In recent columns I have explained how education differences and fundamental changes to labor markets play a dominant role in rising income inequality. In future columns I will write about policies that might lessen the rise of inequality and assuage its negative impacts, if indeed there are any. But before doing so, I want to explain why fixing inequality is no easy task, either conceptually, or in practice.

Labor markets are, like most markets, imperfect things. Research tells us that middle-aged and older women receive a slight, two to three percent, pay deficit compared to men of equal education, occupation and experience. If you have a first name that sounds African-American (Lakisha or Jamal) you are less likely to be hired with the same resume as an Emily or Greg. If you are a white or Asian woman, getting into a good college is much, much harder, all things being equal, than if you are a black man, and forget landing that federal government job unless you are a veteran.

Despite all the drawbacks of labor markets, they do a surprisingly good job attaching wages to individual productivity. Labor markets are a place where income inequality naturally manifests itself, so it is tempting look to redistributing income as a way to remedy inequality. The tax system is the tool we currently use to flatten income, but how well does that work?

My family earns a comfortable middle class income. There are five of us and the dependent deduction for my children reduces our federal tax liability by about 15 percent. As my kids leave, I lose this deduction, but per capita income in our household will rise 250 percent. Clearly, our tax system isn't currently ready to be a tool of equalizing incomes. But, that isn't the real conceptual challenge to this problem of inequality.

Income and earnings are a small facet of employment. Benefits, especially healthcare, are a far larger share of earnings for low-income households, frequently half of total salaries. Incomes alone don't capture real earnings, especially when government also provides significant benefits to low-income families.

More critically still are non-wage amenities. Coal miners face much higher risks that florists, and part of their pay difference reflects that risk. More subtle too, are the sense of importance and non-financial rewards accompanying work. Many public sector jobs exemplify this reality. No one becomes a special education teacher, policeman, infantryman or English professor for the pay.

Still, the biggest distortion involving income inequality is leisure. Workers in the top 20 percent of income work, on average 13 hours per week more than those in the bottom 20 percent. Higher income workers also retire older, working much more over their lifetimes.

Redistributing income is, in practical terms, difficult. More critically though, earnings is only one facet of work. There are extreme differences in the non-wage amenities of work and maybe as much leisure inequality as there is income inequality. Conceptually, there’s no equitable way to fix earnings inequality unless there is also a remedy for risk, benefits, job satisfaction and leisure inequality as well. Labor markets do this best.