February 14, 2007
Why America will need some elements of a welfare state
Is globalisation a leading cause of rising inequality in high-income countries? The outcome of the debate on this question may determine whether the US will remain open to trade. If policymakers do not craft an imaginative response, protection against imports may be the outcome, regardless of its (non-existent) merits. Ben Bernanke, chairman of the US Federal Reserve, laid out the issues in a thought-provoking speech last week.* He embedded his analysis in three principles: “That economic opportunity should be as widely distributed and as equal as possible; that economic outcomes need not be equal but should be linked to the contributions each person makes to the economy; and that people should receive some insurance against the most adverse economic outcomes, especially those arising from events largely outside the person’s control.” The question is how these principles apply today. It has become more compelling as inequality has risen. The remainder of Martin Wolf’s column can be read here (FT.com subscribers only). Discussion from our guest economists is free.
Brad Delong: Martin, you really shouldn’t cite Alan Reynolds of Cato here…
Take a look at Gary Burtless’s assessment of Reynolds’s work (”I take seriously some of Reynolds’s criticisms of the data…. Reynolds points to some serious problems, and in many cases fair-minded experts will agree with him…. [H]e is strongly critical of data series that do not support his views… silent about equal or more serious problems with data sets that show little change in inequality. Rather than do the hard work needed… he cherry-picks evidence to attack researchers whose results he finds displeasing”).
Take a look at what Piketty and Saez - whom Reynolds accuses of being “ready and willing to supply whatever is demanded” to fill “a psychological rather than logical need” - have to say (”Reynolds has picked some of the facts in order to provide a very skewed view. Most of the scenarios described by Alan Reynolds… would imply that… accumulated profits would appear as realized capital gains on the owner individual tax return when the business is sold. Yet, our top 1% income share series including realized capital gains has also doubled from 10.0% in 1980 to 19.8% in 2004…”).
Recall Reynolds’s egregiously false trashing of Steve Pearlstein last year, and his equally egregiously false misrepresentation in the Washington Times of his trashing of Pearlstein.
Recall that we have seen Reynolds on this issue back in 1992, misrepresenting a Fed research study of rising inequality over 1983-1989 to trash it.
Or simply note what Reynolds wrote last December in the Wall Street Journal:
“Senator-elect Jim Webb recently complained on this page of an ‘ever-widening divide’ in America, claiming ‘the top 1% now takes in [twice as large a] national income [share as in] 1980.’… The incessantly repeated claim that income inequality has widened dramatically over the past 20 years is founded entirely on [Piketty and Saez’s] seriously flawed and greatly misunderstood estimates of the top 1%’s alleged share of something-or-other…”
And note that his use of the phrase “past 20 years” was not innocent. As Reynolds wrote on Mark Thoma’s Economist’s View (http://economistsview.typepad.com/), he had said “there is no clear evidence of a sustained and significant increase in inequality since 1988…. I very carefully did not say there was no such evidence about 1981-87…”
“No… sustained.” “No… significant.” “No… clear.” “Since 1988… I very carefully did not say… 1981-87.” You’ve seen this kind of rhetoric before. It reminds me of anti-evolution creationists, lobbyists for tobacco companies, global-warming deniers, and the fringe who claim that the U.S. and western European economies are to the right of the peak of the Laffer curve.
Martin: your social value is highest when you are an effective filter of information. You should give somebody who writes “I very carefully did not say there was no such evidence about [rising inequality for] 1981-87…” the weight they deserve.
Posted by: Brad DeLong | February 14th, 2007 at 3:59 am | Report this commentMartin Wolf: I stand corrected. In any case, if Reynolds is wrong, as you argue, the argument for an American welfare state, including income redistribution, is strengthened, isn’t it, Brad?
Posted by: FT Forum - Martin Wolf | February 14th, 2007 at 10:48 am | Report this commentAdrian Wood: In his characteristically interesting article, Martin asks ‘three big questions’. These are my answers,which are not quite the same as his.
1. Is it true? Yes, for sure, if the question is ‘has inequality risen, particularly at the very top of the distribution?’ There is doubt only about precisely how much.
2. What has caused it? In the middle of the distribution (say between the upper and the lower quartiles), the mechanisms of trade theory, including the extensions made by Robert Feenstra and Gordon Hanson, had a big impact. But most of the hurt they inflicted on unskilled workers was some time ago – in the 1980s and early 1990s.
At the very top, the main cause is different: the intersection of globalisation with a version of the winner-takes-all argument. Reductions in travel and communications costs mean that highly skilled workers can now operate worldwide (not only in capital markets, which Martin mentions), which has hugely leveraged their earnings.
The same process explains why the wageof the average worker in rich countries has lagged and that of the average worker in poor countries hascaught up. The former have lost their privileged access to working with the highly skilled, and the latter have gained it. The formal theory and somestatistical evidence for this mechanism are in an article by Edward Anderson, Paul Tang and myself in Oxford Economic Papers, 58/4, November 2006 (of which I sent Martin a copy a while ago).
3. What should be done about it? Martin, like Ben Bernanke, mentions some useful things, but neither refers to the really big thing, which is to restore the progressivity of the personal tax system – a point already made in these columns by Larry Summers and myself. Higher taxes would be needed to finance the welfare state elements that Martin mentions. Moreover, the enormousearnings of workers at the very top are in an economic sense mainly rents – which can be taxed away without loss of economic efficiency.
The problem intaxing these rents is that the world does not have a unified tax system: no one national tax authority can risk raising taxes on internationally mobile workers. To solve this problem, national tax authorities would need to cooperate. Cooperation would be a major administrative and political challenge, but I wish Martin would help to put it on the agenda for serious discussion (as he has with climate change), since it certainly is not there now.
Posted by: Adrian Woods | February 14th, 2007 at 7:36 pm | Report this commentAllan Meltzer: Your column on inequality speaks to an important political problem — providing opportunity widely. It reaches an answer that is partly correct. As you say, the problem is not globalization. It is inadequate and inferior education for large pats of the population and their children.
