Three Big Clues
Compared with the standard surveys that portray the rich as earning upper-middle-class salaries, Piketty and Saez’s data are like DNA evidence in a case previously investigated using only eyewitness accounts. As it turns out, the DNA evidence reveals three essential clues that were previously neglected.
Clue #1: Hyperconcentration of Income
The first clue is that the gains of the winner-take-all economy, befitting its name, have been extraordinarily concentrated. Though economic gaps have grown across the board, the big action is at the top, especially the very top.
To grasp this point, consider an alternate reality in which income grows at the same pace for all groups in society. In this scenario, the rich get richer at the same rate as everyone else, so the share of national income earned by the rich stays constant. We might call this imaginary country Broadland — a counterfactual parallel to the real world of runaway gains at the top that the writer Robert Frank has evocatively termed “Richistan.”2
Broadland would not be some kind of egalitarian fantasy. It would simply be a country where economic growth was making the income distribution neither more equal nor less. It would, in fact, be pretty close to the situation that existed from the end of World War II until the early 1970s, a period in which incomes actually grew at a slightly faster rate at the bottom and middle of the economic distribution than at the top.
But Broadland is not the world of the past generation. Instead, the share of income earned by the top 1 percent has increased from around 8 percent in 1974 to more than 18 percent in 2007 (the last year covered by the data) — a more than twofold increase. If you include capital gains like investment and dividend income, the share of the top 1 percent has gone from just over 9 percent to 23.5 percent. The only time since 1913 (the first year of the data) that this share has been higher was 1928, on the eve of the stock market crash that ushered in the Great Depression, when it tickled 24 percent.
But the top 1 percent, while seemingly an exclusive group, is much too broad a category to pinpoint the most fortunate beneficiaries of the post-1970s income explosion at the top. The Piketty and Saez evidence allows us to climb higher up the economic ladder and peer into the pocketbooks of the richest tenth of a percent and even the richest hundredth of a percent — yes, 0.1 percent and 0.01 percent — of Americans. The latter comprise the highest-earning 15,000 or so families in the United States, a group in which we would expect Lifestyles of the Rich and Famous to have little trouble finding private jets and opulent mansions.
Plenty of jets and mansions, it turns out: The top 0.1 percent (the richest one in a thousand households) collectively rake in more than $1 trillion a year including capital gains — which works out to an average annual income of more than $7.1 million. In 1974, by comparison, the top 0.1 percent’s average income was just over $1 million. (All these incomes are adjusted for inflation by expressing them in 2007 dollars.) In terms of the share of national income earned, the top 0.1 percent have seen their slice of the pie grow from 2.7 percent to 12.3 percent of income — a more than fourfold increase.
We shall say more about who is in this rarefied group in the chapters to come, but for now, let us simply note that its denizens are not, for the most part, superstars and celebrities in the arts, entertainment, and sports. Nor are they rentiers, living off their accumulated wealth, as was true in the early part of the last century. A substantial majority are company executives and managers, and a growing share of these are financial company executives and managers. High earners in law, medicine, real estate, and other potentially lucrative fields also make an appearance, but they pale in prominence to the “working rich” of the executive world.3
By now it will come as no surprise that the gains within this superrich group are themselves highly concentrated. While things have been good for the top 0.1 percent, the top 0.01 percent (the richest one in ten thousand households) has seen an even more spectacular rise. From less than $4 million in average annual income in 1974, the average member of this select group now earns more than $35 million. From earning less than 1 of every 100 dollars, these supremely fortunate souls now earn more than 1 of every 17 — or more than 6 percent of national income accruing to 0.01 percent of families. This is the highest share of income going to this group since the data began to be collected in 1913.
The more closely we look at changes in the distribution of economic rewards, the more it becomes clear that the big gains have been concentrated at the very, very top.4 According to Piketty and Saez’s revealing evidence regarding pre-tax incomes, we have gone from Broadland to Richistan — from a world in which most of the nation’s income gains accrue to the bottom 90 percent of households (the pattern of the economic expansion of the 1960s) to one in which more than half go to the richest 1 percent (the pattern of the last economic expansion from 2002 to 2007). For those in the tightly circumscribed winner’s circle of the winner-take-all economy, the last generation has truly been a golden age.
Clue #2: Sustained Hyperconcentration
The DNA evidence reveals a second important clue: The shift of income toward the top has been sustained, increasing steadily (and, by historical standards, extremely rapidly) since around 1980.
