3 3 A system is a set of things interconnected in such a way that they
produce their own pattern of behavior over time.
In the previous chapter, I linked the decline of our middle class to a convergence of historic trends. What I didn’t do was ask whether those trends were mere accidents or the result of something deeper. In this chapter, I argue that it’s the structure of our economic system that, day after day, shifts wealth from the middle to the top. This means that if we want wealth to spread more evenly, we need to change our economic system.
Americans aren’t in the habit of thinking, much less acting, systemically; we prefer breaking problems into discrete pieces. Our government agencies, academic disciplines, and nonprofit organizations all focus on specific silos of interest. They develop policies for housing, education, the environment, and so on, but treat our economic system itself—in which every silo affects every other—as off-limits. Wealth distribution is a particularly systemic phenomenon, a result of how all parts of our economy interact. It can’t be understood without viewing it at that level, nor can it be fixed without treating it at that level.
The Italian economist Vilfredo Pareto was among the first to notice that something in modern economies consistently concentrates wealth at the top. Early in the twentieth century, he observed that about 20 percent of the people in Italy owned about 80 percent of the land.1 Looking further, he saw the same pattern throughout Europe. This led him to posit that in market economies, about 20 percent of the people will always acquire about 80 percent of the wealth … because that’s how market economies work.
Pareto’s formula wasn’t purely random; it reflects what mathematicians call a power law, meaning a curve that’s exponentially skewed to one end, as depicted in figure 3.1. The alternative to a power law is a bell curve, which has a large middle with small tails on both ends. What Pareto noticed was that in untempered market economies, wealth distribution follows a power law rather than a bell curve. A century later, this thesis seems as valid as ever: the 80/20 rule understates wealth concentration in the United States today. (Compare figure 3.1 with figure 2.1.)
Figure 3.1: POWER CURVE SHOWING 80/20 RULE
Pareto didn’t say why the 80/20 rule governed wealth distribution; he just noted (to his dismay) that it seemed to do so. In 1992, two American mathematicians, Joshua Epstein and Robert Axtell, dug deeper. Using technology unavailable to Pareto, they built a computer simulation of a market economy (which they called Sugarscape) to see what properties—including inequality—emerged when it ran.2
Sugarscape begins by randomly populating a sugar-laden field with sugar-seeking agents. Like humans, the agents have a random distribution of abilities—for example, some see farther or have more energy than others. Every time the computer recalculates, the agents search for sugar as far as their vision allows, then move to the densest sugar they see (burning sugar as they go), and eat the sugar when they get there.
Sugar, of course, is a proxy for wealth. To track wealth distribution, the computer calculates the amount of sugar each agent has accumulated after every move and, on that basis, sorts them into deciles. When the game begins, the distribution of wealth is like a bell curve with a few rich agents, a few poor ones, and a large middle. As the game progresses, however, the middle shrinks and wealth concentrates. By the end of the game, the distribution of wealth is as Pareto’s law predicts.
Why is this? Epstein and Axtell tested various hypotheses. One is that wealth distribution correlates with the sugar densities (i.e., the natural wealth) where agents are born. Another is that it mirrors agents’ genetic endowments. It turns out that both of these hypotheses are wrong. If the ending distribution of wealth were linked to the random, hence fairly even, distributions of birth location or endowments, wealth distribution would be fairly even as well, and there’d be a permanently large middle class. The reason why wealth concentrates 80/20 is that every time the system recalculates, it amplifies small differences.
It’s important to note that this amplification doesn’t just happen once or even occasionally—it’s continuous. As a result, inequality spirals upward thanks to self-reinforcing feedback. Absent any countervailing feedback, differences that start small become ever wider over time.
