Scott Sumner has become famous in the internet world and elsewhere as monetarism's most capable defender. Sumner has a lot of things to say, but one is illustrative for my purposes here. Sumner argues that advocates of fiscal stimulus often make the mistake in their arguments of assuming away monetary policy as static or accommodating. His point is that you can't do that because the efficacy of fiscal policy always depends on what monetary policy is doing in the background.
This same basic point also needs to be driven home for those who want to talk about how this or that thing will affect the distribution of income in society. Matt Yglesias' review of Tyler Cowen's new book gives us an excellent jumping off point for what I mean. In his book, Cowen apparently argues that coming technological changes will have certain negative effects on median incomes. Yglesias rightly points out (as does Cowen) that this outcome will only come if we fail to implement certain policies that will cause income to be distributed in some other, more desirable way.
When we talk about how economic changes, technological swings, and even education will affect the distribution of income in society, we always sort of assume away our government's distributive policy as if it will or must remain static. But that's not true at all. At any time we can change the huge set of policies that direct the distribution of income in society to something else.
The last few decades of median income stagnation didn't have to happen. Even if you say it was caused by international competition or technological change or whatever else, the point is that if we had put a different set of distributive institutions into place, we could have avoided the maldistribution of income that we have seen. It is not like the median incomes stagnated because the economy as a whole stagnated. Quite the contrary: the economy is much larger on a per capita basis now than it used to be. If we had wanted to make sure median incomes continued to rise, we could have done that. We would have just needed different distribution policy.
Importantly, this observation about the centrality of distribution policy cuts both ways. Suppose you have thought of some way to create a more equal distribution of income within the distributive institutions we currently have. Education stuff is usually touted as doing this. There is no guarantee that once you've done what you need to do to make this income equalizing happen that the distributive institutions wont just change and rob you of your victory. The rich elites, provided they have enough power, can just change the distributive rules to make sure the money keeps flowing their way. Don't think it can happen? Read a Dean Baker book.
This is why I am always harping on distribution policy as the way to handle inequality and poverty. In a rich country like ours, a favorable distribution policy is almost all that matters. If we want a more equal distribution of income, we can make it happen by changing some policy. If we want to make sure technological change doesn't create more inequality, we can make that happen by changing some policy. If we want to make things way more unequal than they already are, we can also make that happen by changing some policy.
Distributive policy is not omnipotent obviously. There are real constraints in the world, namely how much stuff there is to go around. The way you design your distributive policy will also have some effect on how much stuff is produced in the first place. But within those confines, distributive policy is the thing that matters. It is both necessary and sufficient for equitable income distribution. Nothing else is either.
Once you understand what I am saying, you realize that the old leftist cliche is right: who gets what is purely a matter of political power. If the rich elites continue to run the show, it doesn't matter what you do. The distribution will continue to favor them and income stagnation and inequality will never be whipped. Even if you make some strides against those things within the existing set of economic rules, they can use their power to change those rules to preserve the distribution they like. What ultimately matters for inequality is who has the power to control the levers of distribution management and what result they are aiming to produce. Everything else - technology, education, competition, and so on - is completely secondary.
