Here's another example of how complicated it is to combat America's rising income inequality. Economists know that inequality is much lower for younger households than for older households. But unless the reasons for that are clear, formulating appropriate policy aimed at alleviating income inequality is extremely difficult.
A key question regarding the cause of income inequality is whether it's caused by differences in ability or luck. Does enhanced skill, training or education put some groups on higher income growth paths than others? Or is it because for some people, temporary changes in income arising from negative factors such as unemployment or a spell of bad health create setbacks that they never fully recover from, while for others positive shocks, such as the good fortune to find a great job, push their incomes higher?
Regarding the second condition, for most people, the shocks are generally small and roughly balance over time. But for others -- the lucky and the unlucky -- the shocks are very large or persistently positive or negative, and this could lead to more inequality as households age.
Here's where understanding which of these two potential causes -- ability or luck -- matters in determining the types of policy that will be most effective in reducing the growth in inequality as people age. If divergent income growth paths are mainly due to permanent traits such as the level of education and training, then policy should focus on improving programs that offer these skills, particularly for those at the lower end of the wage and salary distribution.
But if the problem is mainly temporary shocks to health, employment and so on that lead to income losses households never fully recover from, then more should be done to alleviate these types of risks to household income.
A recent study by economists at the Federal Reserve Bank of New York attempts to untangle these two potential causes of increasing wage and salary inequality as households age. They begin by documenting how income changes over time for males aged 25-60 (a similar study for women is underway).
They find that "the median worker in the income distribution experiences about a 38 percent rise in his real earnings between ages twenty-five and sixty. There is an impressive amount of heterogeneity: Workers below the 20th percentile actually experience a decline in earnings, while those in the top 1 percent experience a fifteenfold increase."
Next, they examine the persistence of income changes. When a worker experiences a positive or negative change in income, how long, on average, does it last?
The answer depends on the worker's place in the wage and salary distribution. For those at the bottom of the distribution, negative changes to income appear to transitory, while positive changes appear to be nearly permanent. Thus, this group appears to be held back by successive negative, transitory shocks to income rather than from the lack of permanent shocks on the upside. For those in the highest earnings group, the findings are reversed. Negative changes are long lasting, while positive changes are short-lived.
Overall, the results show that most people experience small transitory shocks to income in a given year, a few experience midsize shocks and a small but "non-negligible" number experience extremely large shocks, both positive and negative.
The results also show that an important factor in explaining the widening inequality as households age is large negative, transitory shocks that hit households at the lower end of the wage and salary distribution. These households have difficulty fully recovering from the decline in income that results from these shocks, and thus, even though the shocks are temporary, they turn into permanent setbacks.
The implication is that policies to reduce the growth in inequality as households age cannot focus solely on improving education and job training. Providing better social insurance when households experience income losses due to factors beyond their control such as a Great Recession or health problems must be an essential part of these policies.