Government has been addressing that problem for a generation., A Nation at Risk is 40 years old. Many subsequent reports and political commitments addressed the problem. Governments seem unable to make reforms that work. Research shows that family involvement, concern, and interest can do a lot. We should increase incentives for the students and their parents tied to accomplishment. Also we must separate the redistribution of funding of education from the delivery of educational services. As the late Milton Friedman urged, introduce competition and experimentation. We surely know by now that we will not do much to improve education without increasing incentives for teachers, students and their parents.
One thing I have learned in more than 50 years of teaching is that the main problem is learning. I do not expect to do much for students who do not choose to learn. I can encourage students to learn, try to interest them, motivate them, and encourage them. All that helps, but it works best for students who want to learn.
Posted by: FT Forum - Allan Meltzer | February 15th, 2007 at 11:28 am | Report this commentMartin Wolf: I think I agree with Adrian Wood and apologise for not having cited his paper. It seems to me evident that a more generous welfare state requires higher taxation and, in the US, that must mean higher taxation of those who have done so sensationally well in recent years. I agree, too, on his arguments about tax co-operation. The only question I have is how far these “rents” are the rewards that attract talented people into the tournament. After all, we really don’t want every bright person to become an academic, do we? But the big point I wanted to make is that without more security, it may be impossible to get people to tolerate the present version of liberal capitalism. This is a point often made by Harvard’s Dani Rodrik. I don’t always agree with him. But I do in this case.
Posted by: FT Forum - Martin Wolf | February 15th, 2007 at 5:49 pm | Report this commentWillem Buiter: Martin’s Column “Why America will need some elements of a welfare state”, refers extensively to a recent speech by Ben Bernanke (”The Level and Distribution of Economic Well-Being”). The Column calls for a two-track response - one on the merits of Martin’s argument and the second on the appropriateness of the Chairman of the Fed speaking out on issues such as this. I only deal with the second issue here.
I believe it is a serious mistake for central bankers to express public views on politically contentious issues outside their mandates. The mistake is no less serious for being made so commonly by central bankers all over the world.
Central bank Governors have a lengthy and unfortunate track record of holding forth in public on matters that are outside the domains of their mandate (in the case of the Fed, monetary policy and financial stability) and their institutional and professional competence. With the exception of the Governors of the Bank of England and the Reserve Bank of New Zealand, every Governor on the block appears to want to share his or her views on necessary or desirable fiscal, structural and social reforms. Examples are social security reform and the minimum wage, subjects on which Alan Greenspan liked to pontificate when he was Chairman of the Board of Governors of the Federal Reserve System. Jean-Claude Trichet cannot open his mouth without some exhortation for fiscal restraint or structural reform rolling out. In the case of Chairman Bernanke’s speech, equality of opportunity, income distribution, teenage pregnancy and welfare dependency are clearly not part of the (admittedly broad) three-headed mandate of the Fed: maximum employment, stable prices and moderate long-term interest rates. The regrettable fact that the Secretaries of the Treasury, Ministers of Finance and of the Economy and other political office holders often make the symmetric mistake of lecturing supposedly operationally independent central banks on what they perceive to be the central bank’s duties (mostly lower interest rates and a weaker exchange rate) does not justify the central banks’ persistent transgressions.
When the Head of a central bank becomes a participant, often a partisan participant, in public policy debates on matters beyond the central bank’s mandate and competence, the institution of the central bank itself is politicised and put at risk of becoming a partisan-political football. This puts at risk the central bank’s operational independence in the management of monetary policy and in securing financial stability.
Central bankers, Mr. Bernanke included, should ’stick to their knitting’ (if I may borrow Alan Blinder’s phrase). Being the head of an institution with the national and global visibility of the Fed or the ECB gives one an unparalleled platform for addressing whatever one considers the great issues of the time. The temptation to climb that unique pulpit must be near-irresistible. Nevertheless, unless the text for the sermon concerns monetary policy or financial stability, that temptation is to be resisted in the interest of the institutional integrity and independence of the central bank.
Posted by: Willem Buiter, London School of Economics | February 15th, 2007 at 6:39 pm | Report this commentMartin Wolf: Willem is right. He usually is. Ben Bernanke’s was an interesting speech (and useful for me), but he probably should not have given it. Should the chairman of the Fed even defend free trade? If not him, who? But I do agree with Willem that Mervyn King has been admirably self-disciplined.
Posted by: FT Forum - Martin Wolf | February 15th, 2007 at 7:04 pm | Report this commentMartin Wolf: Let me comment on Allan Meltzer, with whom I largely agree. I think there is more to what is required now than better education. But if there is to be anything close to equality of opportunity, education is the heart of the matter. As I indicated in the column, I would support private provision of services paid for by taxes. Vouchers are one way of doing this. The Swedish system of allowing people to set up schools and receive a given sum per pupil from taxation is another. I also agree with Allan that it makes sense to provide incentives to the pupils and even their parents.
In the US today, however, I think more is required. Some form of universal health insurance is, I believe, necessary and inevitable. It can hardly be less inefficient than the present system.
Posted by: FT Forum - Martin Wolf | February 15th, 2007 at 7:11 pm | Report this commentTito Boeri: I have three remarks.
First, rising inequality in the US is not an “open issue”. It has been widely
documented by studies based on a variety of sources, including household surveys using post-tax records. Thus, the tax measurement problem raised by Reynolds is not relevant.
Second, as documented by Atkinson, Saez and Picketty, a driving force of this rising inequality is inequality at the very top of the distribution, the top 0.5 per cent or an even a smaller fraction of the population. This has little to do with the welfare state, a bit more with the progressiveness of taxation. Something similar has happened in the UK and Germany. It is a matter of globalisation, superstars, stock options, etc..
