April 11, 2015
Policies to shift income from rich to poor may prove less effective than imagined
In a world of inequality the egalitarian thief is an attractive character. From England’s Robin Hood to America’s Jesse James and Mexico’s Pancho Villa many countries lionise brave outlaws who take from the rich and give to poor. Economics, in its down-to-earth way, seems to support their cause: since the cash-laden tend to save more, diverting income to penniless people who will spend it immediately should boost consumption, and GDP. The electoral calculus of redistribution appears favourable, too. The best-off are small in number, so taxing their mansions seems likely to win more votes than it loses. But those who fancy taxing the wealthy should tread carefully. Designed badly, such policies could do more harm than good.
The idea that redistribution could help spur growth has long attracted adherents. In 1920 Arthur Cecil Pigou argued that an annual transfer of resources from the “relatively rich to the relatively poor” would increase national output. Pigou discussed three uses for income: consumption or investment by the rich, and consumption by the poor. Shifting purchasing power to the poor would do little to hurt rich folks’ spending. The outcome of soaking the rich, decreased investment, would be outweighed by purchases of better food, clothing and education by the poor. Thus redistribution would boost output.
Pigou’s argument rests on the idea that poor families would spend more if they had the means, and that the wealthy would be able to smooth consumption if they suddenly lost income. To investigate whether this assumption holds, Greg Kaplan and Justin Weidner of Princeton University and Giovanni Violante of New York University used huge microeconomic datasets to paint a picture of household income and wealth across eight advanced economies. For each household they totted up income from salaries, public handouts and private transfers such as alimony payments. They also measured liquid wealth: cash in bank accounts, along with bonds and stocks that are directly held and so could be quickly sold. The researchers were looking for families that lacked a buffer of liquid assets (or credit facilities) to offset short-term changes in income. This group, whose consumption has to adjust as income changes, are those that live “hand to mouth” and would be likely to spend most from a government windfall.
But the term “hand to mouth” is not as straightforward as it seems. The data show that the median American holds some liquid wealth in bank deposits, as well as illiquid wealth (retirement accounts and houses, net of mortgage debt), but hardly any shares or bonds. Surprisingly, although around 30% of households live from payslip to payslip, two-thirds of these cash-poor people have sizeable illiquid wealth. They do not fit neatly into the Robin Hood bifurcation between rich and poor: their cash barely covers outgoings but they sit on large illiquid assets.
Housing debt is one reason people end up short of cash. Focusing in on American homeowners, the researchers find that of those with small mortgages, only 20% live hand to mouth. But once total debt approaches the value of the house, a much higher number—close to 50%—are income-constrained. Age is also a factor. Although the likelihood of genuine poverty tends to fall with age as workers build up buffers, the chances of being wealthy but cash-strapped peak around the age of 40.
These findings matter because cash shortfalls affect behaviour. Using another dataset that tracks 5,000 American households, the researchers measure the reaction to short-term income shocks between 1999 and 2011. The results confirm Pigou’s hunch. Those with lots of liquid wealth spend just 13% of an unexpected windfall; those living hand to mouth spend 24%. The wealthy-but-income-constrained react most, spending 30% of any windfall, suggesting they are even more cash-strapped. That chimes with a study by James Cloyne of the Bank of England and Paolo Surico of London Business School, which found that Britons with large housing debts react sharply when taxes are raised or cut. In other words, taxing those with large but illiquid assets could cause more of a fall in spending than previously expected.
If policymakers need to draw more nuanced distinctions between rich and poor, they also ought not to assume that hard-up citizens will support redistribution. In a working paper, Vivekinan Ashok and Ebonya Washington of Yale University and Ilyana Kuziemko of Princeton University explain that support for redistribution should, in theory, rise the more a worker’s earnings fall short of a country’s mean income. Yet American attitudes have shown the opposite pattern: support for redistribution has remained flat or fallen as inequality has risen.
Much of this is down to age. Those below 40 follow the expected template: support for redistribution rises in line with inequality. The over-65s are different, perhaps because there are fewer in the “cash-poor” category. In the 1970s, when surveys began, they were more supportive of redistribution than the rest of the population. By the mid-2000s they were much less in favour, doubtless fearing that help for the poor would cut health benefits.
Those twiddling the fiscal dials should mull on these findings. They suggest that the benefits of a fiscal stimulus package would be lower if targeted on the basis of income: short-term largesse should be used on the wealthy, too. It also means redistribution from rich to poor may not be a one-way bet: in particular taxes on the wealthiest should be phased in slowly so they can liquidate assets rather than cut spending. And politicians betting on their Robin Hood credentials should be wary of greying voters. They may be more inclined to back the Sheriff of Nottingham.