Putting a Finger on the Grand (Income) Canyon

Just months ago, newspaper headlines across the country trumpeted the good news: Poverty, at 12 percent, was at its lowest level in two decades and median income was at its highest rate ever. A late September headline in the Minneapolis Star Tribune blared, "Incomes rise, poverty falls; minorities, poor make strides, census finds." All hail America? Hardly, for in what seemed to be a 24-hour relapse, that report faded to black, overpowered by the more nagging issue of income and wealth disparity—the gap between the top vs. that of the bottom, which critics point out has been ballooning.

Like any spirited public debate, there is no shortage of villains and anecdotes concerning the income gap in America and the nation's current system of economic rewards. Some have called the income disparity the "dark side" of the current record economic expansion.

Not all of the arguments, however, stand up very well to close analysis. At an October retreat for the Minneapolis Fed board of directors, a handful of experts zeroed in on the nature of income disparity in the United States, and poked holes in some of the commonly perceived sources of that discrepancy. Not pretending to have all the right answers in a bottle, the speakers said income disparity defies an easily traceable, uncontaminated source.

Two factors that do appear to exert significant influence—education and urbanization—might not be surprising, but can get lost in argumentative din. Moreover, well-intentioned responses intended to narrow the income disparity—higher minimum wages, limits on trade and immigration, capping executive salaries—were not likely to have much of an impact because they did not target the underlying causes, experts told the directors.

Us against them

Arguments over income are largely framed by perspective. From an absolute standpoint, the percentage of people below the poverty line has been declining, and is well below the rates of the 1950s and '60s, considered a golden economic age by many despite the fact that poverty rates were above 20 percent for a good number of years.

But in a relative sense, the income disparity between top and bottom income earners has widened significantly over the decades. Part of the reason for this was a decline in real income for the bottom one-fifth of workers from 1979 to about 1993. But since that time, real income for this group has grown modestly. This means that growing income disparity is more the result of what's going on at the top than what's not going on at the bottom.

As such, the relative income gap overshadows any absolute improvement for the lowest-earning workers. Among many factors for this, "there's always been a strong undercurrent of envy" as it relates to income disparities in the United States, according to Arthur Kennickell, a senior economist for the Federal Reserve Board of Governors, and director of the Fed's triennial Survey of Consumer Finances.

But describing the size of the income gap and identifying its root causes are two separate and difficult tasks, Kennickell and other speakers pointed out. For starters, people mistakenly use different "rich-ness" measures—like income, net worth and wealth—as proxies for one another. "They really are not the same thing," Kennickell said. "There is this confusion, and you see it often in news reports."

That's not to say there is no relationship among these measures, but the link is more tenuous than most would believe. For example, research by the Minneapolis Fed in 1997 found concentration of earnings, income and wealth to be "significantly different." It noted that "wealth is by far the most concentrated" and the correlation between wealth and both earnings and income is "surprisingly low." [See the Spring 1997 Quarterly Review.]

Nonetheless, different measures of disparity do lend insight into how the current economic system distributes rewards, Kennickell said. Such comparisons also provide social commentary, as wealth and income become stand-ins for more romantic and intangible ideals like well-being and happiness, he said. What's important in disparity discussions is that everyone understands the financial nature of the comparison. Even seemingly straightforward comparisons of income can be misleading, Kennickell said. For instance, wealthy retirees often live off past investments and might report little in the way of income; the tax code can also be used in myriad ways to lessen their tax obligations and, by extension, income. Government programs that provide health care, housing assistance and other benefits for the poor and not-so-poor also distort income comparisons.

The dentist's income gap: open wide

All qualifications aside, the existence of a widening disparity gap—whether measured by income, net worth, earnings or wealth—is clearly borne out in current research. Kennickell's research on net worth (the sum of tangible assets, minus liabilities) found that the share held by the top 1 percent grew from about 30 percent in 1989 to 34 percent in 1998.

This doesn't mean the nation is experiencing Robin Hood in reverse—that the rich are stealing from the poor. Rather, the economic pie has expanded and the gains in net worth, in terms of both the rate of increase and total dollars, are accruing to the top much faster than anywhere else.

Financial gains made from 1992 to 1998 as a percentage of total 1998 net worth were fairly even from the 20th income percentile to the 95th percentile, averaging increases of roughly 20 percent to 25 percent, according to Kennickell's research. Gains in net worth of the top 5th percentile during this period, on the other hand, were closer to 50 percent. Add on the proportional effect—that 50 percent of 1,000 is twentyfold larger than 25 percent of 100—and the accelerating income gap starts to become clearer.

Like other public policy issues, critics are eager to identify hand-in-the-cookie-jar sources that might help identify culprits responsible for inequalities. But like many other policy issues, there are few easy villains that might explain the income disparity. Indeed, the single largest factor—education—can hardly be considered evil.