Third, unlike Europeans, US citizens believe that the poor people are just lazy individuals. A worker not paying income taxes because she earns less than 12,000 USD per year is considered a “luckie duckie”. In Europe it is just the other way round: poor people are deemed to be unlucky, hit by the “events largely outside their control” Bernanke was referring to in his speech. Thus,
Posted by: Tito Boeri | February 15th, 2007 at 7:24 pm | Report this commentI very much doubt that the US can adopt any EU-type welfare system. But even if it were to do it, this would perhaps help in reducing the increase in the top decile ratio (as happened in the 1960s and 1970s as observed in the long
series built up by Tony Atkinson), but would not do much to the top 0.5 per cent star incomes. The only way to prevent this new type of inequality to increase is international policy coordination of wage ceilings of CEOs, actors and football players. Is this on the international policy co-ordination agenda?”
Willem Buiter: Poverty bothers me. Inequality does not. There, I said it.
Lest there be any ambiguity, when I say ‘poverty’ I mean absolute poverty, not the income level of the lowest 2 deciles of the income distribution, or whatever definition of relative poverty is currently in favour. I recognise that even absolute poverty concepts are time-, place- and context-specific. But for the US or the UK today, I believe that, with a research grant not exceeding 0.001 per cent of GDP, I could come up with a generally acceptable definition of absolute poverty.
One reason Martin wants us to worry about growing inequality is that it causes declining equality of opportunity? If it did, I would indeed be worried. But where is the evidence and what is the mechanism? I am not even clear about how to formulate the statement in a way that would make it even potentially verifiable. In my world view, both the degree of (ex-post) inequality (in life-time economic wellbeing) and the degree of equality of opportunity are ‘jointly endogenously determined’ - proximately by a socio-economic-political mechanism/process and at a deeper level by technological and cultural/historical factors. Inequality causing equality of opportunity (or the other way around) doesn’t really make sense to me.
It is an interesting policy issue whether there are feasible (or at least conceivable) policy interventions that, in a society such as the US, might produce a lower degree of inequality and a higher degree of equality of opportunity (I would be happy to support policies that produce a greater degree of equality of opportunity, regardless of what they do to ex-post inequality, as long as they do not condemn the losers to poverty).
Martin also argues that rising inequality makes losing a job costlier, more objectionable and so more resisted. Perhaps. But again, the statement has no policy implications because it makes no sense to treat rising inequality as an exogenous event or a policy choice. The relevant policy question is: are there policy options that reduce inequality (which cet. par. might indeed lead to greater resistance to job losses) and do not affect the other determinants of (a) the likelihood of losing a job and (b) the cost of losing a job. Take policies that weaken unions as a policy that increases inequality among wage earners. This policy could also reduce the length of the typical unemployment spell. This would reduce the resistance to job losses both directly (by weakening unions) and indirectly, through its effect on the value of the outside option.
Martin also notes that there is evidence that inter-generational mobility is smaller in the US and the UK than in the Nordic countries and even Germany. Apart from the fact that inter-generational mobility and ex-post equality are not necessarily related in any simple, obvious way, the explanation he offers for the phenomenon – that the relative poverty of the parents is visited upon the educational achievements of their children - is not convincing. It would seem that in countries as rich as the ones under consideration, only absolute poverty of the parents would prevent their children from reaching the highest levels of educational achievement consistent with their innate abilities.
None of this alters the fact that the persistence of absolute poverty, and at times abject absolute poverty, in a country as rich as the US is a disgrace and a monument to political and communal failure. The reforms of the welfare state that are being proposed are, however, unlikely to do much about this. A quick, permanent solution would be achieved by the following package of measures.
Quality pre-school, primary and secondary education should be available to all, regardless of ability to pay. This calls for public funding of pre-school, primary and secondary education. Even generous public funding is not sufficient, however, to ensure quality education. That much is clear from the experience of state schools in deprived inner city areas in both the US and the UK. The power of the suppliers of education in the state sector, especially the teachers’ unions, is an important obstacle to quality education. Competition between state schools and private schools, through the use of vouchers and similar market-friendly mechanisms, must be encouraged. It is likely that legal measures will be required to reduce the power of the teaching unions. Since public primary and secondary education in the US is mainly locally funded and regulated at the state level, nation-wide reform is harder there than in a unitary state like the UK, although local experimentation is easier in the decentralised structure of the US. The public funding of pre-school primary and secondary education may have to be Federalised.
Universal health insurance (or an entirely tax-funded basic health care system, but with a much greater role for private provision of health services than in the UK) is another no-brainer. Any sensible system of universal health insurance will, however, have to get rid of the fiscal distortions that have resulted in the economic nonsense of health insurance being coupled with employment rather than with being alive. The same holds for the fiscal distortions that have coupled pensions not just with employment but with employment in a particular firm. Reforms that would make it as easy for a home worker as for an employee of the central bank to save for a pension and to insure against ill health would also ensure the portability of pensions and health coverage, thus enhancing labour mobility between enterprises.
James Tobin and Milton Friedman agreed on only one thing: the desirability of a negative income tax (preferably levied on a consumption tax base). This was also the only issue on which both of them were correct. It would replace all other entitlement programmes, including Social Security and Medicare. Medicaid would be redundant as a result of the health care reforms of the previous paragraph. I would set the guaranteed income (for every legal resident 18 years or older) at $15.080 per year (this is the proposed new Federal minimum wage of $7.25 per hour times a 40 hour work week times 52 weeks). The minimum wage itself could then be abolished, as an unnecessary labour market distortion.
Focusing on the introduction of these changes would do more for security, well-being and efficiency than bellyaching about rising inequality.
Posted by: Willem Buiter, London School of Economics | February 15th, 2007 at 8:00 pm | Report this commentBrad DeLong: Martin Wolf takes a look in his Magic-8 ball at America’s future and sees… social democracy:
I wish I could see it. But I can’t…
One would have thought that the rise in the value of a sheepskin from a 30 per cent lifetime wage premium over a high-school diploma in 1975 to a 90 per cent premium in 2005 would have called forth an extraordinary wave of public support and financing for investment in education that would have pushed that premium down somewhat: lots more Americans should be getting a higher education now than were getting one in the mid-1970s. But they aren’t.