Figure 1 tells the story. The poor may not be getting poorer, but the rich have been steadily pulling away: in good times (the strong economy of the mid-to late 1990s) and in bad times (the very weak economy of the early 1980s); under Republican presidents (Reagan, George H. W. Bush, and George W. Bush, whose presidencies are shaded in gray on the figure) and under Democratic President Clinton. The only brief reversals occur during the dives in the stock market that occurred in the late 1980s and around 2000. But the occasional setback associated with a decline in the stock market has only been a springboard to new heights. For thirty years, the good times have just kept rolling.
The solution to our mystery, in short, needs to account for a simple fact: The rising share of national income captured by the richest Americans is a long-term trend beginning around 1980. It is a trend, moreover, that is not obviously related to either the business cycle or the shifting partisan occupancy of the White House.5 The partisan to-and-fro and economic ebb and flow surely had some part to play. But something else was at work in creating the winner-take-all economy — something that fostered a sharp divide between broadly shared prosperity and winner-take-all.
Clue #3: Limited Benefits for the Nonrich
We come, finally, to the third clue — and perhaps the most puzzling of all. It is so puzzling, and potentially controversial, that we spend much of the rest of this chapter documenting and delving into it.
The third clue is this: In an era in which those at the top reaped massive gains, the economy stopped working for middle- and working-class Americans. We know that the rich grew fabulously richer over this period. We know that relative to the rich, the rest of Americans lost ground. But what we have not yet investigated is whether they lost ground overall. How well did they fare in an era in which so large a share of economic gains accrued to those above them? Did they become richer and more economically secure? Did they see their chance of rising to the top increase? How much, in other words, did they really benefit from the winner-take-all economy?
The evidence can be summarized in a two-word answer: Not much. When we look at the DNA evidence on U.S. incomes, we find that most Americans experienced extremely modest gains over the era in which the rewards at the top multiplied. This is true, surprisingly and revealingly, even for the mostly highly skilled individuals just below the very top rungs of the income ladder. But the evidence on income gains actually understates the case — by a lot. When we expand our view beyond income to take in the broader canvas of the winner-take-all economy, the argument for thinking that the gains of America’s top-heavy economic growth “trickled down” becomes even weaker. This is not just a story of relative income erosion. The fallout of the winner-take-all economy has reached broadly and deeply into the security of the middle class — and, as recent events reveal, the entire American economy.
Trickle-Up Economics
Ronald Reagan famously asked, “Are you better off than you were four years ago?” Our own version of the question is, “Are you better off than you were a generation ago?” — or, more specifically, “How much better off are middle- and lower-income Americans than they were a generation ago?” The answer has substantial implications for how we judge the economic trends of the last thirty years. After all, if everyone experienced very large gains and the rich just happened to experience even larger gains, this might not cause great concern. Indeed, this is the “trickle-down” scenario that advocates of helping the have-it-alls with tax cuts and other goodies constantly trot out: If a rising tide lifts dinghies as well as yachts, who cares if it does a bit more for yachts?
But trickle-down is not the only possibility. Another scenario might be called “trickle-up”: The rich are getting fabulously richer while the rest of Americans are basically holding steady or worse. What if the modern economy looks less like an open sea, where rising water lifts all boats, and more like a system of locks, where those who don’t get through the gates are left behind? Yachts are rising, but dinghies are largely staying put, locked out (so to speak) from higher waters. Indeed, in this alternative scenario, there is reason to suspect that the dinghies are staying put in part because the yachts are rising — that the rich are closing the locks behind them to capture resources that would otherwise have enhanced the living standards of everyone else.
So which of these scenarios is correct — trickle-down or trickle-up? The evidence is not completely consistent, and there is room for debate at the margins. But it’s increasingly clear that trickle-down economics is not working as its proponents promise. Trickle-up economics, by contrast, seems to be working all too well.
Bringing In Government Taxes and Benefits
To see trickle-up in action, we need a source of evidence slightly different from that provided by Piketty and Saez. As mentioned, Piketty and Saez look at tax records, so the family incomes they report basically add up the private sources of income that people list on their tax forms: wages, salaries, investment income, gifts, and so on. These income sources do not include government benefits, such as Social Security payments. Nor do they take into account the effect of taxes themselves: They are the incomes on which people pay taxes, not incomes after people pay taxes.
These omissions matter for studying inequality, because in all rich nations, including the United States, government taxes and benefits make the distribution of income more equal, taking more from those at the top and providing more to those at the bottom. These omissions also matter because some of the compensation that people receive in the workplace takes the form of noncash benefits like health insurance and retirement pensions. Thus, if we want the most accurate measure of the resources that middle- and lower-income Americans have at their disposal, we need to take into account government’s effect on incomes as well as tally up private noncash compensation. The basic tax data include neither.
