Measuring the Strength of the Two Parties

Measuring the Strength of the Two Parties

By Sean Trende & David Byler - September 18, 2014

One of the frustrating things about analyzing United States elections is that we often proceed without any agreed-upon definitions of the terms were are discussing. Ask 10 analysts the following questions and you’ll get 10 different answers: What is a weak/weakening party? What is a “wave” election? What is a realignment?

In the past, answers have focused (unduly in our view) on the results of presidential elections with respect to the first two questions, and House races with respect to the last one. To be certain, the presidency is important. Yet, in the grand scheme of our federal structure, it is not so important as to warrant disregarding election results from other levels of government.

A party that controls the presidency but is weak everywhere else -- witness the Republican Party in the late 1960s, the late 1980s, or during the Ford administration -- will have real limits on its ability to change the country’s overall trajectory. Of course, such a party isn’t powerless by any stretch. But it is still meaningfully less able to influence policy than a party that controls the presidency and Congress, or the presidency and state governments.

Similarly, a party that possesses significant power in state government, but nowhere else, will nevertheless be in a better position to influence policy than a party that controls nothing.

To address this, we’ve developed an index to track overall strength between the Republican and Democratic parties. It’s still something of a work in progress, and we may revise it further over the next few months. For now, our goal is just to define it and apply it to help flesh out the three related issues above: Partisan strength, degree of electoral change, and realignment theory. We’ll deal with the first today. The second two will be dealt with in a subsequent column.

First, though, let’s briefly explain the index. We refer to five basic components to elected government in the U.S.: The presidency, House, Senate, governorships, and state legislatures.

From this, there are myriad ways that you could assess partisan strength in the country from top to bottom. We use what strikes us as the simplest approach: We look at all five segments of government, excluding third parties. For the presidency, we calculate the Republican share of the popular vote. For the Senate, we calculate the Republican share of seats. For governorships, we also calculate the Republican share. (Of course, because we exclude third parties, Democratic shares are implicitly calculated as well.)

For the House and for state legislatures, we did things a bit differently. For the former we took an average of the Republican share of the popular vote and the share of seats in the House won by Republicans. For state legislatures, we took an average of the share of state Houses controlled, share of state Senates controlled, share of House seats controlled, and share of state Senate seats controlled, since all of these measurements could have implications for partisan strength. All of these measures are calculated from 1946 through 2012 in order to gain a good picture of how current Republicans fare compared to their postwar counterparts. Again, we aren’t claiming this is the only way these things could be operationalized, but we do think it is a substantial improvement over the largely ad hoc discussions that seem to dominate analysis today.

Since each of the five measurements is expressed as a percentage with a theoretical maximum of 100 and a minimum of zero, we can readily add them together to create an overall index that runs on a scale of zero to 500. To help better process things mentally, we subtracted 250 from the score, which means that a government that was perfectly divided between the parties would have a score of zero. A positive score favors Republicans, while a negative score favors Democrats. With that in mind, let’s examine some different potential applications.

It’s commonplace to say that the Republican Party is in decline. When this claim is made, the explanation offered is usually some combination of the party’s supposedly limited appeal at the presidential level and demographic decline. There is certainly a case to be made at the presidential level, although the stronger case is that this is simply the effect of random noise.

But even assuming that the Republican Party is weakening at the presidential level, at the other levels of government, this is much less true, especially if one takes a long view of United States history. For example, despite some self-inflicted wounds in Senate races over the past few cycles, as of 2012 the GOP was still above median strength in the postwar period.

In the House, the GOP has the sixth best showing in the postwar period, despite a 2012 election that many described as disappointing.

What about governorships? After the 2012 elections, the GOP controlled a larger share of the nation’s governorships than at any point except after the 1994, 1968, 1998 or 1996 elections.

Finally, in the state legislatures, Republicans likewise find themselves in roughly as strong a position as they have been in living memory; only in 1946, 1952, 2002 or 2010 did they find themselves in stronger positions in the state legislatures.

But the best picture comes when we put things together in our “official” index:


Viewed in this light, the recent trend line favors the GOP, which remains above the historical average of -3.8, and well above the postwar average of -20.32.

To see how this might play out, let’s make some modest assumptions about 2014. Let’s consider four different scenarios: A bad GOP year, a mediocre GOP year, a good GOP year, and a great GOP year. Without getting too deep into the weeds, we’ve (somewhat arbitrarily) used the following definitions, which in turn create the index outputs found in the right column:

A “good” GOP year would result in an index of 15.9. This would represent the best score for the party since 1930. A great GOP year -- within the outside realm of possibilities (in terms of House seats, it would require a showing of R+9 on the generic ballot and 248 House seats), would leave the party with one of the best scores since Reconstruction.

But perhaps most importantly, even a pretty disappointing GOP year would result in an index that places them above their postwar average and leaves them, on balance, more powerful than the Democrats nationally. Contrast the resulting index with the party’s worst scores in the postwar era: -81 in 1964; -75 in 1976; -61 in 1974; -53 in 1960; -52.5 in 1966; -50 in 1978. Compare it even with 2006 (-14.7) and 2008 (-36.5). Whatever challenges are facing the Republican Party -- and they are real, if overstated -- this would not be a party in overall decline.

We’ll be looking at other applications for this metric over the next few weeks, and we’ll use it to measure the outcome of the election in November, whatever that might be.

David Byler is an elections analyst for RealClearPolitics. He can be reached at Follow him on Twitter @davidbylerRCP.

Sean Trende

Author Archive

David Byler

Author Archive

Follow Real Clear Politics

Latest On Twitter