One of the most controversial findings in Unequal Democracy is that the incomes of middle-class and working poor families have grown much less robustly under Republican presidents than when Democrats are in the White House. The Census Bureau's Historical Income Tables show that, since 1948, middle-class incomes have grown more than twice as fast under Democrats, while the incomes of working poor families (at the 20th percentile of the income distribution) have grown six times as fast under Democrats as they have under Republicans. Only families near the top of the income distribution have done about equally well under both parties. Skeptics have suggested a variety of reasons to doubt these figures. Here are a few - and the reasons why they are not compelling.

1. The pattern simply reflects the fact that Democrats held the White House for most of the high-growth period before the mid-1970s, while Republicans have mostly been in charge in the more recent slow-growth era. While it is true that income growth has slowed considerably, especially for middle- and low-income families, the partisan differences in income growth appear in both periods considered separately. Allowing for shifts in income growth patterns unrelated to partisan control, oil price shocks, changing levels of labor force participation, and other potentially important economic and social conditions leaves the partisan differences in income growth patterns virtually unchanged.

2. The pattern simply reflects anomalous income growth under Reagan (or Clinton, or LBJ, or Eisenhower). Omitting any of these presidents from the tabulation still leaves a considerable gap in income growth performance between Republicans and Democrats. Indeed, with the striking exception of Jimmy Carter, who presided over a major recession in 1980-81, every Democratic president in the post-war era has produced annual real income growth for working poor families in excess of 2%. Only one Republican, Eisenhower, has matched that growth rate; indeed, none of the other five Republican presidents of the post-war era has done even half that well for working poor people.

3. Income growth is lower under Republican presidents because they have to correct the unsustainable policies of their Democratic predecessors. This argument founders on the fact that partisan differences in income growth are even larger in second and third terms than when Republicans succeed Democrats or vice versa. In particular, average real income growth for low- and middle-income families has been even more meager in second and third Republican terms than under first-term Republican presidents. Clearly, these results cannot be attributable to short-term corrections of misguided Democratic policies. Conversely, Democratic presidents have presided over even higher growth rates in second terms than they have in first terms, suggesting that they do not need any help from prudent Republican predecessors to produce robust income growth.

4. Doesn't Congress make economic policy? Skillful presidents can do a great deal to shape policy to their ideological ends, even when they lack working majorities in Congress. That is especially true of Republican presidents, whose ideological ends often demand inaction rather than action from Congress. Nevertheless, the partisan composition of Congress probably has some additional effect on patterns of income growth. However, it is very hard to tell because Democrats had virtually uninterrupted control of the House until 1995 and Republicans thereafter; thus, any effect of variation in the partisan composition of Congress is confounded with the effects of broader economic trends.

5. Perhaps Republican election victories are effects rather than causes of slow income growth for middle-class and poor people. Republicans are no more likely to win when income growth is slow. Rather, voters seem happy to punish Democratic and Republican incumbents alike for bad economic times. In any case, it is hard to see how differences in income growth in the second year of each administration - which is when the biggest differences in income growth patterns between Republicans and Democrats are evident - could affect elections two years earlier. Nor is there any evidence that these differences affect subsequent election outcomes two years later. Instead, voters focus mostly on economic conditions in the year of the election, ignoring the massive partisan differences in income growth produced by Republican and Democratic presidents earlier in their administrations. That is a striking and consequential failure of political accountability.