Fortunately, the Congressional Budget Office — Congress’s nonpartisan budget agency, known as CBO — has developed these broader indicators. CBO does this by combining the basic tax data with the results of income surveys that ask people about government and private benefits. In addition, CBO calculates what people with different incomes are required to pay in federal taxes. The result is considered the gold standard for studying family income trends. Although available only back to 1979—unlike the Piketty and Saez data, which go back to 1913 — this augmented DNA evidence is as close as we can get to an accurate picture of what happened to the income of American households at the bottom, middle, and top of the distribution over the last generation.6
This picture turns out to be stark: The bottom went nowhere, the middle saw a modest gain, and the top ran away with the grand prize.
How Much Did Family Incomes Grow
for the Poor, Middle Class, and Rich?
Let us start with the simplest measure of economic gains: the percentage increase in the inflation-adjusted incomes of households on different rungs of the economic ladder. The first point to make is that the overall economy expanded substantially over the twenty-seven years covered by CBO. Between 1979 and 2006 — the last year currently available — real average household income, according to the augmented DNA evidence, rose by almost 50 percent, a compounded gain of 1.5 percent per year. A household earning exactly the average income had $47,900 in 1979 and $71,900 in 2006, or half again as much.
This is the happy story. The less happy story, at least for the nonrich, is where the gains of that growth went. Start with those at the bottom: As figure 2 shows, the average income of the poorest 20 percent, or quintile, of American households rose from $14,900 to $16,500, a meager 10 percent gain over twenty-seven years, even after taking into account government taxes and benefits and private employment-based benefits.
What about those in the middle? They did better, but not that much better: The middle quintile of households (that is, the 20 percent of households above the bottom 40 percent and below the top 40 percent) saw their average inflation-adjusted income rise from $42,900 to $52,100 — a gain of 21 percent. This may sound good (and because families became smaller over this period, the gains per family member within households are a bit larger). But it works out to a real gain of just 0.7 percent a year, a rate of increase less than half the growth of average income over this period. Not much of a yearly raise.
These numbers look all the more striking when we consider a simple fact: American households are working many more hours today than they were in the late 1970s. This is because women are much more likely to work outside the home than they were a generation ago, augmenting both family income and the amount of time that household members spend in the workforce. Among working-age married couples with children, these extra hours totaled more than ten additional full-time weeks in the workforce (406 hours) in 2000, as compared with 1979. Without those additional hours and income, households in the middle of the distribution would have barely budged upward at all. The incomes of households at the bottom would actually have fallen.7
So who received the gains? The simple answer is those at the top, especially the very top. The average after-tax income of the richest 1 percent of households rose from $337,100 a year in 1979 to more than $1.2 million in 2006 — an increase of nearly 260 percent. Put another way, the average income of the top 1 percent more than tripled in just over a quarter-century. Figure 2 graphically portrays the scale of the disparity. Just getting the 2006 average income of the top 1 percent on the same axis as the average incomes of other groups is a challenge, so stark is the difference.
And, again, the gains enjoyed by the top 1 percent pale in comparison to those received by the top hundredth of 1 percent — a group CBO separated out in its analysis up through 2005. Between 1979 and 2005, the CBO numbers show, the average after-tax income of households in the top 0.01 percent increased from just over $4 million to nearly $24.3 million — more than quintupling in a little more than a quartercentury.
How Much Richer Are the Rich Because of Unequal Growth?
The statistics are stark: Most growth since the late 1970s has gone to the very richest Americans. But what does this mean in terms of the actual incomes of different income groups? Let us return to Broadland. How much more would the poor or middle class take home each year if incomes had grown at the same rate across all income groups between 1979 and 2006, as they more or less did for a generation after World War II? How much better off would Americans at different income levels be if they had stayed in Broadland over these twenty-seven years, rather than moving to Richistan?
The answer is summed up in table 1: Few of the benefits of economic growth at the top between 1979 and 2007 trickled down. If the economy had grown at the same rate as it actually did yet inequality had not increased, the average income of the middle fifth of households would be over $12,000 higher today. The average income of the bottom fifth of households would be more than $5,800 higher. Note, again, that we are assuming no change in the overall growth of the economy, just a broader distribution of the economy’s rewards. Note, too, that we are accounting here for all government taxes and benefits as well as private workplace benefits. And remember: Broadland is not some hyperegalitarian world in which the rich get “soaked”; rather it’s a world in which the rich simply experience the same income growth rate as everyone else, just as they basically had before the late 1970s.
(See our interactive calculator to find out your salary in Broadland.)
So what would have happened to those at the top if they had experienced the same average growth of their income as everyone else? The answer is that they would be pulling down around $500,000 a year on average, rather than the more than $1.2 million the top 1 percent actually earned, on average, in 2006 — a nearly $700,000 difference. Unequal growth has been very, very good for the have-it-alls.