When you stop and think about it, this isn’t a startling discovery. We all know that money begets money and that, relatively speaking, the rich get richer while the poor get poorer. And we understand that this happens because the rich use their initial advantages—money, education, and connections—to gain even more advantages. We also know that because money has the loudest voice in politics, the willingness of government to tax the rich wanes as their wealth waxes, a process that tax-reform advocate Chuck Collins calls the “inequality death spiral.”3
Though not startling, Epstein and Axtell’s finding is nevertheless sobering. It means that small initial differences, such as those in a bell curve, are inexorably magnified until they become extreme differences, such as those in a power law. Which means that, over time, our economic system will necessarily create a small upper crust and a shrunken middle.
This is a crucial point. We know that people have different capacities and drives. Some are smarter than others, and some work harder. But those different abilities don’t explain the far greater differences in rewards. Rather, extreme reward differences are driven by the compounding effects of our winner-take-all economy.
If extreme inequality is a built-in property of our present economic system, we’ve got some deep thinking to do. For this means that even if we educated our children till the cows came home, inspired or cajoled them to work harder, and got them all jobs, we wouldn’t sustain a large middle class over time. Upward income flow would remain self-reinforcing, ending only when the system crashed (and maybe not even then, as 2008 showed). The only way to get a less unequal outcome is to build some equalizing flows into the system.
WHEN SYSTEMIC CHANGES ARE PROPOSED in the United States, there are two directions they can go: toward greater government or less. Typically, the former means more regulating, taxing, and government spending, while the latter involves a larger role for markets. Within my lifetime, there’s been a preference shift from more government to less, but that doesn’t mean inequality can’t be reduced. It can be if the market itself distributes income more evenly.
In 1944, when I was two years old, my father, Leo Barnes, worked as an economist for the Office of Price Administration. (Other OPA employees included John Kenneth Galbraith and Richard Nixon.) As victory in World War II neared, many Americans feared another depression. With millions of soldiers returning and armament factories closing, a vast number of civilian jobs would be needed to keep everyone employed. What every American wanted to know was, where would all those new jobs come from?
My father offered a solution. In an article titled “The Economic Equivalent of War,” he urged the creation of a peacetime government agency that would guarantee full employment by pledging, in advance, to buy up all the durable goods—cars, washing machines, and so on—that American factories produced but couldn’t sell. As during the war, companies would be given production quotas by boards representing industry, labor, and the public; this would keep surpluses from getting out of hand. If surpluses got too high, the agency could order workweek reductions without pay cuts, thereby rewarding workers for greater productivity.4 My father acknowledged that buying unsold hard goods would cost the Treasury several billion dollars a year but contended that such costs would be “incomparably smaller” than those of large-scale unemployment. “The proper shock absorber for economic dislocation in a democracy,” he argued, “is the accumulation of inert commodities, not the living frustration of unemployment.”
Rereading my father’s proposal seventy years later, I must say that I like it a lot. It recognized that despite the harrowing scarcities of the 1930s, our modern economic machine has the ability to produce more stuff than we need with fewer workers than we have. In addressing this fundamental conundrum, my father’s approach was broadly systemic. And, had it been adopted, it would have been good for businesses as well as workers.
That said, my father’s proposal was also distinctly of its time. It relied on more government rather than less, and for that reason alone, it wouldn’t fly today. Fortunately, the system changes we need today don’t require larger government or higher taxes. What they do require is a more even flow of income within our economy. So the question we need to ask is, how can we create such a flow? What new pipes do we need, and what sort of income will flow through them?
A FULLER ANSWER TO THESE QUESTIONS will occupy the rest of this book, but let’s start by narrowing the possibilities.
First, we’ll need new pipes to deliver income on a basis other than labor. These pipes should be capable of being installed in the not-too-distant future. This means they need to mesh with the pipes we have today.
Second, the new pipes should be solidly built. Anything that requires repeated refinancing by Congress isn’t likely to last.
Third, the pipes should have — and be able to retain — a broad base of public support. This requires them to appeal across the political spectrum.