One would have thought that the increasing importance of pension and health benefits in a more medically-capable and longer-lived society would have made American workers enthusiastic about the “flexicurity” agenda pursued by Labor Secretary Bob Reich and others in the first phase of the Clinton administration. But they weren’t - at least not in a manner visible to me or to decisive legislative votes like Sen. Breaux or Rep. Tauzin. And the Labor leaders weren’t either. “Burial insurance. We don’t want burial insurance” was the refrain that I heard from my spear-carrier perch in the back of the room.
One would have thought that initiatives like Barney Frank’s “grand bargain” - the left supports trade liberalization if the right supports social democracy - would have gained more traction with a left that realises that trade restrictions are negative-sum when they aren’t smoke-and-mirrors and a right that recognises the potential economic and soft-power security gains from an even more interdependent world. But they haven’t.
America’s politics are really weird. Lots of things that make obvious and indisputable sense in America simply don’t happen for one or another strange political reason.
In my view, those who benefit the most from America’s open economy are consumers, elderly home-sellers, middle aged mortgage-borrowers, construction workers, and those producing and selling high-end consumer goods who benefit from consumer spending ultimately and indirectly burt surely financed by low interest loans from the People’s Bank of China. They don’t know how much they gain. Those who lose the most from America’s open economy are manufacturing and other workers who find themselves competing with imports. They know how much they lose.
A substantial and relatively rapid fall in the dollar unaccompanied by macroeconomic distress would, I think, make a big difference. But absent that, I don’t see any political coalition assembling in America for freer trade. Maintaining stasis will be the best we can hope to do.
Posted by: Brad DeLong | February 16th, 2007 at 4:03 am | Report this commentMartin Wolf: Let me comment on Tito’s remarks. First, I accept what he says on the data on US inequality, which follows Brad de Long’s earlier comment. Second, I do not think there is much chance of international co-operation to limit incomes of superstars.
Third, I agree that Americans do not care about inequality. As he notes, Americans seem to accept the idea that the poor have only themselves to blame, while Europeans believe the poor can blame society.
I find this American attitude attractive. It is probably more productive, in the long run, since it must motivate people to try harder. But it is hardly true. In particular, it is difficult to believe there is equality of opportunity between someone born to a prosperous and secure upper middle class household and someone born in a crime-ridden inner city area. This is one justification for a more ambitious welfare state. That should be done with generous educational vouchers, not with greater spending on existing public education. I also think that the contingent loan scheme we have introduced in the UK, to finance university education, would be highly desirable in the US.
Again, even if one were opposed to redistribution of income, the efficiency and equity case for universal health insurance looks overwhelming.
This combination would surely reduce at least some of the protectionist sentiment shown in today’s Democratic Party. When such a basic forms of family security is attached to a specific job, its loss is a disaster. But companies cannot underwrite this sort of insurance. It has to be transferred to the state.
So even if one accepts that equality of outcomes is not the objective, as most American do (an attitude with which I have sympathy), there is a strong case for action to equalise educational opportunity and provide universal health insurance. And this might also help reduce the current protectionist attitudes of the left. The Scandinavian, Dutch and British left is more enlightened on trade than the American one, precisely because there is a welfare state.
Posted by: FT Forum - Martin Wolf | February 16th, 2007 at 7:11 pm | Report this commentRobert Wade: I agree with Martin’s bottom line: “[W]ithout more generous government-financed services, the US may be unable to maintain a dynamic, internationally open and socially mobile society”. I also agree that it is not a paradox. It is a paradox only in a simple kind of market economics.
One of the many weaknesses of simple market economics is its equation of “income inequality” with “incentives”, and its fudging of the distinction between “productive” and “unproductive” income differentials. Unproductive ones result from monopoly power, or power-biased organizational arrangements (such as “scratch my back and I’ll scratch yours” in board decisions about CEO remuneration), or categorical discrimination (against women or ethnic minorities, for example). They do not serve as incentives - they do not encourage individuals to adjust their behaviour in response to differences in reward. Productive ones do.
Once the distinction is made between income differentials which incentivize and those which do not, one can understand the absence of an association between measures of national “inequality” and economic growth. Higher 90-10 wage inequality, for example, is not associated with subsequent higher economic growth. Our measures of inequality obscure the distinction and thereby fail to capture the structure of incentives.
Certainly the economic performance of the Nordic countries since 1994 suggests that more egalitarian countries with higher social spending, strong unions and compressed wage differentials can have as good an incentive structure and grow as fast as Anglo “liberal market economies”. A whole raft of public policies - including active labour market policies (retraining schemes, facilitation of women’s employment while bringing up children, and special concentration on raising educational achievement at the lower levels of the skills hierarchy) - helps to reduce the size of the income differentials needed to induce changes of behaviour in a more productive, less socially costly direction.
This connects to Willem’s declaration that absolute poverty bothers him, but not relative poverty, and emphatically not income/wealth inequality. Brazilian levels of inequality - no problem; US 90-10 and 0.01-99.09 inequality soaring over the past 20 years or so - no problem, unless it generates more absolute poverty.
He overlooks quite a lot of evidence on the social costs of inequality, on the link between inequality and the quality of life. Higher income inequality within countries goes with: (1) a higher poverty headcount and a lower contribution of economic growth to poverty reduction (lower ‘growth elasticity of poverty reduction’); (2) higher unemployment; (3) higher crime; (4) lower average health; (5) weaker property rights; (6) more skewed access to public services and state rule-setting fora, and lower standards of public services; (7) lower social mobility, and (8) slower transitions to democratic regimes, more fragile democracies, more social unrest. (Some of this evidence is gathered in the World Bank’s World Development Report 2006, Washington DC, September 2005; and in UNDP’S Human Development Report 2005, New York, 2005. Also the essays in Global Inequality, eds. David Held and Ayse Kaya, Polity, 2007, including my “Should we worry about income inequality?”)
