Over at National Review’s blog, Ramesh Ponnuru wonders why it seems that divided government – aka, gridlock – tends to lead to slower growth in the federal budget.

I mount a defense of gridlock in my new book, Buck Wild: How Republicans Broke the Bank and Became the Party of Big Government. In chapter 8, to be precise. The numbers that Ramesh cites in his post come from a review of my book by Phil Kerpen on the NR website.

By way of jumping into the discussion Ramesh has started, let me clarify a few things first.

In my book I only analyze the real per capita growth rates of government spending in the years 1965 through 2005. I think this is a more useful timeframe for comparison than Ramesh’s seventy-six-year timeframe simply because the post-Great Society welfare state looks vastly different than what existed before. I was mainly curious to see if the correlation holds up within a timeframe that yields more consistency in terms of the entitlement programs being funded in the federal budget.

The methodology I used was modeled after that of Cato chairman William Niskanen and Cato senior fellow Peter VanDoren in the paper they presented to the Public Choice Society meeting in May of 2004. My data falls into the same realm of statistical significance, too. (Incidentally, their analysis goes back to 1949.)

Now on to the fun.

Divided government – gridlock – is the norm, not the exception, in American politics over the past 40 years. The only completely united government scenarios existed during the presidencies of Lyndon Johnson, Jimmy Carter, and George W. Bush. It is obviously true that Democratic united government is more common than Republican united government.

So why might divided government be more conducive to restrained spending? For starters, defense spending varies widely over the past 40 years, but the correlation between wars and united government is quite consistent. American participation in every war involving more than a few days of ground combat was initiated by a united government. You could argue that this is mere coincidence. Or you could argue that united government creates an environment where there is less resistance from Congress when a president wants to exercise his powers as commander-in-chief. The burden of proof is on those who suggest this is simply happenstance.

A president and Congress could certainly offset defense budget increases with cuts in the non-defense budget. Indeed, you often see this in periods of gridlock. The converse is also true: Decreases in defense spending tend to coincide with increases in non-defense budgets under gridlock.

In the periods of united government, however, those sorts of trade-offs disappear. Instead, united government spawns large increases in across-the-board spending on everything. That’s exactly what we saw during the presidencies of Lyndon Johnson, Jimmy Carter, and George W. Bush.

Why this occurs might best be described by the nature of Washington politics. The one thing you can usually count on in DC is partisanship. When Republicans are the beleaguered minority—or a congressional majority fighting a big-spending White House—they are in their element. Big Government is the clear enemy. But once they find themselves in control of it, they are less willing to rein it in.

We can see this by comparing how a GOP Congress treated the proposed non-defense budgets of Bill Clinton and George W. Bush. Remember that once the president’s budget reaches Capitol Hill every year, Congress either increases or cuts the spending request. During the years of divided government under Clinton, the Republican Congress managed to cut Clinton’s domestic spending requests by an average of $9 billion each year between fiscal 1996 and 2001.

Contrast that with the budget outcomes under President Bush—specifically the years in which Congress was held entirely by Republicans. Between fiscal years 2003 and 2006, Congress passed, and Bush refused to veto, non-defense budgets that were an average of $16 billion more than the president proposed each year.

The rules of partisanship imply that a Big Government scheme proposed by a Republican president is more likely to be accepted by a Republican Congress than if it were proposed by a Democrat. That’s exactly what happened with the Medicare drug benefit. It seems quite likely such a drug benefit would not have passed if it were proposed by, say, President Al Gore or President Hillary Clinton. And if the Medicare expansion did get traction in Congress, Republican leaders would probably have been more interested in slowing it down or tacking on substantial reforms instead of abandoning the reform elements in the hope of speeding its passage as they did in 2003.

All the gridlock combinations that have persisted for longer than two years produce average (real per capita) budget growth rates that are half that of the united government average. This includes Republican presidents facing Democratic congresses – like Reagan, Nixon, and Ford – as well as a Democratic president like Clinton combined with a GOP congressional majority in both houses.

It seems to me the main question now becomes this: Even if someone doesn’t believe we would be better off with gridlock, is it still reasonable to believe – based on the historical evidence – that we’d be worse off?