When Richistan and Broadland are compared, another arresting conclusion comes into view. Other than the richest 10 percent of Americans, every income group would have done better if they had experienced the average growth of household income, that is, if they had lived in Broadland rather than Richistan. (The tipping point where Richistan delivers higher incomes than Broadland is somewhere between the 90th and 95th percentiles, though the gap between Richistan and Broadland is small until the very top of the distribution.) Put another way, the entire bottom 90 percent saw their incomes rise more slowly than average household income between 1979 and 2006. If trickle-up economics has a textbook case, this is it.
A Victimless Crime?
The DNA evidence shows that real household incomes for all groups but the very well-off have risen only modestly, and almost entirely because of increased family work hours. Meanwhile, the overall economy has expanded substantially. The explanation for this disconnect, mathematically speaking, is that most of the gains of growth have gone to Americans at the top of the economic ladder.
But this simple mathematical explanation leaves open an obvious objection: The economy is not zero-sum. Perhaps the gains of the rich, as impressive as they are, did not come at the expense of the rest of Americans. Maybe “Broadland” is not just an imaginary but an unimaginable country. After all, didn’t the United States grow much more quickly than other rich nations during this period, allowing the rich to get richer even as the rest of Americans moved ahead as well? In particular, didn’t the United States grow a lot faster than Europe, where incomes are generally much more equal and where the massive rise in income inequality seen in the United States did not occur?
The answer is no. The American economic engine ran hotter in some years than the European economic engine. But on average, between 1979 and 2006, economic growth per capita was essentially the same in the fifteen core nations of Europe as it was in the United States.8 The United States is richer than these nations, but the gap has been surprisingly stable since the late 1970s.
The historical evidence certainly doesn’t suggest an American economic miracle alongside European sclerosis. This lends strength to the supposition that the outsized gains of the rich came at the expense of those lower on the economic ladder, who found themselves enjoying less and less of the economic pie. On average, the economic pie grew at essentially the same rate in the United States as it did in nations where the poor and middle class have continued to enjoy a much larger piece.
Indeed, in one important respect, the pie actually grew more quickly in Europe than it did in the United States. Recall that American households are working more hours than they did in the past. The same is true in Europe, but not nearly to the same extent. Women have entered the workforce to roughly the same degree — in fact, the share of the population in the workforce grew more quickly in Europe than in the United States between 1979 and 2006.9 Yet in Europe average work hours for those in the workforce have declined, so the net effect is only a small increase in overall work hours, compared with a much more substantial rise in the United States.10 As a result, GDP per hour worked — perhaps the best single measure of a country’s economic health — actually rose faster in Europe than in the United States between 1979 and 2006.
To be sure, the story of divergent work hours is complicated: Some of the decline in work hours in Europe is involuntary, due to higher levels of unemployment. But the basic story is that the United States did not grow markedly faster than Europe even as American inequality skyrocketed. Adjusted for the growing work hours of American families — with all the attendant stress and strains that those increased hours have caused — the American economic pie actually grew slightly more slowly.
The Big Zero
If there is a nail in the coffin of the dismissive view about middle-class economic problems, it is the decade that shall not be named: the 2000s. Some call these years the “aughts,” others “the aughties,” still others “the big zero.”11 The last moniker may be the most appropriate. The decade closed with the nation mired in the worst economic downturn in more than seventy years, with the unemployment rate around 10 percent and with the share of Americans unemployed for more than six months at the highest level ever recorded.12 Economically speaking, the 2000s was truly a lost decade. At its end, the stock market was partying like it was 1999. Housing prices had crashed. Home ownership was at 2000 levels.13 Even the most optimistic estimates suggested it would take years to recover from the hemorrhaging of jobs and incomes triggered by the collapse.
Here’s the kicker: The aughties were awful even before the economy began to crumble in late 2007. They featured six years of consecutive economic growth (from the end of 2001 until the end of 2007) in which the median income of non-elderly households actually fell while the share of Americans living in poverty actually rose. This was the first economic expansion on record in which the typical non-elderly household lost economic ground.14 Yet it wasn’t all bad news. Between 2002 and 2007, the real pretax incomes of those in the top 1 percent rose by 10 percent. Per year.15
Beyond Income
Those who are inclined to be dismissive of the evidence just reviewed frequently respond by insisting that the focus on income is short-sighted: We should also look at the chance people will rise up the economic ladder, at their workplace benefits, at their spending and wealth, and so on. True enough. Unfortunately, when we take a broader perspective, the conclusion that our economy has become increasingly winner-take-all only becomes stronger.