But in the “conservative” world view, such evidence tends to be waved away by the conviction that whatever inequalities of rewards which emerge from market competition are justified - provided there is some equalization of opportunities for health care and for primary and secondary education, and provided labour markets are not distorted by union bargaining. Efforts to curb them (eg by progressive taxation) not only cause inefficiencies but also undermine the competitive discipline which is the very foundation of the moral society. Unrestrained inequality is simply the price to be paid for the ultimate good, a disciplined, moral society; though some restraint on inequality may be pragmatically justified in order to avoid even more morality-eroding policies like “protectionism”. I presume this is where Willem is coming from, and as Martin says, he agrees with Willem most of the time.
I wonder how adherents to the conservative world view interpret the new UNICEF study of 21 developed countries as places for children to grow up in. I haven’t done the calculation, but I conjecture that it shows a roughly inverse relationship between 90-10 type measures of after-tax income/wealth inequality and the conditions of children. Certainly the Netherlands and the Nordics are up near the top (except for Iceland, 15th, which has the most Anglo political economy of all the Nordics); the UK is at the very bottom. Willem’s lack of concern about relative poverty sits oddly with the study’s evidence on the link between the proportion of children living in relative poverty and a whole range of other indicators. Might the relatively very high proportion of British children growing up in relative poverty not be linked to the relatively very low proportion of British children saying that they generally found their friends and classmates “kind or helpful”?
Which brings me back to the sheer grudging nature of Martin’s concession that America does need some (modest) elements of a welfare state. Precisely what sort of damage does he envisage being done to “incentives” by doing more than he suggests - including raising minimum wages, which not he but only “the left” wants? Presumably he is thinking not only of incentives to escape from poverty for the poor, but also incentives to earn several more million for the rich?
Lastly, following on from Brad’s comments on Martin’s uncritical use of Alan Reynold’s data on US income distribution to cast doubt on the reality of big gains at the top of the US income distribution, a comment on another subject where more critical use of data is necessary. I have been concerned at Martin’s uncritical use of Sala-I-Matin’s data on world income distribution. In Why Globalization Works, p.153, and in several columns he presents the study as providing serious evidence that world income inequality has fallen sharply since 1978, thanks to “globalization”. Sala-I-Matin’s methods have been taken apart by, among others, Branko Milanovic and Sanjay Reddy (though they have not had access to his full data and programs, which at least as of mid November 2006 he had not released, not even to the journal which published the paper, Quarterly J. of Economics).
Posted by: Robert Wade | February 18th, 2007 at 10:05 am | Report this commentWillem Buiter: Robert Wade’s comments on my statement that absolute poverty bothers me but not inequality are a hodgepodge of conceptual confusions, non-sequiturs and misinterpretations of what are (at best) wonky correlations as causal links.
He starts with the statement that I overlook quite a lot of evidence on the social costs of inequality and on the link between inequality and the quality of life. My statement overlooked nothing at all. It was a statement about my tastes, my preferences. Poverty is bad. Inequality per se is neither good nor bad. I simply don’t care. So the massive and extreme poverty in Brazil bothers me, but inequality in wealth and income distribution in Brazil per se does not bother me. Ditto for the USA. If greater inequality (whatever caused it) were to cause greater absolute poverty, that would, of course, bother me, because poverty bothers me. Clear thinking is helped by an awareness of the distinction between normative and positive statements.
Then comes a long list of things that inequality within countries “goes with” (Robert Wade’s words): higher poverty headcount, lower contribution of growth to poverty reduction, higher unemployment and crime, lower average health, weaker property rights, more unequal access to public services, lower standards of public services, lower social mobility, slower transitions to democracy, more social unrest.
I think it is, on balance, a good thing that so many social scientists have been gainfully employed correlating across time and space (sometimes even with panel data!) measures of inequality with every conceivable measure of disfunctional individual and group behaviour. Who knows what they might have got up to otherwise? Unfortunately, ‘goes with’ is just a couple of weasel words. It’s what you use when you want to imply causation but dare not use the word because there is no scientific basis for a causal interpretation of the association between the variables in question. “Goes with” indeed - in high school I used to ‘go with’ a girl named Suzanne. This kind of contempt for the reader gives the social sciences a bad name.
The statement that inequality does not bother me also does not mean that I don’t care about how wealth or income are acquired. Ill-gotten gains are to be fought everywhere, with due allowance for the costs of prevention, mitigation or cure. Ill-gotten gains are not just those obtained through criminal activity or the use of force. Robert rightly points to the abject failures of corporate governance (especially in the US) that have allowed CEOs to enrich themselves at the expense of their shareholders (and possibly of other stakeholders as well). It includes the abuse of monopoly power by trades unions, whether they be ASLEF, the Teamsters, the BMA or the American Bar Association. It includes the rents extracted from their customers by the integrated gas and power utilities in some continental European countries – rents which are shared by the managers and the workers of these utilities. Income differentials reflecting discrimination (whether legally based or reflecting social conventions and norms) should likewise be tackled at source, by eliminating the discrimination. The assertion that those who are not bothered by inequality equate inequality with incentives is an insulting nonsense. Liberals, conservatives and socialists would find themselves on the same side of the barricade on this issue.
America has a welfare state. It’s probably the worst-‘designed’ and worst-functioning welfare state among the industrial countries. Without any increase in general government spending as a share of GDP, the US could transform its disfunctional welfare state into one that does eradicate absolute poverty. And it can do so without creating the dirigiste, stultifying, anaemic societies that make up too much of the European continent.
Posted by: Willem Buiter, London School of Economics | February 19th, 2007 at 2:08 am | Report this commentMartin Wolf: For the moment, I am going to keep out of the Willem-Robert intra-LSE fight. So let me respond to Brad and one central point that Willem makes.
Willem says, surely rightly, that “America has a welfare state, but it’s probably the worst-”designed” and worst-functioning welfare state among the industrial countries.” This seems evident. To take just one example, what valid social (as opposed to political) reason is there for a welfare system (Social Security and Medicare) that is so overwhelmingly biased towards helping the old rather than the vulnerable people who really matter for the future of a society: families with children? If one is going to provide expensive health insurance for old people, why, as matter of course, doesn’t one provide it for children?
