Gene Epstein of Barron’s devotes his latest “Economic Beat” column to the common election-year claim that the American economy fares far better under Democratic presidents than Republicans.

According to a July 2015 study by economists Alan Blinder and Mark Watson (“Presidents and the U.S. Economy”), the contest isn’t even close.

In what New York Times columnist Paul Krugman celebrated as “partisan differences” that are “actually stunning,” Blinder and Watson report a “1.79 percentage point gap,” which they call “astoundingly large.”

The gap arises by taking real gross-domestic-product growth under all of the U.S presidents since World War II, from Harry Truman to Barack Obama. Average economic expansion under Democrats ran a robust 4.33%, against a much lower 2.54% under Republicans, the study asserts.

These figures are weighted according to the number of calendar quarters during which each president served. For example, growth under one-term presidents, such as Jimmy Carter and George H.W. Bush, gets half as much weight as that under two-termers, such as Dwight Eisenhower and Bill Clinton.

But when I recalculated the data in a way I consider more realistic, I found far less stunning—in fact, astoundingly modest—differences. By simply looking at expansion periods under each president, and backing out recessions, the 1.8 percentage-point gap shrivels to just four-tenths of a point. The unexciting inference: While the impact of government on the economy is quite real, Democratic or Republican labels seem to matter very little for economic growth. …

… BY BACKING OUT RECESSIONS, have I performed the methodological equivalent of rejiggering batting averages by eliminating strikeouts?

No. Recessions usually result from forces that have been building for years, long before the president to whom they are debited took office. The recession of 1974-75 partly stemmed from Keynesian policies started in the 1960s. The even-worse downturn of 1981-82 traces back to the 1960s and the Federal Reserve’s inflationary policies initiated in the early 1970s. The dot-com bubble-induced slump of 2001 dates to the 1990s, while the 2008-09 Great Recession traces its roots to government housing policies begun in the early 1990s.