Home Mail Articles Stats/current Supplements Subscriptions Links
The following article appeared in Left Business Observer #60, September 1993. It was written by Doug Henwood, editor and publisher. It retains its copyright and may not be reprinted or redistributed in any form - print, electronic, facsimile, anything - without the permission of LBO.
It's been a while since LBO reported on the Luxembourg Income Study (LIS). The LIS is both a library of international income and poverty data, massaged into rough comparability (incomparable measures are the bane of international comparisons) and a community of researchers who write about the data. A year's backlog of working papers recently arrived, and here are some highlights; for full references, write or call. Unless otherwise noted, income is income after taxes and transfer payments (welfare, unemployment benefits, etc.), and poverty is an income below half a country's median. (The median is the figure at the midpoint of the income distribution: half of all people have incomes above the median, and half below.) Most data is for the study's "s econd wave," from 1984-87, which supplements the first wave (1979-81). A third wave, circa 1990, is being developed now. This research should be kept in mind as Bill Clinton's plans "to end welfare as we know it" emerge.
Several recent studies are making up for the LIS's slow start at covering women's welfare.
* Single-mother families are worse off than two-parent families in all LIS countries, but their incomes vary widely, according to Yin-Ling Irene Wong, Irwin Garfinkel, and Sara McLanahan. In Sweden and Norway, single-mother families have incomes 85% those of two-parent families; Britain, France, and Germany, 74%; and Australia, Canada, and the U.S., 56%. (At 54%, the U.S. figure was the worst.) Unfortunately, this is early-1980s data, but the clustering probably hasn't changed much. Contrary to the usual idiotic caricatures, 71% of U.S. single mothers are in the paid labor force, the third-highest of the eight; Sweden and Norway outscore the U.S., with other countries lagging well behind. While U.S. single mothers earn wages roughly comparable to those ea rned in other LIS countries, most countries' transfer systems do a better job bringing them above poverty. For lone moms, then, work effort has little to do with economic well-being.
If you want single mothers to work for pay, you can either craft a savage welfare system like ours to force them to -- or offer subsidized day care, child allowances, and a political environment favoring equality between the sexes, which will allow even m ore women to work.
* A typology of welfare systems suggested by Gosta Esping-Andersen has been widely adopted by income connoisseurs. Esping-Andersen divided First World countries into three groups: social democratic countries, the most egalitarian, with income transfers av ailable to all regardless of family status, and with national policies promoting full employment and high wages, as well as gender equality through subsidized child care and affirmative action; corporatist states, which also have high wages and generous i ncome transfer systems, but which favor the family over the individual; and liberal states (in the classic sense of laissez-faire, not the Rush Limbaugh sense), where benefits are minimal and the market rules. Social democratic states include Sweden and the Netherlands; the corporatist, Germany and Italy; and liberal, Australia, Britain, Canada, and the U.S.
Do real lives reflect this? Work by Sara McLanahan, Lynne Capser, and Annemette Sorensen concludes, yeah, pretty well. Employment patterns follow expected patterns. Women are most likely to work in the social democratic countries (thanks to vigorous suppo rts), least likely in the corporatist ones (where women are supposed to be stay-at-home mothers), with those in the liberal countries falling in between (where work is a necessity but no support is offered). Single mothers in corporatist countries are far more likely to work than married ones. Overall, women's poverty rates are lowest in the social democratic countries, next-lowest in the corporatist ones, and highest in the liberal regimes (though Britain's poverty rate was lower, and Italy's higher, tha n might be expected). Contrary to the baying of meatheads, generous welfare systems don't subsidize fecundity; women in the liberal states show higher fertility rates than the more generous ones.
By the way, Clinton's welfare-reform schemes look like a phony corporatism: a pro-family moralizing that is corporatist in rhetoric, while remaining (classically) liberal in its funding.
A study of wage changes and their contribution to inequality by Peter Gottschalk shows that while wage inequality increased in most LIS countries, the impact on household income was greatest in the U.S. Cutbacks in the U.S. welfare state meant that the go vernment, on balance, worsened income distribution (compared to pre-tax and pre-transfer figures). In this, the U.S. was virtually alone. Even Thatcher's Britain did a better job of softening growing wage inequality than Reagan's America.
Work by Timothy Smeeding, Barbara B. Torrey, and Lee Rainwater comparing the welfare of the elderly in eight countries (Australia, Canada, France, Germany, Netherlands, Sweden, U.K., U.S.) shows the U.S. with the most unequal distribution of income among over-65s. "Newly elderly" married couples in the U.S. enjoy incomes the highest multiple of average incomes in the eight countries, but older single U.S. women are the worst off of all, with a poverty rate 3 to 40 times higher than their peers abroad. One reason is that U.S. retirement benefits are less generous than in the other seven; "in general," the authors say, "the smaller the role of public sources of income for the elderly the higher the level of income inequality." Further, "the United States do es the least adequate job of preventing poverty among the elderly." This is the system that's allegedly bankrupting us with its generosity. And despite the alleged profligacy of Medicare and Medicaid, the U.S. elderly poor spend over 20% of their income o n medical care.