Brad’s explanation is that American politics are “really weird”. Lots of sensible deals just don’t get done. This is a correct description. But it is hardly an explanation. So the question is: why?
Is it the sheer scale of the country and the resulting difficulty in forming broad coalitions (Mancur Olson’s famous point about “encompassing organisations”) and so a tendency for zero- or negative-sum political outcomes?
Is it the constitutional setup, with federalism, the division of powers and weak (domestic) government?
Is it the strength of special interest lobbies that result from the country’s scale and its constitutional set up?
Is it the gerrymandering of congressional seats and the consequent incentive to reach for the extremes in the primaries that are, in most cases, the “real” election?
Is it the intense ideological rifts and, if so, what caused them when other high-income countries have been moving away from deep (left-right) political divisions over the past two or three decades?
Is it a cultural divide, which is essentially religious/secular?
If, as I fear, the Democrats are going to go protectionist because it is the politically acceptable way to (appear to) help its vulnerable constituency and if, as I expect, the Republicans are now (rightly) discredited, because of the frightening incompetence (and worse) of the Bush administration, we have a big problem on our hands. Like many non-Americans, I very much want to see an end to the administration’s politics of paranoia, but not to see it replaced by the politics of protection.
Posted by: FT Forum - Martin Wolf | February 19th, 2007 at 6:59 pm | Report this commentRobert Wade: Willem Buiter’s comments are a sad illustration of what happens when jeering contempt drives argument. In this case, jeering contempt for those not satisfied with his bland assumption that everyone can agree that fairness simply equals the absence of monopoly rents or power intervention in free markets; and who want to use their skill in regression analysis to find out what “goes with” inequality. “Goes with” is plain English for correlation, and one would think it could be taken for granted in an economists’ forum that correlation is without prejudice to the question of causation.
It would be interesting to know why Willem is bothered by absolute poverty (though not relative poverty or inequality); and, of course, how he defines and measures absolute poverty (PPP$ 1-a-day? 2-a-day?). Is he bothered by absolute poverty just for productivity reasons, in the spirit of the Virginia planter who said, in 1837, that the most important subject in slave management was sufficiency of food: “The master who gives his field hands half a pound of meat per day and two quarts of meal…is better compensated by slave labor, than those who give the ordinary quantity” (quoted in Alan Berg, The Nutrition Factor, Brookings, 1973) ? If there’s more to it than that, what else? Does Willem include the deprivation of citizen dignity which over the last ten years has been put into the “absolute” category of social exclusion, but is clearly a matter of dignity relative to that of other citizens?
I am bothered about absolute poverty, relative poverty and inequality, for both moral and instrumental reasons, and consider that any responsible government should be concerned to keep a lid on all of them. The recent UNICEF study of child well-being suggests one big reason to be concerned about what Willem says he is not concerned about, namely relative poverty. That Britain has the second highest percentage of children (0-17 years) living in households with equivalent income less than 50% of the median (next to the US at the unglorious top) is probably an important causal driver of British children’s bottom rank (21 out of 21) in child well-being. And likewise for the four Nordics, which have the four lowest percentages of children in relative poverty and occupy four of the top seven positions in child well-being. I take this evidence as part of a case for substantially stronger public action to reduce relative poverty in Britain.
Beneath our differences in reading evidence and style of conversation is a more basic difference in premise about the state. I tend to emphasize the enabling potential of the state, compensating for market failure. Willem, in the great mainstream of Anglo-trained economists, tends to emphasize the restrictive effects of the state, which induce market failure. This is the lens through which he sees “too much of the European continent” as occupied by “dirigist, stultifying, anaemic societies”, presumably in contrast to America, which is much closer to the mainstream model of virile capitalism (dysfunctional welfare state aside). Does “too much” include the Nordics, and Netherlands, Spain and Switzerland, which account for the top seven child well-being positions in the UNICEF study? In the context of developing countries, the mainstream’s emphasis on the “restrictive state” over the past several decades has been hugely damaging (for them, not so much for us in the rich countries), because it causes an allergy to anything resembling “industrial policy”, and so hinders state action in an area full of market failure, namely, innovation.
The differences in premise about the state are themselves part of a larger difference in worldview, of the kind suggested in my original comment. These differences (as well as the similarities) would certainly be worth exploring in a civil way.
Posted by: Robert Wade | February 26th, 2007 at 10:51 am | Report this commentMartin Wolf: I want to apologise for the long delay in my response to the various important comments from Willem and Robert on my column. I am closer to Willem. But I do not agree with him on everything. So here is my reaction to his proposals:
I agree, above all, that I cannot see why inequality itself is objectionable. Consider the following two alternatives: a society in which everybody’s real income is $100 and one in which 50 per cent have $100 and the other 50 per cent have $200. I find it difficult to imagine that the second society is worse than the first. It appears to me then that inequality in itself is not a bad. What matters is how it came about (through theft or productive activity and so whether it is an aspect of a dynamic economy or the product of a zero- or negative-sum civil conflict), whether it can be diminished at little or no cost and whether it does some further damage to what I would see as a good society.
Where I agree with Robert and disagree with Willem is that inequality does seem to me to inflict some such damage. One way it does so, I suggest, is via its impact on equality of opportunity. There is much evidence, for example, that, other things being equal, it is helpful to children to be brought up in a family with two married parents in a stable relationship. Where inequality of earnings among men is very high, however, a large proportion of men cease to be useful as husbands: they become less capable providers than a welfare state. This generates an explosion of single-parent families that produce men whose social role is that of casual worker, criminal, layabout or feckless “baby-fathers”. This is not a good environment in which to bring up children who will go on to thrive in the sort of sophisticated society we now have. I do not think these obstacles to equality of opportunity are just a matter of absolute poverty, as Willem suggests. They are also a matter of self-discipline, learned from suitable adult role models, particularly parents. I think many young people, particularly boys, now lack such models.
