Optimal inequality for economic growth, stability, and shared prosperity: the economics behind the Wall Street Occupiers Protest?

The American 'experiment' with unfettered finance and huge inequality has not produced the promised economic nirvana. Rather, the evidence suggests that it has adversely affected the economy in various ways.

(pages 5-11 of printed journal)

By Richard B. Freeman


Protest illustration

In autumn 2011, Occupy Wall Street moved into the financial district of New York City to protest income inequality in the US. Dividing the country between the upper 1 per cent whose income had increased massively and the 99 per cent whose income had stagnated in preceding decades, the occupiers claimed to speak for the majority under the slogan 'We are the 99%'.

In similar protests throughout the US and in countries from Australia to Mongolia, the occupiers accomplished what US academics and unions had failed to do: move inequality from a sub rosa topic that invoked charges of inciting class warfare to the centre of economic discourse. President Obama, who earlier defended Wall Street's million-dollar bonuses as the just reward of 'savvy businessmen', spoke in October 2011 of 'broad-based frustration about how our financial system works...[in which] a lot of folks who are not doing the right thing are rewarded'. Republicans who initially denounced the occupiers as a mob 'engaging in class warfare', or who castigated them for not being rich, switched gears to make sympathetic comments.

When I gave the Miegunyah lecture 'Optimal inequality for economic growth, stability, and shared prosperity' in August 2011, I had no inkling the occupiers were less than a month away and that their demonstration would open the space for serious discussion of inequality. I have added the subtitle to this paper to reflect their contribution.

My analysis focuses on three questions:

- Is there a level of inequality that optimises economic growth, stability, and shared prosperity? My answer is yes. The relation between inequality and economic outcomes follows an inverted-U shape, so that increases in inequality improve economic performance up to the optimum and then reduce it.

- Did inequality in the US exceed the optimum level in the 2000s, as the occupiers believed? With the highest inequality of any major economy, the US is the best case for addressing the question of whether inequality can go beyond the optimal level in a market democracy. Arguably, the huge rewards at the top of the US income distribution drove the financial chicanery and excessive risk-taking that led to the implosion of Wall Street and the great recession. Inequality reached levels that harmed growth, stability, and shared prosperity.

- What danger does excessive inequality present to the future of market democracy? High inequality creates a society in which wealthy 'crony capitalists' dominate corporate and government policies, subvert market competition and corrupt democracy in order to maintain their position atop the income hierarchy. This system differs from the state capitalism found in China, Russia and some developing countries, in that large corporate interests and the barons of wealth rather than central governments control economic policy. I call it economic feudalism.

The relation between inequality and growth, stability, and shared prosperity

Economists have greater tolerance for inequality than most people. They stress the importance of the monetary incentives associated with inequality in motivating people to undertake economically valuable behaviour. The communist experiments proved that inequality can be too low for an economy to function well.

But inequality that results from monopoly power, rent-seeking or activities with negative externalities that enrich their owners while lowering societal income (think pollution or crime), adversely affect economic performance. High inequality reinforces corruption by allowing a few 'crony capitalists' to lobby politicians or regulators to protect their economic advantages. When national income goes mostly to those at the top, there is little left to motivate people lower down. The 2007 collapse of Wall Street and bailout of banks-too-big-to-fail showed that inequality in income and power can threaten economic stability and give the few a stranglehold on the economy.

Stipulating that inequality can be either too low or too high for optimal economic performance implies that output follows an inverse-U shaped curve with respect to inequality. Figure 1 illustrates the relation. When inequality is zero and output is modest, there is no incentive to try harder or invest in risky activities. In this state, increases in inequality raise output. Then, when inequality is extremely high, output is also low. The few people with the skills or background to compete for the top jobs work hard, while everyone else coasts because they have little or no chance of reaching the top. Between these extremes lies the optimal level of inequality (I*).

Figure 1: Inequality-output relation is an inverse-U.

Increases in inequality raise output if economy is to the left of I*. Decreases in inequality raise output if economy is to the left of I*.

Figure 1, Inequality-output relation is an inverse-U.

The peak output I* does not necessarily reflect the ideal point. Societies may choose to trade-off total output for more-preferable distributions of output, for instance, by preferring inequality at F (because, say, the politically-dominant majority gains more from lower inequality than they lose from lower output). This is Okun's (1975) famous equity-efficiency trade-off. Or a society may prefer inequality above I* (because, say, a wealthy politically-dominant minority gains more from higher inequality than they lose from lower output). Society pays for this in the form of lower output.

What is the evidence for the inverse-U shaped curve in figure 1? International comparisons show that countries with higher levels of GDP per capita have lower Gini coefficient measures of inequality. But this pattern presumably reflects the effect of economic development on inequality more than the effect of inequality on output. Comparisons of growth rates of GDP per capita among areas that differ in their initial level of inequality or whose inequality changed would seem to be a better way to identify the impact of inequality on economic outcomes. The studies that do this report widely different results depending on the data set and country sample, and thus are not particularly helpful.

