(March 25, 2016) Three things are striking about the rise in economic inequality since the 1970s in the United States. (1) It doesn’t really matter which measure of income distribution we choose; they all show a rise in inequality. (2) There are many different competing possible explanations and interpretations. (3) We do not need to agree on the explanation to agree on what are sensible policies to lessen inequality.
(1) Measures of inequality
There are many different measures of inequality. Even when just measuring inequality of income, one might look at the Gini coefficient, the poverty rate, income going to the upper 1% or ¼ %, the high-low range, the wage share, or median income. Also of interest are measures of the distribution of wealth and measures of inter-generational mobility. They can give different answers. If the topic is global inequality, for example, the historic economic success of many Asian countries has generated a reduction of inequality by some measures (e.g., a big fall in the poverty rate) but not by others (an increase in the high-low range).
In the United States, however, all measures have pointed the same direction since the turn of the century. They all reflect that the benefits of economic growth have gone to those at the top. The share of income going to the top 1/2 of 1 percent, for example, is up to 14%, where it was in the 1920s.
(2) Diagnoses
Normally one would think that diagnosing the cause for such a fundamental shift would be a necessary step in prescribing a cure. In that case one might be discouraged by the over-abundance of plausible explanations that have been offered and the difficulty in choosing among them.
Thomas Piketty’s Capital in the Twenty-first Century emphasized what he saw as a very long term trend arising from a high return to capital which causes inherited wealth to accumulate at a faster rate than earned income grows. The 2013 book did much to restore priority to the subject of inequality on the agenda of economists. But research suggests that widening US inequality comes primarily from within earned income, rather than arising from the difference between earned income (wages and salaries) and unearned income (return on capital). According to the Piketty-Saez figures, of the substantial 1970-2010 increase in the share of income going to the upper 1% of Americans, 68 percent is due to increased inequality within labor income. (32 % is due to increased inequality within capital income, and 0 % to a shift in income from labor to capital.)
Several causes of increased inequality of earned income have been identified.
- The first of the explanations is technological change (such as information technology) that raises the demand for skilled workers faster than the supply. It can explain an observed widening gap in wages or incomes between skilled and “unskilled” workers, defined according to whether they are college-educated. But this has little to do with the gap between the upper 1% and the rest.
- The second explanation is “assortative mating,” according to which highly accomplished professional men no longer marry their secretaries but instead marry highly accomplished professional women.
- The third is the “winner take all” aspects of many professions, from dentists to university professors to movie stars. Because modern media tell us who is the best dentist in town or the best movie actor in the world, relatively small differences in abilities win far bigger differences in income than they used to.
- According to the fourth explanation the very high compensation of corporate executives, especially in the financial sector, is not a return to services that are in demand because they are socially valuable (like having gone to medical school or having been born with acting talent). Rather, managers essentially get to set the terms of their own pay, through compensation packages that reflect failures of corporate governance, tax law, and financial engineering. Options, for example, have failed in their original goal of relating pay to performance.
- A political economy explanation is especially popular: the rich have captured the levers of power, through campaign donations, and so are able to get policies adopted that are favorable to them.
(3) Must the diagnosis determine the treatment?
It sounds reasonable that to address a problem one first has to figure out what caused it and then undo that cause. But this is not always the right approach. One does not necessarily have to figure out why a particular physical injury occurred to determine the best medical procedure for treating it.
Consider, for example, another explanation for inequality that is particularly popular in this election year: trade. To avoid the contentious question of whether trade has helped or hurt lower-income American workers generally, let’s focus on one sector where the claim that trade has hurt the employees is almost certainly true: the auto industry. 50 years ago workers who lacked a college education but were lucky enough to get a job in the auto industry could earn high wages — much higher than other US workers, let alone those abroad. But Detroit back then had a near-monopoly, which allowed it to produce cars that by modern standards cost a lot, got poor gas mileage, and broke down often. A subsequent flood of auto imports from Japan and elsewhere ended all that. One result was that the US auto industry adjusted and today is globally competitive. Another result was that the wages of auto workers were badly hit. (Notice that these were the effects of trade, not of regional trade deals like NAFTA.)
Even if one cares only about the auto workers’ wage losses and believes that trade is the sole reason for it, what remedy does one arrive at? To protect Detroit fully against foreign competition, the government would have had to raise tariffs on auto imports astronomically, so that domestically produced cars today would cost several times more than foreign produced cars. Does anyone seriously think that is a practical policy solution?
(4) Policy prescriptions
Many of the same policy prescriptions to ameliorate inequality apply regardless what is the cause. Most of them are ways to make the tax system more progressive. This includes lowering the effective marginal tax rate for low-income workers – what President Obama called “Making Work pay” (in 2009, when he passed a refundable tax credit for low income workers, until Republicans took back the Congress and eliminated it]. Enhancing the earned-income tax credit is a live option today. Obama also proposed in January’s State of the Union message expanding wage insurance, which currently helps workers who lose their jobs because of trade but could be made to help as well those who lose their jobs due to technological change. Another possible proposal would eliminate the payroll tax for low-income workers.
In light of looming deficits in social security and Medicare, it would probably be good to make such changes in the tax code revenue-neutral. (Ten years ago, reducing the deficit should have been a clear priority; five years ago, stimulating the economy should have been the short-term priority. Today there is not so strong a presumption either way.) The Republican candidates for president this year, as usual, propose massive tax cuts, focused on the rich, without a plan how to pay for them.
The lost revenue could be balanced with some measures from the following list:
- End the tax break for “carried interest” — by taxing it at regular income tax rates, rather than at low capital gains tax rates (which is a pure a gift to hedge fund managers) — as Hillary Clinton wants to do.
- Reduce the estate tax exemption (originally expanded by George W. Bush].
- Raise the cap on payroll taxes for upper-income workers.
- Remove oil subsidies and use the gas tax to fund the federal highway trust fund.
- Tighten some distorting tax deductions like mortgage interest for upper income households.
A lot should be done on the spending side too. Examples include universal high-quality pre-school, health care insurance for all, and infrastructure spending. Many of these measures have been pursued by President Obama.
It is striking how little this list of policy prescriptions depends on the precise diagnosis of the problem.
Many of those who are upset about inequality — perhaps lacking patience for the minutiae of fiscal policy — are attracted to the banner of Bernie Sanders (or the banner of Donald Trump). They like the argument that one must break up banks in order to address the root cause of the problem, which is thought to be that the rich use campaign contributions to influence politicians.
Money in politics is indeed a big problem. But what do politicians use all that money for? It doesn’t go into their pockets (at least not in the US). Rather the money goes to running for office: campaign advertisements and getting out the vote. In my view, voting for the right candidates (by which I mean those who will enact the right policies) is a far more direct strategy than the convoluted Rube-Goldberg chain of causality which says one should vote for the people who want to break up the banks in order to reduce the amount of money in politics so that there will be fewer ads trying to dissuade one from voting for the right candidates.
[A shorter version of this column appeared at Project Syndicate. Comments can be posted there or at the Econbrowser site.]