We’ve been measuring inequality wrong – here’s the real story

Income Inequality Thomas Nast Style Democracy Chronicles Flickr BY CC2.0

(The Conversation) Despite appearances to the contrary, this year’s presidential follies have managed to feature at least a few policy discussions amid all the name-calling.

Income inequality in particular has animated voters on both sides of the partisan divide, but the solutions advocated by candidates from each party are markedly different.

Democrats claim higher taxes on the rich and more benefits for the poor are the best ways to reduce inequality. Republicans argue what we really need is more growth, accomplished by lowering taxes to spur work and investment with, it seems, benefit cuts to make up lost revenue.

Remarkably, this debate has taken place based on partial and inappropriate indicators of U.S. inequality. Each party is dead certain about how to address inequality, yet neither knows what it is. Neither has a comprehensive and conceptually correct measure of inequality. The right measure is not how much wealth or income people have or receive but their spending power after the government has levied taxes on those resources and supplemented those resources with welfare and other benefits.

In a just-released study, we provide the first picture of actual U.S. inequality. We account for inequality in labor earnings and wealth, as Thomas Piketty and many others do. And we get to the bottom line: what does inequality in spending look like after accounting for government taxes and benefits?

Our findings dramatically alter the standard view of inequality and inform the debate on whether and how best to reduce it.

The methodology

Our study focuses on lifetime spending inequality because economic well being depends not just on what we spend this minute, hour, week or even year. It depends on what we can expect to spend through the rest of our lives.

Measuring lifetime spending inequality for a representative sample of U.S. households was a massive, multiyear undertaking, which may explain why ours is the first such study.

It required two big things. The first was developing software that properly measures lifetime spending, taking into account all possible survival scenarios households face (e.g., a husband dies in 22 years and a wife in 33 years). Second, it required accounting, in meticulous detail, for all the taxes households will pay and for all the benefits they will receive under each scenario. Our list included everything from personal income taxes (with its copious provisions) to estate taxes to Social Security benefits (eight kinds). Our paper lays out all the gory details.

The raw data came from the Federal Reserve’s 2013 Survey of Consumer Finances (SCF), which we ran through a computer program called The Fiscal Analyzer (TFA). We designed TFA to calculate the present value of the annual spending, including final bequests, a household can sustain given its “resources” (current wealth plus the present value of their projected future labor earnings), its taxes and benefits, and limits on its borrowing capacity. Our lifetime spending measure appropriately weights the spending arising under each survival scenario. The weights are the probabilities of the survival scenario in question and account for the fact that the rich live longer than the poor.

One final methodological point: since we are comparing lifetime spending inequality, it makes no sense to compare households of different ages, with very different lifespans. So we divided them up by age cohorts (30-39, 40-49, etc).

Next we ranked the households in each cohort according to the size of their resources, as defined above. Finally, we split the households into five equal groups or quintiles, with the lowest quintile having the lowest amount of resources and so on. We also considered households ranked in the top 5 percent and top 1 percent based on resources.



disinformation®­ curates the most shocking, unusual and quirkiest news articles, podcasts and videos on the web, most of which are submitted by the site’s visitors.