In a laboratory experiment, Alex Gelber and I (Freeman and Gelber, 2010) organised subjects into groups of six and asked each person to solve a packet of mazes. We rewarded them for the number of mazes they completed under three incentive systems: a low inequality system in which everyone in the group received the same amount of money regardless of what they produced; a high inequality system that gave a large prize to the person who solved the most mazes and nothing to anyone else; and an intermediate incentive system that gave increasing rewards to persons who ranked higher in the maze competition. Figure 2 shows that this design produced an inverse-U relation between inequality and output. The group with no incentives had the lowest output, the group in which only the top person earned a prize had a modestly higher output while the group with the middling level of inequality solved the highest number of mazes.

Figure 2: Reported number of mazes solved in maze experiment at given incentives

Figure 1, Reported number of mazes solved in maze experiment at given incentives

Outside the laboratory, evidence suggests that firms that incentivise the bulk of workers do better than firms that give incentives primarily to a few high-paid executives. Presumably as a result, companies in the US and elsewhere have increasingly adopted profit-sharing and gain- sharing modes of pay, employee stock ownership schemes, and all-employee stock options.

The rising tide of inequality

The main thrust of economic reform in most advanced countries post Reagan-Thatcher has been toward policies that benefit people in the top rungs of the income distribution. Many goals lowering marginal tax rates on high incomes, deregulating finance and other industries, lowering welfare-state safety nets, privatising state-run businesses and giving tax breaks to capital income were not ostensibly to enrich the rich but to create incentives for those at the top to make decisions that would improve resource allocation and raise output. Policy analysts expected that the benefits from increased inequality would eventually trickle-down to all workers and broadly improve economic well-being.

This did not happen in the US. Since the 1970s upper-income earners have gained nearly all of the improved productivity while incomes have stagnated for most workers. In 1970 the upper 1 per cent received 8 per cent of national income; in 2007 they received 18 per cent. Most of this increase went to the upper 0.1 per cent of income recipients. In 1970 this group had 2.7 per cent of national income and incomes 27 times the mean income; in 2007 they had 12.3 per cent of national income and incomes 123 times the mean. If the share of income going to the upper 0.1 per cent had been constant, and the economy performed as it did, the income of all people in the remaining 99.9 per cent, including those in the lower 90 per cent of the upper 1 per cent, would have risen by nearly 10 per cent.

The US is an extremophile in inequality. The CIA's World Factbook reports that in 2007 the US Gini coefficient was 45. Ranking countries from lowest to highest inequality, the CIA places the US in 93rd position among the 134 countries for which it provides data. If the US was a developing country, it would be in the bottom one-third of the ranking by inequality; if it were in Africa, where inequality is high, its Gini coefficient would rank it 20th of 35 countries.

Who is in the upper 0.1 per cent? In 2005 some 63 per cent of people in the upper 0.1 per cent worked in finance and real estate (Bakija et al., 2010). Much of their income was capital income. In 2011 the top 0.1 per cent received 38.1 per cent of all US capital income while the top 1 per cent received 56.8 per cent. The IRS reports that in 2007 the 400 persons with the highest adjusted gross incomes earned 10 per cent of all capital gains, and 4 per cent of both interest and dividends. These high income earners benefited from the lower tax rates on capital income than on labour income and from the increased share of national income going to capital. If they were executives they also benefited from compensation systems linking earnings to capital income through stock options and restricted stock payments.

The concentration of people in finance and real estate at the top of the income distribution reflects in part the repeal of the Glass-Steagall Act, which Congress enacted in 1933 to establish walls between the deposit-taking commercial banks and the riskier investment banks that issue securities in order to limit financial speculation and the risk of financial meltdown. The repeal permitted a massive expansion of finance, created banks-too- big-to-fail and drove huge increases in Wall Street incomes. In 1990 total compensation for security and commodity brokers was 31 per cent of the compensation for federal civilian employees; by 2007 it was 93 per cent. In 2007 finance absorbed 40 per cent of business profits. Wall Street hired some of the country's brightest university graduates to develop new financial instruments, which it peddled as essentially risk free. Had these new financial instruments reduced risk worldwide, the bankers and financiers would have earned their high pay instead, they leveraged the new instruments, lowered mortgage standards and increased risk until Wall Street imploded.