Stagnant Social Mobility
Americans have always believed in upward social mobility — both as an ideal and a description of reality. And if social mobility had been rising along with inequality over the last generation, then the growing concentration of income at the top might be less of a concern. One year’s poor household might be next year’s middle-class or even rich household. Social mobility would soften the sharp edges of a growing class divide.
Alas, the evidence is overwhelming that upward social mobility has not increased at the same time that inequality has skyrocketed. Indeed, according to a number of innovative new studies, American mobility may well have declined over the last generation, even as inequality has risen. This is true of both individual mobility (“Am I richer than I was a decade or two ago?”) and of intergenerational mobility (“Am I richer than my parents were?”).16 Over a typical decade, for example, just under four in five people stay in the same income quintile or move a single quintile up or down.
We know less about the long-term mobility of the top 1 percent, but all indications are that people in this rarefied group usually don’t drop very far down the ladder for any length of time. Looking just at wage and salary income, for example, more than 70 percent of the highest-earning 1 percent of American households in 2004 were among the highest-earning 5 percent of households in 1994. Only around one in ten had risen from the bottom 80 percent — down from around one in seven in the 1970s.17 And of course, as the gains of economic growth become ever more concentrated, almost by definition fewer individuals will be able to move into the shrinking ranks of the economic winners.
Compared with other rich nations, moreover, U.S. intergenerational mobility is surprisingly low, in part because the gap between income groups is so much bigger. The American Dream portrays the United States as a classless society where anyone can rise to the top, regardless of family background. Yet there is more intergenerational mobility in Australia, Sweden, Norway, Finland, Germany, Spain, France, and Canada. In fact, of affluent countries studied, only Britain and Italy have lower intergenerational mobility than the United States does (and they are basically even with the U.S.).18
The differences are often stark. In the United States, more than half of the earnings advantage (or disadvantage) of fathers is passed on to sons. In Canada, only about a fifth or less is. And almost all of the difference is accounted for by the fact that Americans are much more likely to be stuck at the bottom or secure at the top than are Canadians. In short, when mobility is accounted for, the picture of the winner-take-all economy looks no better. If anything, it looks worse.
Broken Benefits
It is often argued that workplace benefits, like health insurance and retirement pensions, alter the story of the winner-take-all economy — and that, in particular, the middle class does better when you take into account the growing share of compensation that comes in the form of such benefits, rather than cold, hard cash. Actually, the CBO numbers we have been looking at already take into account health and pension benefits, so the anemic gains of working- and middle-class Americans cannot be explained away by suggesting that they are getting better and better benefits at their place of work.
But let’s take the argument a little further. Are Americans getting better benefits tied to their jobs? Not when it comes to retirement benefits. Employers contribute less to such benefits than they did in the 1970s, and workers are less likely to have an employer-sponsored pension than they were in the late 1970s.19 And many fewer have a guaranteed defined-benefit pension that pays them a fixed income in retirement. Instead, most Americans who have pensions rely on defined-contribution plans that place all the risk of retirement savings on them. This risk has been driven home by the recent stock market drop, which reduced the median balance in 401(k)s by a third between 2007 and 2008.20 As notable as the decline is the end point: the typical amount in a 401(k) in 2008 was a paltry $12,655. Given this, it’s no shock that the share of Americans who are at risk of retiring without adequate income has risen substantially since the early 1980s.21
On the other side of the benefits equation, employers and workers are certainly spending more on health insurance—much, much more. The question is what this spending has bought. Are Americans better insulated against medical costs than they were a generation ago? No. Are they more secure against the medical crises and costs that, studies suggest, are associated with more than half of the rising number of personal bankruptcies?22 No. Health costs that outstrip the growth of earnings year after year after year — pushing more people out of coverage and more people into hardship — can hardly be seen as an unqualified benefit for the middle class.
That’s especially so because these higher costs are not inevitable. The United States spends vastly more than any other rich nation on health care — both as a share of the overall economy and on a per-person basis — even though we are the only affluent democracy that doesn’t guarantee insurance to all its citizens. In 2007, the price tag for our exorbitant system came to $7,290 per person. The next most profligate nation, Norway, featured spending of $4,763 per person. Canada’s spending per capita is essentially half the U.S. level.23
Yet the United States has fewer doctors, hospital beds, and nurses per person than the norm among rich nations, and Americans (while less healthy overall) visit doctors and hospitals less often and have shorter hospital stays. Indeed, by some measures our care looks surprisingly substandard. For example, recent analyses of “amenable mortality” — deaths that could have been prevented with timely care — find that the United States has the highest rate of preventable death before age seventy-five among rich nations, and that it is falling farther and farther behind.24
And, of course, the United States has had a higher share of citizens without coverage than any other rich nation, and that share has been growing. In 2006, more than 46 million Americans younger than sixty-five were uninsured — nearly 18 percent of the non-elderly population. Back in 1979, the number of uninsured was 27.5 million, or less than 15 percent of the nonelderly population.25 Add in workplace benefits and the situation of most Americans looks even more dire than when we focus on income alone.
