Data-driven journalism is a blessing. But one of its less salubrious side effects is the tendency on the part of some — mostly non-data journalists — to see everything can be that counted as meaningful.
It’s not.
{mosads}This brings us to the discussion about relative turnout in Republican and Democratic primaries to date.
Because the number of voters in each state can be counted and the numbers participating in each party’s primary can be compared, there is a tendency to see the comparisons as meaningful. As a result, I get questions almost daily about whether I am worried that more Republicans are voting than Democrats.
“Why should that worry me?” I ask. “Because,” comes the reply, “it says that Democrats are less enthusiastic about their candidate than are Republicans, which should translate into lower Democratic turnout in November.”
The problem with this assumption is that there is not a scintilla of evidence to back it up.
Check the aggregate national data first.
Since we started using primaries extensively to select candidates, more voters have turned up for Democratic primaries than for Republican contests in seven cycles.
Democrats went on to win the general election in just three of those seven years.
Republican primaries drew more participants in three of those years and won the general election in just one: 2000, when the GOP lost the popular vote.
One of the biggest landslides of recent years was Ronald Reagan’s 19-point victory in 1984. That year, almost three times as many Democrats voted in primaries as Republicans.
Reagan’s performance in 1980 was not quite as strong, but he still bested Jimmy Carter by a hefty 10-point margin, even though about 50 percent more voters turned out for Democratic primaries than for the GOP’s.
Maybe you’d argue to restrict the analysis to open seat years when a president is not seeking reelection.
In two of those four elections, the winning party had lower primary turnout. In one of them, 2000, the party with higher turnout won the presidency, while losing the popular vote, and in one year the party with the most primary votes went on to a clear win in the general election.
Even in key states there is no relationship between primary turnout and general election performance.
2000 was an open seat presidential election. In the Maine primary, Republican turnout was over 50 percent greater than Democratic turnout, yet Al Gore won the state.
That same year, more Democrats than Republicans participated in the Arizona primary, but George W. Bush went on to win the state in November.
In that year, the number of Democrats voting in the Arkansas primary was four times greater than the number of Republicans, yet, there too, Bush won the general election.
Political scientists Leonard Williams and Neil Wollman transformed these kinds of anecdotes into data, analyzing more than 300 state presidential primaries from 1972 to 2004, and found no relationship between a primary turnout advantage and winning the general election. Indeed, the party with the higher primary turnout lost the state more often than won it.
Linking the relative primary turnout of Democrats and Republicans to general election outcomes may sound plausible, but “sounding plausible” is not the same as being true.
And it’s not true — the evidence clearly indicates there is no linkage.
More plausibly, and more accurately, primary turnout results from locally and nationally competitive primary contests, while general election victories emerge from a host of other factors.
Indeed, the competitiveness of a primary in a particular state is one of the least important determinants of general election outcomes, while being one the most important determinants of primary turnout.
When thinking about November, pay no attention to those primary comparisons.
Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the minority leader of the Senate and the Democratic whip in the House.