Felix Salmon is super-pessimistic about the prospects for government action on reducing income disparities:
[W]hen you have a progressive tax system, especially when there are surcharges on people making seven-figure incomes, you also have a system where for any given level of national income, the greater the inequality, the greater the government’s tax revenues. And indeed federal revenues have been rising faster than median wages for decades now, thanks to the rich getting ever richer.
Given the government’s insatiable appetite for cash, it’s only natural that it would prefer to tax plutocrats, spending some of that money on poorer Americans, rather than move to a world where poorer Americans earn more (but still don’t pay that much in taxes), and the plutocrats earn less, depriving the national fisc of untold billions in revenue.
The government’s interests, then, are naturally aligned with those of the plutocrats — and when that happens, the chances of change naturally drop to zero.
Emphasis mine -- and that's where my problem with Felix's argument lies: Does a progressive tax system really lead to higher inequality?
I took data on national income, income inequality as measured by the ratio of those in the 95th income percentile to those in the 20th income percentile, and government tax receipts between 1947 and 2007. It turns out that when you keep national income fixed (whether your measure is GDP, GNI, or personal income), tax revenues and inequality are inversely correlated.
For example, in 1968 those in the 95th income percentile earned 4.5 times more than those in the 20th percentile. By 2007, that measure had increased to 7.1 times. How would personal income tax revenues in 2007 have been different if inequality had stayed at the 1968 level? A rough calculation puts the value $370 billion higher than the actual $1.1 trillion that came into government coffers through individual income taxes.
I also used a couple other measures for inequality pulled from a famous study by Emmanuel Saez and Thomas Piketty: 1) the share of income by the top 10% of earners and 2) the ratio of incomes between those in the top 1% and the bottom 99%. The results were similar in that higher inequality was associated with lower tax revenue -- though the magnitudes were quite different, in both directions. So I don't quite see how the "government’s interests ... are naturally aligned with those of the plutocrats." As far as tax revenues go, history suggests that the government should tax the rich more--which would reverse a four decade decline in the progressivity of our tax system.
P.S. The GLS regression I used above was a very simple one (log(tax receipts) = log(national income) + inequality) , so controlling for other factors could potentially weed out an effect of inequality on tax revenue that is positive. But other more robust research doesn't find that to be the case.