Consumption to the Rescue?
Can another definition of economic well-being come to the rescue? It has long been known that inequality of spending (or “consumption”) is less than inequality of income. The reasons are obvious: The rich save more of their money than the poor; at the same time, two major groups in society — retirees and young adults — often spend more than they earn, thanks to savings and loans. So we should expect less consumption inequality than income inequality. The issue is whether we see the same basic trends in inequality of consumption that we do in inequality of income, namely, a big increase.
Tackling this issue turns out to be extremely difficult, in part because the main source of evidence, the Consumer Expenditure Survey, largely misses the high-income folks who’ve benefited from the winner-take-all economy. Still, it is increasingly clear there is no “consumption paradox.” Consumption inequality is not as great as income inequality, as we’d expect (especially since the evidence fails to capture the big winners at the top). But the two have grown at more or less the same rate over time.26
Here again we find smart analysts ignoring the most notable feature of the winner-take-all economy — that it’s, well, winner-take-all. Have lower-income Americans been borrowing more to sustain their spending? Definitely. Are the less affluent experiencing lower price inflation than richer Americans thanks to cheap Chinese goods?27 Perhaps. But none of this is going to alter fundamentally our view of American inequality. The access of the poor to easy credit or Wal-Mart prices has little bearing on the growing gap between the superrich and the merely well-to-do.
Drowning in Debt
Furthermore, the flip side of consumption is savings and investment — and the wealth that comes with savings and investment. After all, it’s not as if the rich are giving all the money they don’t spend to charities for the poor. In fact, money that high earners do not spend gets turned into wealth, and the share of wealth held by the rich is both high and growing. In 2004, the wealthiest 1 percent of households had an average net worth of nearly $15 million.28 (At the very top of the ladder, the wealthiest 400 Americans — according to the famous Forbes 400 list — had an average net worth of $3.9 billion in 2008, more than six times the 1985 average for the Forbes 400.)29
The average net worth of the bottom 80 percent of households, in contrast, was around $82,000, and that includes the wealth that households had in their homes. Average net worth for the bottom 40 percent of households was a paltry $2,200 in 2004, less than half the $5,400 that this group enjoyed in 1983. Strikingly, over the entire period between 1983 and 2004, only 10 percent of all wealth gains went to the bottom 80 percent of Americans, an even more skewed pattern of growth than seen in income.30
Perhaps more striking, 17 percent of households in 2004 had zero or negative net worth—they owed more than they had. (And this was before the collapse of real estate prices put more homes “underwater” — with home loans exceeding home values—than at any time in U.S. history.) Here we find a major clue to why consumption is not always a good measure of income: For a time, at least, people can spend beyond their income by borrowing, and before the 2007 downturn, Americans were borrowing at record rates. But, as the late economist Herbert Stein is reported to have said, “If something can’t go on forever, it won’t.” The inexorable increase in household debt was not sustainable without comparable income gains — and those gains did not occur. Once again, scouring the scene turns up plenty of additional proof that the winner-take-all economy has produced limited gains for those outside the winner’s circle.
The clues are undeniable: Not only did those below the top reaches of the economic ladder find themselves falling ever farther behind the have-it-alls; they reaped surprisingly few of the gains of the massive expansion at the top.
This is an economic puzzle. It is also a political puzzle. Democracy may not be good at a lot of things. But one thing it is supposed to be good at is responding to problems that affect broad majorities. How could events and trends like these evolve with so little response from democratically elected leaders? Indeed, as we shall see, the puzzle is even deeper. For government was no mere bystander in many of these developments. It actually pushed them along. Why?
Before we turn to these questions, however, we have one last piece of unfinished business: We need to explain why the prime suspect that America’s economic detectives have fingered is, at most, a modest accomplice to the crime.
The Usual (but Wrong) Suspect
Tune into the cable money stations or read the business press and you are likely to hear an account of rising inequality that goes something like this:
“Education is the key to understanding broad inequality trends.”31
“To explain increasing inequality we must explain why the economic return to education and to the development of skills more generally has continued to rise.”32
“We have an economy that increasingly rewards education and skills because of that education.”33
These quotes were not chosen at random. They are the pronouncements, respectively, of the former head of President George W. Bush’s Council of Economic Advisers, Gregory Mankiw, a Harvard economist; Fed chairman Ben Bernanke (another economist, formerly of Princeton); and, finally, former President George W. Bush himself.
