(This essay was published in Hong Kong Economic Journal on 18 December 2013)
Charles Murray’s book, Coming Apart: The State of White America, 1960-2010, admits that middle America is falling apart, but college graduates in highly skilled occupations are doing well. Between 1960 and 1980, the divorce rate of working class whites rose from about 5% to about 15%. The trend continued and between 1980 and 2010 the rate increased from 15% to 35%.
The well educated saw a parallel rise between 1960 and 1980: the divorce rate rose from about 1% to about 7.5%. No big surprise. The shocker is that the well educated divorce rate was flat from 1980 to 2010. The difference between the two groups is reflected in the rates for children growing up in broken homes: a steady increase for the working class, a low plateau for the well educated. Murray actually shows that the percentage of well educated people in happy marriages has sharply rebounded, while the percentage of the working class in happy marriages has crashed.
Polarization of Classes
Murray claims that class divisions have vastly expanded over the last half century. College graduates in high skilled occupations are not just doing well economically; they stay married and continue to be industrious, honest, and religious. The working class, in contrast, has fallen apart. Never mind their stagnant wages; they have almost completely lost touch with how to live successful, meaningful lives. The consequence for intergenerational poverty is ominous to say the least. A permanent underclass is in the making.
Murray uses a highly innovative storytelling technique to explain his analysis by inventing two fictional towns: Fishtown is the working class town and Belmont is the town of professionals. For those interested and curious you are urged to read his primer at http://www.newcriterion.com/articles.cfm/Belmont—Fishtown-7250.
I learned a lot from Murray’s book. His results on the white American working class did not surprise me. I already knew that their marriage rates were way down, their divorce rates were way up, and their children were growing up in broken homes. And I knew that lots of working class men dropped out of the labor force long before the big economic downturn in 2007-09. What I did not realize was that college graduates have been almost completely immune to these changes.
Does Murray’s study of white America have any parallels in Hong Kong?
Political Explanations of Inequality
Let us first examine the rise of measured household income inequality in Hong Kong. The Gini-coefficient is a common measure of income dispersion. From Table 1 we can see that it has increased steadily since 1971 and rose from 0.430 in 1971 to 0.537 in 2011, especially in the 1990s. Why has this happened?
Table 1: Household and Individual Income Inequality 1971-2011
The political left and most social democrats have overwhelmingly responded to rising inequality by blaming it on two factors: first, the inequities of the capitalist market economy and, second, the presence of cronyism between business and government.
The idea that markets fail to help the poor and disadvantaged is of course part of the well-known accusation that trickle-down economics has failed. The accusation is largely ideological and impossible to either prove or disprove because it is not a testable scientific hypothesis.
Cronyism is of course not a new point either. Adam Smith himself had long warned about its dangers in his Wealth of Nations (1776) and concluded that the best antidote against it was to make markets as free and open as possible. He spoke strongly against barriers to entry that protected monopolies and he advocated free trade. Professors Raghuram Rajan and Luigi Zingales of the University of Chicago have revived Smith’s argument against cronyism in their eloquently written Saving Capitalism from the Capitalists, 2003.
Demographic Explanations of Inequality
An important but often neglected cause of measured household income inequality is demographics. Demographic change has been a very important part of the increase in household income inequality in the past 30 years. To understand the significance let me make some observations about the household and individual income distribution patterns shown in Table 1.
Household income at the high-end of the distribution, the top 90th percentile (P90), continued to rise steadily from 1976 to 2011. Household income at the low-end, the bottom 10th percentile (P10), of the distribution stagnated after 1996. The household income gap between the top and bottom therefore has widened. In particular, the household income ratio of P90 to P10 more than doubled from 6.22 in 1976 to 13.11 in 2011. This phenomenon of a wider income gap between the rich and poor has recruited many to the ideology of the political left and the social democrats. But it has not really added much to our scientific understanding of why the gap has widened.
It is, however, worth pointing out that changes in the distribution of individual income differ quite considerably from those of household income. Individual income at the high-end of the income distribution, P90, rose steadily from 1976 to 2011, but so did individual income at the low-end, P10, at approximately the same rates. At the P10 end, individual income stagnated in the period of the Asian financial and economic crisis 1997-2003, but it rebounded from $4700 in 2006 to $6000 in 2011 – something that did not happened to household income. The gap in individual income measured by the income ratio of P90 to P10 did not increase very much over 35 years, rising only from 5.00 in 1976 to 6.33 in 2011. The patterns of household and individual income distributions are therefore very different. So what explains these differences?
One of the reasons for the rise in measured income inequality is the big increase in elderly and one-person families, who are usually down at the bottom of the income distribution. For example, the number of elderly individuals per thousand households has increased from 207 in 1971 to 397 in 2011 (see Table 2). Similarly the number of single person households has increased from 145 in 1971 to 171 in 2011. There may be economic reasons for these demographic changes. Some of them have to do with what is happening in the economy, but some of them have to do with social changes that have nothing to do with the economy.
Table 2: Changes in Divorced, Elderly, Single-Person and Single-Parent Households 1971-2011.
Since the last century well educated men in high skill occupations have been marrying highly educated women. But before the 1970s most of these highly educated women did not work. The higher the earnings the man had, the less likely was his wife to work in the labor force. In those times, the positive pairing of educated men and educated women dampened down household income inequality because the rich people were paired off with people who were not working and not earning any income at all.
Critical Significance of Differences in Household and Individual Income Inequality
But in the last 30 years or so this has been completely reversed, to the point where the high-earning men and women are now married to spouses who are also high-earning men and women. Today, the pairing of highly educated men and women who are both working has powered up the top of the household income distribution. Yet throughout this period the individual income gap between high-earning and low-earning individuals has not changed very much.
These economic changes within the family are tremendously important. One of the things we must realize is that it is critical for us not to think of inequality in terms of a single number. You have to look at what is behind different parts of the household income distribution. Upon closer examination everything becomes a lot more complex. Policies that are focused on a single number or a limited set of indicators can produce unexpected outcomes with perverse effects. It also explains why relying on single targeted public policies to treat complex social objectives that have multiple highly interrelated dimensions, like poverty, often falls seriously short of expectations.
Another huge change in family structure in Hong Kong in the last 30 years or so is the big increase in the divorce rate and in the number of single parent households. The number of divorces granted per 1000 households has been steadily rising since the 1970s. The number rose from 0.79 in 1976 to 8.27 in 2011. The increase in the number of divorced individuals per 1000 household also accelerated over time; and rose from 16.7 in 1976 to 117.4 in 2011. The percentage of single parent households among ever-married households (age ≤ 65 with children age ≤ 18) increased from 6.0% in 1976 to 15.4% in 2011.
Why do increases in divorce rates affect measured income inequality? How does this worsen intergenerational poverty and lead to the vanishing of the middle class? These vital issues will be discussed in my next essay on 1 January 2014. My column will take a break next Wednesday as the Hong Kong Economic Journal will not be published on Christmas Day.
Charles Murray, Coming Apart: The State of White America, 1960-2010, Crown Forum, 2012.
Raghuram Rajan and Luigi Zingales, Saving Capitalism from the Capitalists, Princeton University Press, 2013.
Fifth essay in the series on Rekindling Hong Kong’s Magic