This is the first part of what will likely be a three-part series examining the results of Congressional and Senate races in presidential-election years since 1960.
The purpose of this exercise is to get a quick and dirty analysis of how often, and to what extent, the Presidential ticket has a clear effect on downticket races.
The conclusion I have come to is that presidential coattails generally matter a good deal less than is perceived. Furthermore, even when there is a clear correlation between results in the Presidential election and in downticket races, this often has more to do with external circumstances affecting the party brand name as a whole, and less to do with the personal appeal of the presidential candidate.
To put it another way, I think that the relative strength of the party (or weakness of the opposing party) independent of the candidates, has a lot more to do with the viability of a presidential candidate than vice versa. Presidential candidates don't usually create "wave" elections; they ride them.
In order to do this analysis, I’ve gone through
-The electoral results in the last 16 Presidential elections, to determine exactly how many House and Senate seats were gained and lost by each party in each election.
-The success of each party in winning close Senate races in those years (I used a fairly broad definition of "close" as within eight points).
--The "popular vote" totals for House races-the percentage of voters nationwide who voted for a Republican or Democrat for Congress, vis-à-vis the percentage of voters who voted the same way for President, to see how closely they generally match.
Because there are relatively few gubernatorial races in Presidential election years, I’ve skipped those for the time being, but I may revisit those races later.
Before I get into the results and my conclusions, I want to state that there are any number of other factors which may affect this admittedly simple, quick and dirty analysis, that could serve to counter the conclusions I’ve made. A few of these are:
1) preexisting conditions affecting party brand name. If a situation should arise where a particular party is exceptionally fractured, like the Republican Party in 1964, or the Democratic Party in 1968 and 1980, that frequently affects the success of the party as a whole, regardless of the presidential candidates running. As I've said, I think that these preexisting conditions usually have more effect on the party's success in Congressional races than the candidate himself does. But it's also possible that a candidate's personal popularity can help congressional candidates overcome an adverse political environment, in ways that would not necessarily show up here.
2) I’ve determined what happened in Senate races which actually wound up being close, but I can’t really say which races, if any, would or would not have been close if there had not been the possible influence of the Presidential race. There is some evidence that sometimes, races which wouldn’t have otherwise been close wind up being surprisingly close perhaps due to the influence of the Presidential campaign, even if the candidate from the opposite party of the Presidential winner does win his or her race.
Jim Exon’s 1984 Senate race in Nebraska is one good example of this.
3) Most importantly, the absence of any kind of "alternative model". We don’t really have any way of knowing what the results would have been in Presidential or Congressional races if someone else had been heading the tickets in past elections. I can’t say for sure what would have happened in Congress in 1980 if John Connally and Ted Kennedy had been the presidential nominees, and it may well have been very different from what actually did occur.
That said, here’s what I’ve come up with.
My conclusion, as I said, is that coattails generally matter less than people think they do. It’s possible, on occasion, that in a major presidential landslide, presidential coattails can have a fairly clear effect...but even then, it’s not clear whether the positive downticket effects for the presidential candidate’s party have anything to do with the personality of the candidate or their specific appeal, or whether they are tied to external factors affecting the brand name of the entire party (which includes, of course, the presidential candidate).
In other words, it’s entirely possible, even probable, that the downticket Republican landslide in 1980 had less to do with Reagan's personal appeal, and more to do with a fractured Democratic party, the ascendancy of a conservative philosophy which had been incubated since 1964, and the conclusion of an extremely turbulent decade in American politics.
I will say that because the "coattails" phenomenon has not often been borne out in the past, there’s no reason why it couldn’t be valid in this election. That said, I don’t think that the coattail effect has been especially strong, especially often, in the past.
In this installment, I will look at the presidential election years from 1960 to 1972, which involved two very close elections and two blowouts of historic proportions, which seems a fairly good place to start.
1960:
Kennedy won the election by roughly 120,000 votes, but the Republicans gained one Senate seat and 20 House seats.
The 1960 presidential election was obviously very close, so the minimal change in the Senate is maybe not surprising. The Republicans’ 20-seat pickup in the House looks impressive, but I doubt it had a great deal to do with Kennedy himself. Democrats had had a terrific landslide in 1958, picking up 49 seats, and they had an inflated majority in the House at the time, holding over 60% of House seats even after the election (in which Comgressional Democrats earned a much, much smaller percentage of the actual vote, 54%).
That number, 54%, also indicates that Congressional Democrats significantly outperformed Kennedy in this election (which they didn’t always do; in 1964, for example, Johnson outperformed the national vote for Congressional Democrats by four points).
Close races: Montana and North Dakota (Democratic victory, open-seat), Iowa (Republican victory, open-seat), Idaho and South Dakota (incumbent Republican reelected), Michigan (incumbent Democrat reelected), Delaware (incumbent Democrat defeated).
Judging from the fairly even break in close elections, and in Senate races overall, and by Congressional Democrats outperforming Kennedy by five points, I don’t know that there were any especially strong positive or negative coattail effects in this election.
