The New York Times
By: Robert H. Frank
April 2, 2011
LIKE everyone else, government officials want to look good. That often leads them to enact policies that promote favorable movements in the indexes by which they are judged. But when those indexes are imperfect, bad choices often result. And that's nowhere more evident than in economic policy.
By far the most visible official summary measure of individual economic well-being is per capita gross domestic product -- the annual market value of a country's final goods and services divided by its population.
The many shortcomings of per capita G.D.P. have been widely discussed. For instance, its failure to account adequately for product quality improvements in areas like computers causes it to understate well-being. And its inclusion of spending on burglar alarms and pollution-control equipment causes it to overstate well-being. The index also completely ignores the effects of changes in the distribution of income.
Although the last flaw is potentially far more serious, it has received little attention, because the costs of income inequality are notoriously hard to measure. Yet a simple supplementary index can be calculated from readily available data that captures one of the most important ways that changing distribution patterns have affected middle-income families.
I call it the toil index. It measures the number of hours that median earners must toil each month to be able to rent a house in a school district of at least average quality. (Could the median earner aspire to any less?) Unlike per capita G.D.P., which, apart from brief recessions, grew at a strong and steady rate from the end of World War II until the recent downturn, the toil index has been much more volatile. Its movements suggest that recent increases in income inequality have imposed substantial economic costs on middle-income families.
From 1950 to 1970, incomes grew rapidly and at about the same rate -- almost 3 percent annually, on average -- for families at all income levels. From 1970 to 2000, however, that pattern changed sharply. Incomes of the top 1 percent grew more than threefold, while median household income grew less than 15 percent.
Although conventional wisdom has long held that a widening income gap is a problem, there has never been a practical way to measure its actual costs. The toil index tackles that issue.
The index rejects the standard economic assumption that well-being depends primarily on absolute consumption. Instead, it assumes that the context of that consumption is often far more important. Context matters because the brain requires a frame of reference to make any evaluative judgment.
For example, is a particular family's house adequate? The answer invariably depends on the quality and size of other houses in the surrounding area.
Rising inequality has shifted the context that governs housing choices. Higher incomes at the top have led the wealthy to build bigger mansions, shifting the frame of reference that shapes demands for those with slightly smaller incomes, who travel in overlapping social circles. The near-rich respond by building bigger houses as well, shifting the frame of reference for others just below them, and so on, all the way down the income ladder.
This shift has affected not only subjective evaluations, but also the cost of achieving basic goals, like sending one's children to a good school. School quality is an inherently relative concept, too, and good schools tend to be in more expensive neighborhoods. The toil index rests on the positive link between a neighborhood's average housing price and the quality of the school that serves it.
This link implies that the median family must outbid 50 percent of all parents to avoid sending its children to a below-average school. Families that failed to rent or buy a house near the median of the local price range would have to send their children to below-average schools. The only alternative to seeing their children fall behind is to keep pace with what others are spending.
How long must the median earner work to achieve that goal? During the immediate postwar decades, when income distribution was relatively stable, the toil burden for meeting the rent of that median-price home actually declined slightly, from 42.5 hours a month in 1950 to 41.5 in 1970, according to my calculations.
But once inequality began rising sharply, the toil burden began rising in tandem. The median new single-family house in the United States grew from 1,570 square feet in 1970 to more than 2,300 square feet by 2007, an increase that can't be explained by the paltry growth in median earnings during those years. What changed was the context that governed housing choices.
By 2000, the median worker had to work 67.4 hours a month to put his or her family into the median home. The toil index thus fell by 2.4 percent from 1950 to 1970, but rose by 62.4 percent from 1970 to 2000. Yet all the while, steadily rising per capita G.D.P. painted a substantially rosier picture.
SOME economists invoke the celebrated Pareto Efficiency Principle to argue that rising inequality actually doesn't matter. This principle, named for the Italian economist Vilfredo Pareto, counts any change as an improvement if it helps some people while not harming others. Because the rich now have much more money than before and others have no less, these economists argue, society as a whole must be better off.
But that view misses something important. As the toil index shows, rising inequality has been costly even for families whose incomes have risen slightly. Such measures remind us that what really matters is families' ability to achieve their most important goals.
Robert H. Frank is an economics professor at the Johnson Graduate School of Management at Cornell University.