With these filters in mind, let’s consider some possibilities. One is more progressive taxation than we now have—that is, raising taxes on the rich and flowing the added revenue into the US Treasury. Progressive taxes have been applied in varying degrees for a century or so, with generally positive results. But two things limit their effectiveness as a middle class prop: the power of the rich to evade taxes and the fact that, while progressive taxes may slightly reduce the wealth of the rich, they don’t automatically lift the middle or the bottom.
Consider also a variation of my father’s plan in which the government buys socially useful things such as bridge and highway improvements and clean energy systems, rather than surplus washing machines. Something like this was attempted in 2009 when, in the wake of financial collapse, President Obama persuaded Congress to enact an $825 billion “stimulus” package with spending on infrastructure, renewable energy, and education. A onetime grab bag of this sort, however, is a far cry from a permanent set of pipes. What’s missing is a way to sustain the government’s purchasing.
A more promising precedent is social insurance, a system for sharing the risks of unemployment, sickness, disability, and old-age poverty. Social insurance was invented in Germany in the 1880s and gradually installed in all industrial economies during the twentieth century. It’s been more effective than progressive taxation in lifting the middle and lower classes, and it has the enormous virtue of automatic funding and distribution. It’s a systemic solution to systemic problems that capitalism creates but doesn’t solve. That is its genius and relevance here.
Equally interesting is that social insurance was the invention not of socialists but of the conservative chancellor Otto von Bismarck, who wanted to strengthen the German nation, build a prosperous economy, enlarge the middle class, and blunt the appeal of real socialists. Social insurance helped him do all these things at once.
Germany’s system was a form of insurance because participants paid into a common fund and received benefits when needed. It was called social insurance because everyone was covered, and employers as well as the government chipped in.
Fast-forward to the United States in the 1930s. With unemployment at 25 percent, President Franklin Roosevelt’s first task was to put millions of people to work. But another aspect of poverty required a different remedy. This poverty resulted not from financial collapse but from demographic changes that had begun decades earlier. America was no longer a rural society of multi-generational households; we’d become an urban, industrial nation of single-generation households. This meant that millions of aging parents were no longer cared for at home by their children. Moreover, because of improvements in public health and sanitation, people were living longer than ever. The result was that by 1939, more than three-quarters of Americans over sixty-five lived in poverty.5
Wisely, Roosevelt’s advisers (especially Labor Secretary Frances Perkins) saw old -age poverty not as a temporary problem awaiting an upturn in economic activity, but as a systemic problem requiring a systemic solution. What emerged was the thirty-five-page Social Security Act, which, though brief in words, was monumental in impact.
The Social Security Act created a trust fund into which active workers and employers pay and out of which retired workers receive annuities—that is, monthly payments for life—much as Tom Paine proposed in 1797. The result is a self-financing, multigenerational system in which each generation supports its predecessors in return for being supported by its successors. The fund’s trustees keep operating expenses low (they’re currently 0.5 percent of the money handled),6 and no private entity skims money off the top.
This simple system works so well that, once it was in place, future Congresses expanded it many times. They also added coverage for unemployment, workplace injuries, and medical care past age sixty-five. Today, social insurance amounts to 9 percent of our economy, and the poverty rate among elderly Americans is below 10 percent.7
Despite (or because of) social insurance’s success, a portion of America’s financial industry never liked it. These companies believed it was their right to protect Americans against risk and to be richly rewarded for doing so. So, not long after Social Security began, they mounted a nonstop campaign to privatize it.
One showdown came in 2005, when newly reelected President George W. Bush sought to create “personal accounts” out of a portion of Social Security contributions—accounts that would be managed by Wall Street. Much to Wall Street’s surprise, however, there was so much opposition, even among Republicans, that the proposal never came to a vote. In the 2012 presidential election, Republican candidate Mitt Romney picked up the torch by calling for a voucher system to replace Medicare. This did not propel him to victory.