Another worrying consequence of extreme inequality, this time at the top, is enormous inequality in political power. We know that there is a huge collective action problem in democratic politics. Where people are immensely wealthy it will pay them to intervene directly in the political process in support of candidates they believe in or of their own narrow material interests (in low taxes, for example). The result is a plutocracy, not a democracy. This is not the sort of society in which I particularly wish to live.
This does not mean that inequality can be attacked at no cost. Nor does it mean that this should be the overwhelming priority of economic or social policy. But it is an important consideration. I also suggest there are policies that would do something to reduce these forms of inequality or their consequences. The problem of low wages for less skilled workers can be tackled through wage subsidies. If wage subsidies are attached to individual working adults rather than families, there will be an incentive to form stable partnerships. Of course, if society wishes to ensure that a woman with children is as well off, whether or not she is with their father, then this policy will fail. Equally, policies to restrict the cost of elections may deal with some of my concerns about the rise of plutocracy.
Willem is right when he notes that there is no point in bemoaning some problem if there do not exist relevant policies that can ameliorate it, without introducing additional costs. There are always trade-offs. So, as he notes, weakening trades unions may increase inequality but also reduce the duration of joblessness. In this case, the direction of the impact of this change on resistance to job losses (and so the pressure for protectionism) is ambiguous.
Where Willem and I clearly agree, however, is on the high priority for the generous public funding, rather than provision, of schooling up at least to the age of 18. We agree, too, on universal health insurance, though with greater private supply of health services than under the UK’s National Health Service. We agree, again, on the absurdity of linking pensions to employment at a particular company. The absurdity of this has become evident to everybody in the UK and US in recent years.
I share Willem’s desire for something like a negative income tax, provided it was linked to income from work, at least for the able-bodied. I don’t think large transfers to the idle able-bodied are tolerable. They break what I think most people view as an acceptable social compact. Transfers to the disabled, children, students and the old are another matter.
Posted by: FT Forum - Martin Wolf | March 1st, 2007 at 12:18 pm | Report this commentWillem Buiter: I would like make three points in response to Martin’s latest comment in the discussion of ‘inequality’ prompted by his column on the welfare state in America.
The first relates to Martin’s argument about ‘baby-fathers’. Both the moral and the public policy aspects of the issue are clear. The responsibility of parents for their offspring ought to be inescapable. Like the ‘you break it, you own it’ sign found in china shops, signs stating ‘you contribute to its conception, you are responsible for it until the child reaches the age of majority’ should be prominent in any location where acts leading to procreation might be engaged in or considered. At the very least, this means financial responsibility for the well-being of the child (the cost of housing, food, clothing, education, health care, socialisation, fun and entertainment). It should go beyond that to include financial responsibility for the transgressions of the minor child. Equally important is the moral duty to spend time with your child and to support it emotionally as well as in practical ways.
It is not clear, however, what any of this has got to do with relative poverty or inequality. Even within the UK and the USA, there are ethnic communities with similar levels of financial resources but with substantial differences in the incidence of single-parent families, absentee fathers and neglected children. Inequality (or even absolute poverty) are not causally related to self-discipline or a sense of responsibility in any simple and obvious manner. What is the mechanism linking inequality to the excessive consumerism, irresponsibility, hedonism and individualism-verging-on-egotism that are the ugly face of modern western societies?
I also agree that just as parents are responsible for their children, communities are responsible for the weak, the vulnerable and the helpless among them. That includes the young, the old, the sick (including the mentally ill) and mentally handicapped and those who have for other reasons fallen on hard times through no fault of their own. But again, what has this got to do with inequality?
The second point relates to equality of opportunity and the extent to which inequality of outcomes limits equality of opportunity (for the next generation especially). I am a firm believer in equality of opportunity, but recognise that the concept has to be defined carefully, because it is quite slippery and can be twisted into something very close to equality of outcomes (through the ‘equal opportunity to achieve a given outcome’ interpretation). Equality of opportunity clearly means absence of discrimination and of other, artificial, man-made obstacles to individual achievement. The interesting question is to what extent, and for how long, the community (through voluntary mechanisms, through the state or through other collective mechanisms) should try to compensate for unequal endowments, broadly defined to include not only one’s biological/genetic inheritance, but also the family, social and financial resources one inherits. Here I favour an end to ‘collective compensatory interventions’ (both interventions in the form of special assistance to those with low endowments and, a-fortiori, those leveling interventions that involve handicapping those with high endowments) when a person becomes a competent adult (and for as long as (s)he remains a competent adult). The nature of the support for the young is the familiar list of nutrition, health, education and socialisation. To take the procrustian equalisation of outcomes for competent adults further than this, is to enter a dangerous road whose ultimate destinations are fascism and communism.
The third point concerns the risk that, in a society with extreme inequality of income and wealth, enormous inequality in political power will result. It is a valid concern. Both in the US and in the UK, there are worrying signs of plutocratic excess undermining democracy, liberty and the rule of law. In the US, when rich criminals have a significantly better chance of beating the rap than poor criminals committing the same felony, equality under the law is made a mockery of. The dominant role in British politics of the Murdoch press is a clear example of the unacceptable face of plutocracy. If anything, including extreme wealth in a highly unequal society, were to infringe on the unalienable Rights of others (among these are Life, Liberty, and the pursuit of Happiness) action would be required. The action should be targeted directly at the problem: the impact of extreme inequality of income and wealth on the inequality of political power. It is not obvious that the best, let alone the only way to address this problem is by measures aimed at reducing inequality; instead, legislation and regulation aimed at preventing the abuse of private wealth to undermine democracy and liberty may be more effective in achieving the desired result. Transparency and openness about the use of private wealth in political activity would be an important first step. If honours are to be sold, let them be sold in an open, competitive and transparent market! Limits on the concentration of ownership and control of the media are another. Tax payer financing of political parties a third. It may be that deliberate measures to limit the inequality of wealth and income (such as more progressive taxation and more severe inheritance taxes) will be required to prevent undue concentration of political power in the hands of a few very rich people, but the evidence to support this position is yet to be presented.