It might seem that the common element connecting these quotes is that those responsible for them have ties to the Republican Party. After all, rising inequality was an inconvenient reality for a GOP president (and, before 2006, a GOP Congress) intent on cutting taxes for the wealthy. The fact is, however, that these three quotes express what was, until recently at least, the overwhelming consensus view on inequality among economists, a view summarized in the ungainly acronym SBTC.
The SBTC Seduction
SBTC is not a regional telephone company. It stands for “skill-biased technological change,” and it is still by far the dominant explanation for American inequality trends.34 According to the SBTC argument, the last thirty or so years have witnessed a massive shift toward more knowledge-based employment. In the popular version of the argument, this shift has been greatly accelerated by the globalization of the American economy. This transformation has made formal education and advanced skills much more valuable, fueling a growing divide between the highly educated and the rest of American workers.
In some versions of the argument, skill-biased technological change is driven by computers and the Internet. In others, the main culprit is the failure of the educational advancement of most workers to keep pace with the growing skill demands of a global knowledge economy.35 The account of the crime, however, is the same: SBTC did it.
There are just two problems: SBTC isn’t even charged with the right crime. And the suspect has an alibi.
Why Educational Gaps Can’t Explain
American Top-Heavy Inequality
If there is an Exhibit A in the case that SBTC did it, it is the rising “college wage premium” — the extra amount that college graduates earn relative to their less educated peers. Each year, the College Board publishes a report entitled, “Education Pays,” in which it announces that the gap between those who have finished college and those who have not is large and growing. More sophisticated economic analyses usually emphasize the effects of education across the full spectrum of educational achievement, before college and beyond. But they reach the same conclusion: a growing “return to schooling.” From here, it’s a short leap to the view that the rising bang for one’s educational buck is the main cause of growing inequality.
Only it’s not. The return to schooling — and especially to a college degree — has risen. But, as we’ve seen, rising American inequality is not mainly about the gap between the college-educated and the rest, or indeed about educational gaps in general. It is about the pulling away of the very top. Those at the top are often highly educated, yes, but so, too, are those just below them who have been left increasingly behind.
There’s more: The college educated did well relative to those below them, but not because they experienced massive economic gains. Rather, they merely managed to avoid the devastatingly slow growth at the bottom.
Assume that the story about a new educational elite is true. The top 20 percent of the income distribution should therefore be composed almost entirely of college graduates. (The share of Americans with a college degree was 29 percent in 2007.) What, then, happened to the household income of someone at the 80th percentile, the starting point for entry into this supposedly favored class? The answer is that it has grown extremely slowly compared with the income of folks at the very top — roughly one-fourth as quickly per year as the average income of the top 1 percent.
That middle-class income growth, moreover, is mostly because of increased household work hours, not increased individual earnings. Hard as it may be to believe, a typical entry-level worker (ages 25–34) with a bachelor’s degree or higher earned only $1,000 more for full-time, full-year work in 2006 than did such a worker in 1980 ($45,000 versus $44,000, adjusted for inflation).36 And college-educated workers are substantially less likely to receive health insurance in their first job than they once were—almost four in ten now start out in the labor market without health benefits.37 So much for the enormous general rewards of a college degree.
To be sure, some workers with advanced education make enormous sums. But that’s exactly the point: A huge amount of inequality occurs among workers who have heeded the advice of “Education Pays” and sought a college degree. Economists call this “within-group inequality” (that is, inequality among people with the same education or skills), and it is one of the strongest pieces of exculpatory evidence on SBTC’s side. That is because within-group inequality, by definition, cannot be explained with reference to education, since it occurs among people with the same basic characteristics. And within-group inequality accounts for a major part of the rise in inequality since the 1970s, especially at the very top — where almost everyone has a good education.38
Maybe SBTC was at work among these workers in more subtle ways — some argue that college-educated workers with the skills to do routine tasks have lost out to those who do higher-level “abstract tasks” — but the case against SBTC, considered so strong at the outset, becomes harder to make.
But forget about the weak case, because the suspect has a strong alibi.
Why Didn’t Other Rich Nations Experience SBTC’s Wrath?
SBTC is attractive to economic detectives because it is an all-purpose criminal — one that, its prosecutors argue, can explain both the decline in American inequality in the first half of the twentieth century and its rise in the last thirty years. It is a suspect whose influence should be seen over long spans of time and, more important for the present discussion, across national borders. After all, other rich nations have computers and the Internet too — indeed, quite a few are more networked than we are — and most rich nations are more exposed to the global economy than we are. If SBTC did it here, it should have done it elsewhere, where the same technological and global shifts were taking place.