1964:
The 1964 presidential election was one of the biggest popular-vote landslides in modern American history, and the Republican party was as divided at this point in time as it has probably been in its history, with a bitter split between moderates (led by Nelson Rockefeller) and conservatives (led by Barry Goldwater).
Democrats gained 36 seats in the House, expanding their majority to 295-140, the largest majority enjoyed by either party in the postwar era (and frankly, probably a majority of unsustainable size). Despite the astronomical number of Democrats elected that year, Johnson actually outperformed Congressional Democrats in this election, as they won 57% of the popular vote to Johnson’s 61%.
The effect was much less pronounced in the Senate. We gained two seats, which doesn’t sound like much, but it gave us a majority of 67-33, larger than any majority we’ve had since. Republicans barely held on to the seat Goldwater vacated, in fact, winning the Arizona Senate race just 51-49.
On the other hand, Republicans did beat a Democratic incumbent in California, despite Goldwater losing badly in that state. And they came within 100 votes of ousting a Democratic incumbent in Nevada, a state Johnson also won, and came relatively close to winning two Senate seats in Tennessee (both of which they would take by 1970).
So perhaps the coattail effect wasn’t as pronounced as one might have thought.
Democrats won more of the close races (Nevada, New Mexico, Oklahoma, Ohio, both Tennessee seats, Wisconsin, Wyoming) than Republicans did (Vermont,
It’s difficult to deny that there were major Democratic successes downballot, and that probably had quite a bit to do with the national Republican party’s woes, sympathy for Democrats after Kennedy’s death, and Johnson’s successful campaign.
But I also don’t see this as an especially good predictor of what might happen in 2008, as I certainly don’t expect a 61-38 blowout in a presidential election.
In close races this year: Republicans won Arizona, California, Delaware, Hawaii, Pennsylvania and Vermont, while Democrats won Nevada, Ohio, Oklahoma, two Tennessee races, Wisconsin and Wyoming.
1968:
Richard Nixon won the presidential election very narrowly, and Republicans gained five seats in the House and five in the Senate.
The five-seat pickup in the House is pretty modest, and it’s about what you’d expect for a year in which the Presidential election is decided by only 500,000 votes.
Still, the Democratic margin in the Congressional popular vote was exceptionally small, with Democrats winning the vote total just 50% to 48%. Nixon and Humphrey each won roughly 43% of the vote in the presidential election, due to George Wallace’s candidacy.
This was an exceptionally turbulent year for the Democratic party, obviously, and for the nation as a whole; Between Kennedy’s assassination, the unpopular war in Vietnam, the controversial Democratic convention and the independent candidacy of the former Democrat Wallace, it was about as bad a set of circumstances as one could imagine. And yet, Democrats nearly won the presidential election, essentially treaded water in the House, and despite losing five Senate seats retained a 57-43 majority in the chamber.
In the close Senate races that year, Democrats won in Alaska (Mike Gravel, w00t!), California, Connecticut, Indiana, Iowa, and Missouri, while Republicans won in Illinois, Kentucky, Ohio, Oklahoma, Oregon, Pennsylvania and Utah.
I suppose there’s an argument to be made that the Republicans’ Senate successes could be tied to Nixon and Humphrey, but if Nixon led a Republican wave that year, I’m not sure why Republican gains in the House were so small.
And even if Democratic losses in 1968 can be tied to Humphrey, I don’t expect the political environment in 2008, for Democrats, to be as it was in 1968.
1972:
Nixon wins in a landslide, 61-38, over George McGovern. However, the Republicans gained a relatively modest 12 seats in the House, and Democrats gained two Senate seats.
Democrats actually earned a greater percentage of the vote in Congressional elections during the 1972 presidential wipeout than they did in the hotly contested 1968 race; they garnered 52% of the Congressional vote to 46% for Republicans.
Needless to say, that was a LOT better than McGovern did, 14 points better. I really don't think they would have done that much better with a stronger candidate atop the ticket.
Democrats won close Senate races in Colorado, Delaware (word, Joe Biden!), Georgia, Kentucky, Maine, Montana, and Rhode Island, while Republicans won in Idaho, Nebraska, New Mexico, North Carolina, Oklahoma, and Oregon.
As such, I have a difficult time seeing major coattails attached to Nixon’s candidacy, despite the magnitude of his victory over McGovern.
So, for these four elections-two close, two landslides-we have:
1960, where the winner's candidate lost seats in the House and Senate despite his vaunted personal charm.
1964, where the Republican Party had one of its worst electoral years ever, on all levels;
1968, where there may have been some positive correlation between Nixon's election and the Republicans' Senate successes, but that's hardly clear, and it wasn't especially strong if it was present; and
1972, where despite the Nixon landslide, Republicans lost seats in the Senate and made modest gains in the House.
Overall, I can't say I see a strong correlation between the Presidential ticket's popularity and their parties' success in downticket races for these elections. Next week, we'll look at 1976, 1980, 1984 and 1988.