In a world of increasing sophistication, the market is paying a premium for ever-higher levels of education, according to Kevin Murphy, an economics professor at the University of Chicago and 1997 winner of the American Economic Association's John Bates Clark Medal, which goes to the economist under 40 who has made the most significant contribution to the field.

By now, most are familiar with the college premium—the fact that wage earners with a college degree typically earn more than those with a high school diploma. But Murphy's research shows that a still-higher premium has appeared for wage earners with a graduate degree, resulting in a sort of three-tiered income standard (see Chart 1).

A worker with a college degree still receives a wage premium of about 50 percent above the wages of those with only a high school diploma. That premium, however, is only about 10 percentage points higher than the wage premium collage grads enjoyed almost 30 years ago (see chart). The income disparity becomes more obvious when you look at the wage premium for those with graduate degrees, which took off from about 160 percent in 1981 to almost 215 percent by 1997.

However, education levels by themselves do not explain the wage gap—evidenced by the fact that the wage premium for both bachelor's and graduate degrees saw a decade-long decline from about 1971 to 1981, Murphy told the directors. Another critical factor is demographics—the number of people achieving higher education levels. In the 1970s, the rate of college attendance surged coincidentally with a surge in the number of college-age people (young baby boomers). So, there were more of them, and more of them decided to go to college. This growth of college-educated workers outstripped employers' need for such individuals, Murphy said, which pushed down the value of these workers relative to high school-educated workers for about a decade.

"The relative quantity of labor is the key," Murphy said, pointing out that beginning in the 1980s, enrollment rates were still high but the subsequent pool of Generation X college-goers was smaller, bringing back relative scarcity of college-educated workers.

The third leg of this stool is employer demand for educated workers. Murphy told directors that employment at companies with a high ratio of college-educated workers has grown significantly and steadily for the past three decades see Chart 2). So when graduating baby boomers hit the workplace full force, it took the private market some time to pick up the slack. This temporarily distorted the wage premium, which returned only when the stock of educated workers had been winnowed down in relation to demand.

Montana: a shooting gallery of disparity ducks

Aside from education, however, public debate often floats a variety of other ducks to shoot at as the primary source of income disparity.

Montana, for instance, has seen its state ranking in average income steadily drop over the past 50-plus years from 10th in the late 1940s to 48th by 1999—the result of a decline in average work earnings (rather than, say, shifts in either capital income or transfer payments). Critics variously argue that this earnings dip comes from the state's job mix, a poor business climate, demographics, even the amenity trade-off people make for a higher quality of life in the scenic beauty of Montana.

None of these culprits offers a very fulsome explanation for Montana's plunge in rank, according to research by Ron Feldman, an assistant vice president, and Jason Schmidt, a financial analyst, both of the Minneapolis Fed. "I don't think any [of these sources] are going to get you very far," Feldman said.

Feldman and Schmidt looked at a number of indicators that might explain lagging average work earnings in Montana. Ironically, education offered little insight in Montana's case. The percentage of Montana residents with a high school diploma surpasses the U.S. average, and the percentage with college degrees is only very slightly lower.

The high-amenity quality of life in Montana, Feldman said, "for all I know, could be very important" in explaining the decline, but such a relationship "is very difficult to capture in any way that's meaningful."

There are plenty of other theories for the decline. For example, many people point to the general decline in manufacturing jobs, and a shift in the state's job base away from natural resource extraction (like timber and mining) and toward lower-paying retail jobs. But Feldman and Schmidt found that Montana average earnings would be lower even if the state's economy were rebuilt.

First, they gave Montana a hypothetical job mix that exactly mirrored the U.S.'s industry mix, and then applied the average Montana wages earned in these various industries. They found that average wages were still far below the U.S. average. Similarly, they went back in time and reconstructed the Montana economy to have a job mix similar to the one it had in 1969, and projected average current wages. Again, they found that Montana's wages lagged (see Chart 3).

"What jobs people have [in Montana] doesn't explain much of the difference," Feldman told the directors.

Poor business climate is also a common whipping horse, Feldman said, which "is very common-sensical. It's something you'd want to look at." However, available research generally has shown that Montana ranks in the middle of the pack on business climate—which most people informally define as business taxes and environmental regulation. But many other factors are at play with regard to locational decisions for businesses, including available labor, energy costs and proximity to markets, Feldman said. He added that when state rankings on business climate "start to get serious, states act" to improve their position.

Even demographics offer little insight to Montana's eroding income ranking, Feldman said. Compared with other states, Montana has a higher concentration of men (who typically earn higher average wages), a higher percentage of workers in their peak earning years (45 to 54 years old) and a lower minority population (who typically earn lower average wages). All of this points to the fact that average Montana wages "relative to the United States are lower in similar kinds of industries," Feldman said (see Chart 4).