Embarrassingly for the SBTC-did-it consensus, however, SBTC seems to be picky about where to strike. When it comes to rising inequality, the world isn’t flat after all. American income inequality is the highest in the advanced industrial world. As one labor economist wryly puts it, “If there were a gold medal for inequality, the United States would win hands down… [S]tandard measures show that the United States more closely resembles a developing country than an advanced country on this measure of economic performance.”39
Yet gaps in skills, as measured by years of schooling, are not larger in the United States than they are in other affluent nations. They are actually smaller. Inequality is dramatically higher in the United States not because of greater skill gaps or greater returns to education, but because within-group inequality is greater than it is in other rich nations. Indeed, there is more inequality among workers with the same level of skills (measured by age, education, and literacy) in the United States than there is among all workers in some of the more equal rich nations.40
The Uniqueness of America’s Winner-Take-All Economy
SBTC’s alibi appears even stronger when it comes to the meteoric rise of earnings at the very top, because that rise has been substantially more meteoric in the United States than in other rich nations.
Figure 3 shows the share of income, excluding capital gains, going to the top 1 percent of households in twelve nations: Australia, Canada, France, Germany, Ireland, Japan, the Netherlands, New Zealand, Sweden, Switzerland, the United Kingdom, and, of course, the United States.41 The first bar shows the average share in the mid-1970s (1973–1975); the second shows the average share around the millennium (1998–2000).
One feature of figure 3 that jumps out is that the United States did not look all that exceptional in the early 1970s. Germany, Switzerland, Canada, even France — all had a higher share of national income going to the top 1 percent a generation ago.
Yet that has changed dramatically. The United States is now at the top of the advanced industrial pack, with regard to both the level (16 percent) and the increase (virtually a doubling) of the top 1 percent’s share of income. Half of the nations in figure 3 — a diverse group that includes France, Germany, Japan, the Netherlands, Sweden, and Switzerland — experienced little or no increase in the share of income going to the top 1 percent. Apparently in these countries SBTC was AWOL.
It’s true that the other English-speaking nations in this group — Australia, Canada, Ireland, and the United Kingdom — have followed a path more like the United States’. Still, the United States is well in the lead in the competition for the gold medal for inequality. Whereas the United States experienced a doubling of the income share of the top 1 percent, the other English-speaking nations saw only an average rise of around half that in percentage terms.
The English-speaking world has certainly emulated the American pattern more closely than other nations have. But this is hardly proof that government policy doesn’t matter, since these nations have also generally emulated U.S. public policy more than other nations have. What’s more, the trajectory of the two countries that are most often compared to the United States, the United Kingdom and Canada, cannot be viewed as wholly independent of the rise of America’s winner-take-all-economy. As we saw when we started to parse the composition of the top 0.1 percent, the rise in the compensation of the highest earners, especially corporate executives and financial managers, drives much of the outsized gains at the top in the United States. Companies in English-speaking nations compete for these workers, and thus have faced the most pressure to match the massive salaries on offer in the States.42 This appears to hold particularly true for Canada.2 While this contagion effect is hard to pin down, there is little doubt that some of the increase in top incomes in other English-speaking nations reflects competitive pressure to match the more dramatic rise in the United States—a rise that we shall see has a great deal to do with U.S. public policy.
The cross-national window just opened puts the rise of the winner-take-all economy in surprising perspective. The hyperconcentration of income in the United States — the proximate cause of the death of America’s broad-based prosperity — is a relatively recent development. It is also a development that sets the United States apart from other rich nations, calling into serious doubt the usual explanation for America’s winner-take-all economy, SBTC.
But if SBTC didn’t do it, who did? Enter the unusual suspect: American politics.
[*] We should emphasize that the Piketty and Saez data only allow us to say how well different income groups are doing, not how well individual households fare over time — an issue with traditional surveys of income as well. We can say the rich of today are richer than in the past, not how much change there is from year to year in who is rich and who is not. But, as we shall see, taking into account the upward or downward income mobility of households does not change this basic picture. It may even strengthen it, since long-term income mobility is much more limited than Americans believe, and may have declined since the 1960s. In any case, income groups are not statistical fictions. If the rich grow much richer, while the poor and middle class do not, the structure of society will look very different.
[*] Perhaps most telling, there is little sign of the same meteoric rise at the top among French-speaking portions of Canada. Executives in Quebec do not appear to be competing in the same common labor market that has allowed American pay levels at the top to diffuse to the rest of Canada.
Copyright © 2010 by Jacob S. Hacker and Paul Pierson. Reprinted by permission of Simon & Schuster, Inc., NY.