Posted by: Willem Buiter, London School of Economics | March 3rd, 2007 at 5:19 pm | Report this commentMartin Wolf: Let me comment on Robert’s responses and apologise for my delay in doing so.
For the reasons I have given in my discussion of Willem’s views, I agree with Robert that inequality has real costs, but disagree with him on the ease with which it can be eliminated. Consider, for example, the contrast he draws between the US and Nordic countries, which are the extreme examples among the high-income countries. Before accepting that the latter are an acceptable role model for the former, it is surely necessary to take a few obvious points into account.
The US is a continental economy of more than 300m people. None of the Nordics would be more than a small state in the US. The US is both ethnically and culturally highly diverse and has a population built out of successive waves of very different immigrants. The Nordic countries are (still) extremely homogeneous. Yet, despite the enormous differences among the immigrant communities, the US has the highest GDP per head of any country in the world, at purchasing power parity. I would also imagine that the income per head of people of Scandinavian stock in the US is at least double the average for Scandinavians in Scandinavia. This really is an astonishingly successful economic system, given the inflow of immigrants from poor countries.
Finally and surely most importantly, it is inconceivable that a state and society similar to that of Sweden or Denmark could be created and sustained in the US, given its history, constitutional traditions and so forth. We must accept differences among countries provided they are reasonably functional. In fact, we must often accept differences even when countries are dysfunctional (as far too many now are), because nothing can be done about them.
As to the UNICEF study of children that Robert cites, I admit not to have studied it. But I am under the strong impression from the analyses I have read that it adopts methods designed to find exactly what it does. This is particularly true of the use of the term “relative poverty”. Since we already have the perfectly good word “inequality”, the phrase “relative poverty” seems to me to add just about nothing. Use of it is an intellectual trick designed to persuade people who care about poverty that inequality is the same thing.
The phrase also leads to peculiar results. Consider two societies with very different average incomes, neither of which contains many people who would be regarded as absolutely poor by global standards. Assume too that the richer society is considerably more unequal than the poorer ones. Then it is easy to conclude that it contains more poor people than the poorer society, even though a large proportion of the people living in the latter society are absolutely worse off than anybody living in the former. Surely, the right thing to say, in such a case, is that the richer society has fewer poor people than the poor one, but is also more unequal.
Robert chides me for being grudging in my admission that the US needs a welfare state. But, as I have already said above (and he himself insists in other contexts), one must take societies largely as they are. It is not for me to tell Americans that they really want to live in Sweden. I am far less convinced, in addition, of the beneficial consequences of high minimum wages. So many countries have destroyed their labour markets in this way.
What I think an outsider can do is point out the internal inconsistencies, as he or she sees them, in positions held by people of influence in a given society, such as Ben Bernanke. My aim was to show that even someone with his values should recognise the need for a more generous welfare state. In other words, one can analyse the choices people confront even if their starting point is not one’s own.
On the data questions Robert raised, it is in the nature of column-writing that one cannot examine all data sources and analyses with equal care. That is why I appreciate Brad’s comments. But it is close to inconceivable that global inequality among households has not fallen over the past quarter century, given the relatively rapid rising incomes of China and India. With the greatest respect, I did examine the work of Milanovic and Reddy when I wrote my book on globalisation and found it quite unpersuasive. The former, in particular, published a well-known paper that not only relied exclusively on household income and consumption surveys, despite the inconsistency of these sources with national accounts (which also automatically cross-check among different data sources), but focused on a short time-period that “happened” to include the Tienanmen Square crisis in China and the balance of payments crisis of 1991 in India.
Is this, then, objective science? Robert likes data showing worsening global inequality in the age of globalisation. Let us simply agree that people are tempted to use data and methods that give them the conclusion they want. This is all too human. Since I, too, am human, I am no doubt prey to this fault, but no more, may I humbly suggest, than those on the anti-capitalist left.
Robert is right that the big differences are in attitudes to the state. Classical liberals mistrust the active state. The left embraces it. The big question here is whether, when one looks at the world as it is, as opposed to what would like it to be, one finds that states are generally beneficent agencies that use their powers wisely and with restraint. Actually, even Robert does not believe that, as his views of the present US administration makes clear. Let me grant that some interventionist states have been quite effective. Is that then a general rule? Is Japan a credible model for the developing world? Is Sweden a credible model for a continental super-power? I would suggest that the answers have to be no. Look at developing countries as they really are. First let these states fulfil their basic functions - honest and effective administration of fundamental services and provision of basic infrastructure - before we encourage them all to do industrial policy.
Posted by: FT Forum - Martin Wolf | March 4th, 2007 at 6:22 pm | Report this commentLet me make two comments on what Willem has written most recently.
First, while I agree that different ethnic groups show very different behaviour with regard to marriage and the family, I would also argue that the move towards single-parent families, with the mother “married to the state” is neither an accident nor the consequence of shifts in values that have nothing to do with economic forces. Marriage is, among other things, an economic arrangement, one in which spouses pool resources to rear children. If men have little earning power, particularly relative to what the state provides to lone mothers, they are not worth marrying. As men become aware of this, their behaviour will reflect it. If boys are then increasingly brought up without a father, they will not know how to be one. So I would suggest that rising inequality of earnings among men is one explanation for the growth in single-parent families, with the longer-run consequences for child-rearing and so equality of opportunity that I have mentioned.
Second, we agree on the dangers to democracy from extreme inequality in incomes and in economic power. I agree with Willem on funding of parties and on the need to limit control of the media by a limited number of tycoons. The position of Rupert Murdoch - not even a citizen of the country - in the political life of the UK is inconsistent with any democratic theory of which I am aware. But where inequality is big enough, politicians may find it too politically dangerous to make the targeted policy changes Willem (and I) would support. I don’t think that’s yet true in the UK. But it may be true in the US.
So we end with some disagreements and some agreement. But I have greatly enjoyed these debates.
Posted by: FT Forum - Martin Wolf | March 5th, 2007 at 10:58 pm | Report this comment