In terms of the ratio of the earnings of workers high in the income distribution to workers low in the income distribution, in 2009 the earnings of people in the 90th percentile was 4.98 times that of people in the 10th percentile in the US. This compares to a 3.05 average for advanced OECD countries and 3.33 for Australia. From 2000 to 2009 the ratio in the US increased by 0.49 points, which compares to an average increase of 0.07 points for the advanced countries. As an indicator of the position of low paid Americans, in 2010 the US federal minimum wage was US$7.25 while the Australian federal minimum was AUD15.51. Using the exchange rate, the minimum wage workers in Australia earned twice as much as minimum wage workers in the US.

Part of the increase in upper-bracket earnings took the form of a massive growth in CEO pay. Estimates by Business Week and the Institute of Policy Studies show a CEO-to-worker pay ratio of 42:1 in 1980, of 107:1 in 1990 and of 325:1 in 2010. Much of this high CEO pay took the form of stock options, restricted shares or bonuses. When share prices rise, the owners of options and shares benefit even if price rises reflect factors outside their control. When share prices fall, boards dominated by executives often issue new options at the abnormally low market prices which pay off handsomely when the market recovers. Some firms did this immediately after 9/11, turning a national disaster into a way of lining their own pockets. Similarly, in December 2008 when the financial sector needed massive government support to survive and the federal government ran a huge deficit to stimulate the economy, Goldman-Sachs issued nearly 60 million options to its executives, who benefited from the taxpayer bail-out and ensuing recovery.

Much of the inequality in US incomes occurs among individuals with nominally similar skills. The dispersion of pay in the US among workers with similar scores on adult literacy and numeracy tests exceeds inequality in pay among all workers in Germany, the Netherlands, and Sweden. Similarly, the dispersion of pay among US workers with the same age, gender, ethnicity, years of schooling and so on exceeds dispersions of pay in many advanced countries. Equally surprising, Barth et al. (2010) found huge differences in pay among workers with similar measured skills by the firms in which they work.

If the US had high rates of social mobility, or if the economy had performed exceptionally well in the period of high and rising inequality, one might argue that inequality approximated the optimal level.

The US economy performed well enough pre- 2007 to convince many analysts that it was the best-performing major capitalist economy, but subsequent events proved this view illusory. Instead of reducing risk, Wall Street had enveloped the real economy in a highly-leveraged financial house of cards an estimated $22 of derivatives for every dollar of goods and services produced in 2009. Instead of creating a flexible labour market that could rebound quickly from a major economic shock, the US rapidly laid off workers in the recession but then increased employment slowly. Joblessness grew and many workers left the workforce because of the lack of demand for labour. Poverty increased and even fully-employed workers came to rely on food stamps to keep their heads above water.

The US experience casts grave doubt on the proposition that more inequality brings more efficient economic performance and that economic reforms should invariably favour corporations and the wealthy on the notion that they are 'job creators'. But does the experience tell us something more that the high level of inequality contributed to the financial collapse and weak recovery, and thus exceeded the optimal level?

Even before the crisis, some analysts such as Bebchuk and Fried (2004) showed that executive compensation systems were not working as promised. But most experts were not particularly troubled. It took the financial disaster and recession to encourage more critical thinking about the pay- performance nexus. More and more studies are finding that rewarding only those at the top does not improve economic performance. Bianchi and Freeman (2012) found that firms that incentivise executives with high-powered stock options have slightly worse performance than firms with lesser incentives.

When the financial sector was booming, there was a similar complacency among researchers. Many examined Wall Street through the lens of the efficient market hypothesis, which assumes more or less that whatever the market produces must be right. Alan Greenspan's 2002 paean to credit default swaps 'As the market for credit default swaps expands and deepens, the collective knowledge held by market participants is exactly reflected in the prices of these derivative instruments' reflects this Panglossian view of finance.

Financial crises often uncover widespread crime through which the financial sector makes huge earnings by fraud, insider trading, options backdating, securities scams, financial misreporting or amoral practices bordering on illegality. Even before the Wall Street disaster, FBI Suspicious Activity Reports highlighted a rising trend in mortgage fraud while SEC data showed substantial financial misreporting. Investigators uncovered and successfully prosecuted the top executives of Enron and other firms. In many cases, the firms accused of financial misfeasance agreed to pay large fines as long as they were not forced to admit criminal negligence. I anticipate that ongoing investigations into the behaviour underlying the financial implosion will uncover much venality, chicanery and crime motivated by the chance to make huge sums of money.

Finally, there is the potential link between high levels of inequality and the behaviour of regulators and legislators. Imagine this: you work for the US Securities and Exchange Commission (SEC) and have to judge whether a particular firm has defrauded its customers. You earn $120,000 a year. The firm regularly hires experts from the SEC at $500,000 a year. You are thinking of leaving the government. The evidence is complicated. How do you decide? The term 'regulatory capture' represents what economists think often happens: you decide the evidence supports your potential future employer. If inequality were less and the firm paid experts $150,000 a year, I hypothesise less capture of regulators and more capture of firms/individuals defrauding their customers.