The only significant relationship Feldman and Schmidt could find that might explain this phenomenon had to do with geography. "It's the absence of large cities" or proximity to large cities, Feldman said. Per capita income is higher for big-city workers compared with small-city and rural workers, and rises with the increasing size of a city, something Feldman called the "urban wage premium" (see Chart 5).

Holding all other things constant, wage earners in cities of more than 500,000 earn 30 percent more than those in rural areas, which dominate the Montana landscape. The few larger cities in Montana (Great Falls, Billings) barely make it above the 20th percentile in population among U.S. cities, he pointed out.

Theorizing about the underlying factors of this urban wage premium, Feldman said, "It's got to be productivity that's driving this." Urban areas offer businesses higher density, which lowers the cost of finding labor and moving goods and ideas, and allows a higher degree of specialization in both end products and labor training. Urban areas also are better at "capturing aggregate knowledge spillovers," Feldman said, a fancy way of saying urban areas are better able to maximize the skills of workers, which ratchets up the rewards to both individual workers and the larger society.

"It's not that I don't think these [other] things aren't important, but urbanization seems to have a bigger role" in the wages people earn. Feldman also noted that average rural wages in Montana are higher than the U.S. average, "so people [in Montana] are doing well by this measure."

Unfortunately for policymakers, Feldman said, the urban wage premium "doesn't develop overnight but over time," which means quick-fix policy prescriptions are unlikely to have much of an effect.

Minimum wage: poorly engineered solution to wrong problem

In a similar vein, Preston Miller, vice president and monetary adviser for the Minneapolis Fed, told directors that increasing the minimum wage was the wrong lever for correcting the income disparity gap.

Miller said he assumed the intent behind the minimum wage—namely, the transfer of wealth from the fortunate to the unfortunate—was a social insurance objective largely agreed upon as a good thing. As such, he said, "I view the issue as an engineering problem" in finding the best way to achieve that objective.

For starters, Miller pointed out that minimum wage laws are poorly targeted, that as many as four of five workers receiving minimum wage are not the intended target—namely, a head of household. As such, minimum wage laws appear to have little broad effect on putting more money in pockets of low-earning households. Rather, they appear to artificially raise the wages of supplementary and teenage earners. Second, Miller said, few workers are earning only minimum wages, particularly given the current labor shortage.

Designers of redistributive programs need to understand the reward structure of today's economy, Miller said, which is based on three factors: luck, effort and human capital. Each of these factors is influenced by the others (it's lucky to be born smart; intelligence is part of human capital, but is maximized through effort).

Based on this tripartite, the social safety net should redistribute based on a person's luck, rather than his or her effort or human capital, Miller said. "Paying for lack of effort," he added, encourages and rewards "even less effort."

He also said Congress' "shrug and a wink" attitude toward the minimum wage was disingenuous, often sitting on legislation until its implementation was moot for wage earners. He said he would prefer to see Congress pursue policy that would ratchet up the skills of low-wage earners. "We've got to help them help themselves."

In doing so, government programs should introduce as little distortion as possible into the market, Miller said. "It is costly to interfere in markets when there is no clear evidence of market failure."

Most redistributive programs can be categorized into two basic groups, Miller said: "Before market closing" refers to those programs that interfere with market activities to achieve some social objective, including things like price supports or controls and job protection.

"After market closing" programs wait until after the market has conducted its business, so to speak, and then smooth out the day's winners and losers. The earned income credit, rental assistance and school meal subsidies are good examples of such programs, Miller said. In general, he added, Europe depends more on the "before market closing" design, while the United States favors the "after market closing" approach.

Each has had both positive and negative effects, Miller said. Europe has been successful in preserving workers' jobs, but at a cost of labor mobility and entrepreneurship. In the United States, employment volatility is more common, but so too is innovation, higher productivity and economic reward. Volatility, Miller said, "is the stick to go along with the carrot."

Ultimately, the Minneapolis Fed's interest regarding income disparities goes beyond merely trying to understand its underlying causes. The Community Reinvestment Act was passed almost 25 years ago (and subsequently enforced by the Federal Reserve System) to ensure low-income people and businesses in low-income areas have proper access to credit, which can provide the leverage needed for income growth.

Robert Townsend, an economics professor at the University of Chicago, told directors that despite the law, there is "very little data on what's going on in individual communities." As a result, there is a tendency "to think that banks are the only game in town" for individual or business credit. Overlooked are informal networks and other nontraditional credit sources that are common in many low-income communities.

Townsend conducted groundbreaking research two years ago on the access and use of different credit sources in two minority neighborhoods in Chicago. Based on this work, Townsend has been brought in to advise a similar study of the Hmong community in the Twin Cities by the Minneapolis Fed. "We need to better understand how these informal relationships work."