Which would you rather have: a 10 percent return on your investment or a 20 percent return? Okay, maybe that's a no-brainer.

But what if, in the first case, inflation was running at a Japan-like one percent per year while in the second it looked more like Iran, above 15 percent? In real terms, the first option now looks much better than the second.

This process of adjusting for inflation is common to almost all types of economic and financial analyses. But one area where it’s not used systematically is in calculating income inequality. Superficially, it’s understandable why this is the case: People who live in the same country face the same prices for goods and services no matter what they earning. An American consumer pays -- roughly -- the same amount for a Cadillac whether his income is $12,000, $120,000 or $1.2 million.

What’s missing from this line of thought, however, is that people in different income groups don’t buy the same things. A person earning $12,000 is a whole lot less likely to purchase a Cadillac than the other two earners. And if changes in the cost of living are unequal across income groups, then our measures of inequality could be misleading.

Using data on household consumption of a wide variety of products, two University of Chicago economists, Christian Broda and John Romalis, argue that thanks to cheaper imports from countries like China and the growth of retailers like Target and Wal-Mart, different income groups do indeed face different levels of inflation, with those at the bottom seeing smaller rises in price levels.

And if you buy their story, then measures of income inequality also change quite substantially.

For example, between 1973 and 2006, incomes for those in the 10th percentile (that is, the people who make more than only 10 percent of the population) grew by just 13 percent while growth was much higher---41 percent--for those in the 90th percentile. But when taking into account differing inflation levels, Broda and Romalis calculate that real income growth for the 10th percentile rises to 47 percent while growth for the 90th percentile rises to 51 percent. One major caveat with their result is that it's based on purchases of non-durable goods (like food and clothing). If poorer people have been facing higher inflation when it comes to durable goods (which seems unlikely), then Broda and Romalis's results may be overly optimistic. Still, the pair write:

“We take this to mean that the poor, relative to their past selves, have been experiencing quite rapid progress. As a consequence, we believe that more thought may need to be placed on exactly what we care about when we study inequality, or indeed when we make any conceptually similar comparisons."

--Zubin Jelveh