Hazchem: from market capitalism and democracy to economic feudalism

The greater the concentration of income in a society, the more the wealthy will act collectively to advance their interests against that of other citizens. 'Crony capitalism' describes economic systems where wealthy insiders work in concert with politicians to enrich themselves. Many blame the economic problems of developing countries, such as the late-1990s Asian financial crisis, on this phenomenon. Until the latest financial crisis, most economists viewed the US and other advanced capitalist countries as largely immune to 'crony capitalism'. 'Regulatory capture' the process by which regulators become so entwined with the group they regulate that they worry more about the interests of that group than the interests of the public was supposed to be a problem limited to specific industries, regulations and agencies rather than a systemic disease that afflicted the entire economy. The Wall Street crisis and its aftermath changed this view.

The evidence shows the extent of such behaviour in the US. In the mid-2000s nearly 35,000 lobbyists were registered to lobby for Congress alone. Thousands more lobbied other government agencies and the courts. About 43 per cent of Congress who left government to join private life registered to lobby. Between 1998 and 2008, the financial sector finance, insurance, and real estate spent US$3.44 billion on lobbying alone and contributed US$1.74 billion to political campaigns for a total investment in rent-seeking of US$5.18 billion. The US Supreme Court decision that individuals and corporations can spend unlimited sums of money in political campaigns has strengthened the 'crony capitalists' power to affect the 2012 elections.

In 1961, before inequality had increased to the point in which the wealthy would invest massively in politics and lobbying, President Eisenhower warned the US about the danger of groups that had unwarranted influence and misplaced power the industrial military complex. Today, I believe he would warn the country about the unwarranted influence and disastrous rise of misplaced power of a different complex the 'financial-political complex'. The increase in inequality is intrinsically connected to the rise of this danger to the country, both as a cause of the spread of 'crony capitalist' operations and, given the success of these operations, as an effect as well. It has brought the US to a level of inequality far beyond the optimum in figure 1. It risks turning market capitalism into what I call economic feudalism a social system in which the upper 0.1 per cent, or even fewer, dominate the economy and politics.


Can the US reduce inequality toward its optimal level and reverse the movement toward economic feudalism? I believe that the concerted efforts of concerned citizens can do this; and the activities of the occupiers are a sign of change in society. The increased opposition of conservatives some in the 'tea party' movement to 'crony capitalism' is a sign that the campaign to rein in the excesses of inequality in the economy and polity cut across ideological lines.

Part of the push back will come from individuals and groups using low-cost Internet-based information and communication technology and social media to increase transparency and pressure business and government to reject policies that benefit the few at expense of the many. Internet campaigns forced the Bank of America to drop a planned banking fee on its low income customers and forced Verizon to withdraw a $2 payment fee for customers paying their telephone bills online.

Part of the push back will come from inside the financial-political complex, as more and more people realise that finance has become a huge rent- seeking machine instead of a system for allocating capital from investors to its best use in the real economy. And part will come from the economic superiority of firms that distribute the incentives and rewards to a larger proportion of the workforce; and from the innovative activities of entrepreneurs that use the latest science and technology to create new products and processes in the real economy.


Bakija, J, Cole, A & Heim, BT 2010, 'Jobs and income growth of top earners and the causes of changing income inequality: evidence from US tax return data', November http:// web.williams.edu/Economics/wp/Bakija ColeHeimJobsIncomeGrowthTopEarners.pdf

Barth, E, Bryson, A, Davis, JC & Freeman, R 2010, The contribution of dispersion across plants to the increase in US earnings dispersion (mimeo), NBER.

Bebchuk, LA & Fried, J, 2004, Pay without performance: the unfulfilled promise of executive compensation. Harvard University Press, Cambridge.

Bianch, J & Freeman, R 2012, 'Are executive stock options effective incentives or effective rent-sharing?' (in process), NBER.

Freeman, RB 2007, America works: critical thoughts on the exceptional US labor market, Russell Sage Foundation, New York.

Freeman, RB & Gelber, A 2010, 'Prize structure and information in tournaments: experimental evidence', American Economic Journal: Applied Economics, 2:1, pp. 149164.

Okun, A 1975, Equality and efficiency: the big trade off, The Brookings Institution, Washington DC.

Top ^

A condensed version of the Miegunyah Lecture delivered at the University of Melbourne on 24 August 2011.

Richard B. Freeman holds the Herbert Ascherman Chair of Economics at Harvard University, and is Director of the Labor Studies Program at the National Bureau of Economic Research.

Authorised by: Brooke Young, Director, Marketing and Commercial Engagement
Maintainer: Web Communications Officer, Faculty of Business and Economics

Disclaimer & Copyright | Privacy | Accessibility

The University of Melbourne ABN: 84 002 705 224
CRICOS Provider Number: 00116K (More information)