Official poverty figures in the U.S. and elsewhere focus on the percentage of the population below a certain income, the poverty line. But such poverty rates don't tell how deprived the officially poor really are. In one country, the poor might be plentif ul, but enjoy incomes clustered just below the poverty line; in another, there might be fewer poor, but those unfortunates are very poor. A third country might have a bounty of very miserable people. So the poverty gap -- the percentage by which the incom es of the poor would have to be raised to bring them up to the poverty line -- is also an important measure of poverty, as are measures of income distribution among the poor. Some measure that combines all three aspects -- depth, width, and relative depri vation -- into a single index would be a useful standard of comparison.
Lea Achdut and Orit Kristal studied eight LIS countries (Canada, France, Germany, Israel, Netherlands, Sweden, the U.K., U.S.) to flesh out the forms of poverty prevailing among them. According to mid-1980s data, by all measures -- headcount, poverty gap, and three versions of a composite index -- the U.S. had the highest poverty rates. After that, standings are harder to read. Sweden has the lowest poverty rate, but the second-highest poverty gap (after the U.S.); Sweden doesn't have many poor, but its p oor are under deep water. Israel has the second-highest headcount (after the U.S.), but the second-lowest gap (after Germany). Combined measures, all versions agree, show Germany, the Netherlands, and Sweden to be low-poverty countries; France and the U.K ., mid-level; and Canada, Israel, and the U.S., high-poverty (all in that order). U.S. combined figs were considerably higher than Israel's.Other news on the income front.
In a study of wages in 17 countries, the Organisation for Economic Cooperation and Development (OECD, a Paris-based official think tank sponsored by 24 rich industrial countries) found "dispersion" widening in most, though not all. (Dispersion, the OECD says, is a better word than inequality, since it avoids "implicit value-judgements." Whew!) The U.S. started the 1980s with the most unequal distribution of wages of the 17, yet managed, from these already high levels, to see the second-highest increase in dispersion, after Britain.
Several popular explanations are trotted out, which are refuted by the OECD's own data. It can't be that an influx of young people into the labor force depressed wages, since the OECD's index of the number of young people in the labor force shows their ra nks declining in eight countries during the 1980s, unchanged in two, and up in only three, and there by an average of just 5%. It can't be deindustrialization, either; within-industry inequality grew in most countries. Nor can it be a shortage of skilled labor, which might drive up high-end wages as employers bid up wages in the face of short supplies, since the number of college-educated workers increased in most countries. Demographic explanations don't fly either; inequality within demographic groups - - people of the same age and occupation, or same educational level, or family status, or work experience -- increased in several countries studied. The OECD wonders if these strange results aren't the result of "an increased importance of skills not well measured by educational qualifications, such as the ability to work with other people." Mystery solved! The reason for polarization: "does not work well with others."
The countries that didn't experience an increase in wage inequality during the 1980s -- Denmark, Finland, Germany, Italy, and Norway -- are ones with strong centralized wage-setting mechanisms. Formal and informal pay setting schemes were scrapped or weak ened in several countries that saw an increase in wage inequality -- Australia, Britain, Spain, and Sweden. Minimum wages eroded in others that showed more "dispersion" -- Canada, France, Portugal, Spain, and the U.S. Another culprit, the OECD barely conc edes, is the decline in unionization.
Wage setting was exposed more nakedly to market forces, and "dispersion" was the natural result.
A study by Rudy Fichtenbaum of Wright State University shows that the decline in unionization contributed significantly to the polarization of incomes in the U.S. This contradicts the conventional economic model, which asserts that unions increase inequal ity. Many economists do little more than add an overlay of math to the conventional notion that union workers are lazy, underfired, and overpaid. Conservatives with a sharp PR sense point to gaps that unions allegedly open among workers; unions, thundered right-wing icon Friederich Hayek in 1980 (when unions were considerably more powerful than now), "have become the biggest obstacle to raising the living standards of the working class as a whole. They are the chief cause of the unnecessarily big differen ces between the best and worst-paid workers." The organized extort fat paychecks, leaving everyone else in the dust.
Fichtenbaum adds to evidence that the right-wing view is fanciful nonsense. Unionization, several studies have shown, reduces the profitability of unionized firms, and since a significant part of upper class income comes from interest and dividends (i.e., profits), changes in unionization should affect income distribution. Unions, to be blunt about it, shift income from capital to labor -- which is why capital fights labor organizers tooth and nail.
As a what-if experiment, Fichtenbaum created statistical simulations of U.S. income distributions at two levels of unionization -- the actual level in 1950, and a fanciful 70% rate. (In 1950, roughly a third of the workforce was unionized; now, it's about a sixth, one of the lowest rates in the world. Sweden's rate is around 90%.) His estimates, which use the Gini index are reproduced nearby [not in web edition; sorry]. While they should be taken only as guides, not gospel, and the 70% union densit y figures taken with an extra dose of skepticism, the drift is clear: union decline has polarized American incomes, and higher unionization rates would de-polarize them. Returning to 1950s levels of unionization would bring the U.S. Gini up to that of Australia; it would require much higher unionization rates to get to the middle of page 4's chart.
PS An LBO examination of state-by-state unionization data, generously provided by Barry Hirsch and David Macpherson of Florida State University, shows a significant correlation between unionization rates and per capita income (with more unionized s tates showing higher incomes), though this association of course says nothing about distribution.
Home Mail Articles Stats/current Supplements Subscriptions Links