The opponents of social spending have been routed from the political battlefield by the coronavirus crisis of 2020-2021. The United States hurriedly doled out more than 10 per cent of a peak year’s GDP in the Cares Act of March 2020 and added almost as much under the American Rescue Plan of March 2021, in the form of unemployment compensation and other direct transfers to workers and the poor. Britain’s Conservative government ramped up public spending on health care and other social services. The conservative Christian Democratic coalition of Germany and other European Union governments did likewise. They and the European Central Bank even exempted Greece, Italy, and other heavy EU debtors from the usual belt-tightening strictures. Financial markets smiled through it all, and the losses of jobs and output have been reversed.
In fact, even before the COVID crisis, the world’s historical experience had already delivered a verdict against the common claim that social spending drags down the growth and level of GDP. My new book re-affirms a “free-lunch puzzle”: larger social-spending budgets have not accompanied any net loss of GDP, or in skills, or in work. So say the experiences of over 20 countries over the last 14 decades. Testing the effects of the size of total social budgets means comparing whole national bundles of social policies to see how they correlate with economic outcomes. There is always wisdom in looking first at the whole forest, before approaching any trees. Even if one tries to control for other factors, one still finds no clear negative effects. Without any such costs, Europe’s welfare states have quietly produced greater equality, cleaner government, and even longer life.
How can that be? How can taking a quarter of national income in taxes and spending it on social programs do no net damage to GDP, work, and skills? Wasn’t there merit in the suspicion that transferring resources to people who need it would dampen work incentives, both for the recipients and for the taxpayers?
There are good economic explanations for this free-lunch puzzle. Only when focusing narrowly on unemployment compensation and certain specific welfare programs have economists uncovered negative work effects in the real world. Yet these are offset, or outweighed, by the more clearly positive parts of the social-spending bundle. Two quick examples should make the point easily enough: Tax-based state schooling, which the whole world has adopted, has clearly raised skills and productivity; and public health expenditures have also delivered longer and more productive lives.
The lack of any significant negative correlation between social spending’s share of GDP and the level or growth of GDP is all the more remarkable since short-run gyrations in GDP should cause a false bias toward a negative correlation. To see this bias, imagine a short-run slump in GDP, as in a recession or depression. The slump will cut the GDP denominator. At the same time, the slump should raise the social spending numerator by raising such ‘automatic stabiliser’ social spending as unemployment compensation and assistance to poor families. Result: a negative shock to GDP should show that the economy is doing worse at the same time that social spending is rising as a share of GDP. Having GDP shocks automatically trigger increases in the share of GDP devoted to social spending should show a negative correlation between social spending and the level (or growth) of GDP, inviting the false inference that the rise in the social-spending share lowered GDP.
Watch for this deceiving negative bias in writings about the year 2020, when the coronavirus crisis slashed GDP and caused governments to hike social spending. And yet, in 2020, as already hinted, it was the coronavirus that made a whole host of rich nations do the high-jump “Fosbury Flop:” Their outpouring of emergency aid suddenly lifted them over the bar into what might be considered welfare-state status, spending more than a fifth of GDP on social programs.
Why was the response to the 2020 crisis so different from earlier refusals to help those in need? The 2020 crisis, like the Great Depression of the 1930s, convinced those with political voice that anybody could suffer in such times – “that could be me.” It probably will, like the Great Depression of the 1930s, weaken the resistance to a more permanent and universal government social insurance. Even after the emergency has passed, continuing much of the new government aid may prove politically popular. It may prove impossible to squeeze the genie of larger social spending back into the bottle. Fortunately, the larger genie need not harm economic growth.
When the evidence keeps stacking up on one side of the scales, there comes a time when one side should concede. That time should have come earlier, but at least the crisis of 2020-2021 should finally bury the glib assumption that social spending is bad for economic growth in the real world.
Peter H. Lindert
Peter H. Lindert is Distinguished Professor of Economics at the University of California, Davis.
This article was first published in LSE Business Review
Stay updated with all the insights.
Navigate news, 1 email day.
Subscribe to Qrius