You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Changes in the House: a Prediction

How Many Seats Will the Republicans Lose?

Fewer voters will go to the polls in two weeks than did in the last Presidential election. This decline will occasion no surprise. Presidents know that voters take midterm congressional elections far less seriously than Presidential elections; they also know the resulting threat that mid-term elections pose for them.

Lyndon Johnson presumed that the 38 Democratic congressmen elected in 1964 would be likely casualties in 1966. So be rammed through his Great Society program before the '66 election. The results were remarkable: save for the repeal of right-to-work, and DC home rule, all of Johnson's agenda was adopted, including landmark educational, employment, and civil rights packages, model cities, and economic development legislation. Johnson's intuition served him well. Midterm losses in 1966 were greater than anticipated: the Democrats lost 47 seats.

Unlike Johnson, Nixon has not had control of the House. With a Democratic majority of 243 to 192, it has overridden the President's education appropriations veto and threatened increases in other budget items. No wonder Nixon is joining Agnew on the hustings.

The source of Republican hopes in November is the Kevin Phillips thesis of the decaying of the Democratic Party. The New Deal coalition of the South, minority groups, laborers, and eggheads seems to be splintering, as hardhats and Southerners increasingly cast Republican votes. Those who predict a Republican gain in congressional strength thus assume a realignment of major proportions between the two great parties. In our opinion, this assumption is premature. There is little in the 1968 election that suggests an impending major realignment comparable to 1896 and 1932. The great issues—Vietnam, crime, race, and disorder—have unquestionably created instability in the electorate. Also, Nixon's total vote in '68 would have been much larger than 43 percent if Wallace bad not made the race. But there is little justification for comparing Nixon's victory to FDR's triumph in 1932. Without Wallace, Nixon's total vote would not have been appreciably greater than 50 percent and Republican gains in the House would not have been much larger than the five seat net gain. Though there are indications of change, there are no compelling reasons to suppose that these changes are at present of sufficient magnitude to create a Republican majority in the country. Likewise, there is no reason to expect Republican gains in the House this fall, much less Republican control.

There are, on the other hand, persuasive reasons to expect Republican losses. Among them are the historical pattern of gains in Presidential election years followed by losses in succeeding midterm elections (the surge and decline theory), the limited number of swing districts available to a Republican challenge, and the consequences of the business cycle for the party in power. The surge and decline argument for a Nixon loss is a straightforward extrapolation of the past. Only four times in this century —1908, 1916, 1956, and 1960—has the party winning the Presidency failed to increase its strength in the House; only once since the Civil War, 1934, has the Presidential party not lost House seats in the following midterm election. This recurring pattern occurs because the ballyhoo of Presidential elections draws to the polls voters who are greatly influenced by images of candidates, and who vote predominantly for a Presidential candidate and his party's congressional candidates. In mid-term, these marginal voters stay home, and congressmen they had elected two years earlier are sent home.

A first cousin to the surge and decline thesis places great emphasis on the number of incumbents each party has standing for reelection. Since swing districts have dwindled to less than 25 percent of the total, all save the newest members will probably return to protect the Democratic majority in the House.

The business cycle theory has a number of versions. The most common asserts that in bad times voters exact their retribution from the party holding the Presidency. An interesting twist on this theory is that bad times always hurt Republicans, because unemployment stimulates the turnout of Democratic, blue-collar workers. MIT economist Duncan Foley has determined that every one percent increase in the October unemployment rate since the previous election costs the Republicans fifteen seats in the House. Should the unemployment rate October, 1970 be, say, 4.5 percent (in fact, it is 5.5) Foley would predict a Republican loss of 15 seats. Just to hold their own, the Republicans would have to find issues appealing enough to offset the effect of unemployment - no mean task in a midterm election of marginal interest to most voters.

To arrive at a single prediction, we have taken into account the present party strength in the House, the unemployment rate, and the type of election - Presidential or midterm. Our analysis is based on a study of all House elections since 1916. Our prediction is this. Leaving aside for the moment the effects of political and economic cycles, the party in power can expect to retain 87 percent of its membership simply because of the electoral strength of its seasoned incumbents. Republicans therefore enter the race with a base of 168 seats. But the number of seats usually lost in midterm is not proportional to the number of seats won in the Presidential surge. Election results are more accurately predicted by ignoring the size of the surge and forecasting a constant midterm loss of 17 seats for the party in power. This midterm decline would reduce Republican seats to 151, a loss of 41.

The increase in the unemployment rate will exact additional seats from the Republicans. Our estimates are milder than those of Foley. We have found the price of unemployment for the party in power to average about four seats for every one percent increase in the rate since the previous election, depending on fluctuations in the unemployment rate before the election, the consistent rise in unemployment over the past eight months should cost the Republicans between four and eight seats - or a total of 45 to 49 seats.

The power of the Presidential office partially offsets these predicted losses and has in the past contributed an average of 13 seats to the President's congressional strength. Thus, we arrive at a final prediction — a loss of 32 to 36 Republican seats in November.

This might seem startlingly high. It is worth noting, however, that the average loss of the Presidential party in midterm elections since 1916 is 35 seats. Our predicted loss, moreover, is very close to that of Gallup. Gallup has argued that an average Presidential popularity rating yields an average midterm loss for the party in power. Since Nixon's current popularity rating is 56 percent, Gallup predicts a loss of 34 seats, the average midterm decline of' the party in power over the last six midterm elections.

Assuming the coming election conforms to the patterns of the past, the Republicans will fare badly. If they win seats, or simply avoid losing any, the 1970 election will be historically significant. Such a victory would be the strongest evidence yet that a new party hegemony is being established.

A final word. We have been talking of numerical, quantitative changes in party strength, not qualitative shifts - and those are important, though hard to measure. Old-line Democrats, North and South, are increasingly vulnerable. Since 1966, five Democratic committee chairman have been defeated in primary challenges. Among this year's primary casualties were Leonard Farbstein (NY), defeated by a women's rights crusader, Bella Abzug; Philip Philbin (Mass.), a hawk and second man on the House Armed Services Committee, defeated by peace candidate Robert Drinan; and George Fallon (Md.) defeated by peace candidate Paul Sarbanes. Power is slowly changing hands.

This article originally ran in the October 24, 1970, issue of the magazine.