have something to lose. Perhaps Indian politicians are now convinced of the converse proposition, that restrictive import policies, at least for onions, can bring quick electoral disaster.
The evidence is also beginning to suggest that reformers do well at the polls. Mrs Thatcher, whose style of reform may not have enchanted all of us but who certainly brought market-oriented reform to Britain, had the longest contiguous term as Britsh prime minister since Lord Liverpool in the 1820s. Carlos Menem, Alberto Fujimori, and Fernando Henrique Cardoso, are the most conspicuous Latin American presidents to have been re-elected after introducing market-oriented reform programs. The same story can be told in Australia and New Zealand after they introduced reforms in the mid-1980s. In fact, of the 11 episodes of reform that we studied in a conference I organized in 1993,1 the government that introduced reforms was re-elected at least once in no less than seven cases.
Nor is the evidence that economic reform can be a winning political issue purely anecdotal. An econometric study of electoral performance in Latin America over the 13-year period 1982-95 concluded that a record as an economic reformer was the second most potent vote-winning strategy, second only to success in resolving a civil war.2 Obviously this refers only to a particular historical situation, but it suffices to raise the possibility that the public has a better understanding of where its interest lies than do those politicians who at times seem so hesitant about embracing reform.
Some Policy Implications
Let me conclude my remarks this morning by drawing out some implications for what the above analysis implies for those of us economists who share the judgments about where the public good lies that I sketched at the start of this paper.
First, it surely imposes on us a professional duty to do our best to inform the public about the nature of the benefits that we see from reform, given that reform often fails to find a political constituency because the general public is ignorant of its potential benefits. There may be a certain tension between this duty and our other professional duty to be objective, but we should not resolve this by a purist refusal to get engaged in the policy debate. We need to try and maintain a proper balance, arguing our case in as transparent a way as we can, making clear where there are judgments with which others can reasonably disagree, acknowledging where limitations on our knowledge prevent a definitive conclusion, but still seeking to make a persuasive case for actions that we believe would advance the public interest. Those of us who are sensitive to concerns of fairness and distributional equity will naturally seek to mould our policy proposals to reflect those concerns.
Second, let us not imagine that our advocacy will lead to a sudden and dramatic policy switch. I think there is no doubt that the cumulative force of the policy advice given by economists does influence the course of public debate, and thus ultimately of public policy, but the impact of any one of us on any one occasion is unlikely to be measurable, let alone decisive, in a country the size of India.
Third, I am convinced that advocacy is ill-served by dogmatism. Those who go round asserting that all markets should be liberalized, when it is pretty obvious that it was the vastly premature liberalization of the capital account that created the conditions for the explosive contagion that has laid East Asia low, risk discrediting all liberalization. Those who assert as a matter of faith that everything should be privatized risk discrediting all privatization when there are examples of failed privatizations as glaring as the Bangladeshi banks. It can be argued that resolution of the East Asia crisis would be served by the large-scale injection of public equity into the banks (as happened in Chile after 1982), and such policy proposals need analyzing on their merits and should not be rejected on the basis of knee-jerk dogma. Reforms sometimes fail, and when they do we need to research why, as honestly as we can, and be prepared to modify our policy proposals in the light of our findings.
Fourth, I suggest that we should endeavor to give options to policymakers, who after all like to earn their keep. If you tell them that the choice is between Scenario A, "business as usual", and Scenario B, your wish-list, the easiest way for them to make their mark is to delete all the bits of Scenario B they do not like. If instead you offer them more than one way of achieving the essential objectives being sought by reform, then their choices are already an inherent part of the reform design. Give policymakers a chance to be constructive before you blame them for messing up your reform programme.
Today's seminar is about reforms and the poor. That many reforms are better for the poor than the status quo seems pretty clear, as I have already argued. Whether reforms in general are pro-poor more than pro-rich is a different question, and one about which I am much less sure of the answer; perhaps I will have a better idea later in the day. How reforms can be designed with a pro-poor slant is another basic question, which I am delighted to find you are addressing at this workshop. I look forward to hearing your debate on these issues.
1. John Williamson, ed., The Political Economy of Policy Reform (Washington: Institute for International Economics, 1994).
2. Carlos Gervasoni, "Economic Policy and Electoral Performance in Latin America, 1982-1995", M.A. thesis, Stanford University, 1995.