As it happened, Roosevelt anticipated opposition to social insurance and planned for it. When asked why Social Security was funded by payroll contributions rather than general taxes, he answered: “We put those payroll contributions there to give contributors a legal, moral, and political right to collect their pensions. That way, no damn politician can ever scrap Social Security.”8
SOCIAL INSURANCE AS IT NOW STANDS can’t solve the problems of the twenty-first century, but it offers several useful lessons.
Policies come and go; institutions endure. Social insurance is far more durable than tax laws or most other public policies. That’s because it’s not so much a policy as a set of self-financing institutions. As such, it’s built into the fabric of people’s lives. People make regular contributions and expect checks to arrive as promised. Undoing such institutions isn’t easy.
Universality beats means testing. Virtually everyone in America is covered by social insurance; this gives it a huge middle-class constituency. It also makes social insurance very efficient: it serves more people for less money per capita than private insurance does or can, in part because it has no marketing or underwriting (deciding who is eligible) costs.
Universality also avoids the pejorative distinctions that come with means testing. If only economic “losers” get benefits, they become “takers,” “moochers,” or whatever is the slur du jour. Those who don’t get benefits resent those who do, and those who do feel bad about themselves. No one is happy with the arrangement.
Means testing, in short, divides society while universality unites it. To put it another way, means testing necessarily pits one class against another—the very definition of class warfare—while universality treats us as a single society.
Build external costs into current prices. Because of its contributory structure, roughly half the cost of supporting workers in old age—the part that’s covered by employer contributions—is internalized into current prices. That’s because businesses include their social insurance contributions in their cost of goods sold. Other currently externalized costs ought to be treated the same way.
Build the pipes first; then add water. When Social Security began, payroll contributions were 1 percent, benefits were around $20 a month, and large categories of workers were excluded. Over time, as the system became better known, it also became more popular. Benefits grew, more people were covered, and health care was added.
President Roosevelt’s Committee on Economic Security, which drafted the original Social Security Act, did so with this long-term vision in mind. “A program of economic security, as we vision it, must have as its primary aim the assurance of an adequate income to each human being in childhood, youth, middle age, or old age,” the committee wrote in its report to Congress. “A piecemeal approach is dictated by practical considerations, but … whatever measures are deemed immediately expedient should be so designed that they can be embodied in the complete program.”9
THERE ARE TWO POTENTIAL ROUTES through which non-labor income can flow: one runs through markets, the other through government. My preference is for the former. But as markets are constituted today, nonlabor income doesn’t flow to nearly enough people. Some important pipes are missing.
The most obvious absence is a set of pipes that collect property income from multiple sources and spit it out to everyone. Such a system wouldn’t be hard to build; it’s essentially a mutual fund with some 300 million shareholders, each with a nontransferable share. It’s safe to say that the software and hardware to run such a fund already exist. The barriers to building it are political, not technical.
In the remainder of the book I make the case for building such pipes, using as income sources wealth that already belongs, or should belong, to all of us equally. The governing principle behind these pipes would be everyone-gets-a-share rather than winner-takes-all. The new pipes wouldn’t displace our existing ones—like social insurance, they’d complement them—but their equal distribution of a new sort of property income would offset the distorted distribution of the currently dominant sort. The result would be a market economy with a large and secure middle class, even as labor income declines.
To fully understand the case for adding these new pipes, it’s necessary to understand how our existing ones work. This isn’t always easy, but the basic patterns, if not the intricate details, can be discerned.
Our present pipes are designed to carry a mysterious substance I will henceforth call rent. Rent is mysterious because, despite its immensity, there’s actually a “vast conspiracy” to keep it shrouded. (By “conspiracy” I mean here a loose network of billionaires, CEOs, lawyers, lobbyists, PR firms, politicians, and economists who benefit, first or second hand, from the current flow of rent.) The conspirators are keenly aware that the more hidden they can keep rent, the more of it they’ll collect—and conversely, that the more broadly rent is noticed and understood, the more broadly it will be shared.
So let me tell you about rent, or more precisely, two kinds of rent: extracted and recycled. One is the dominant form today; the other is the kind we need more of tomorrow.