Will 10% poll margins be enough for the Democrats?

It is always election season in the United States. Every second November sees either a Presidential election or important midterm elections that determine control of Congress and other political offices across the country. Currently, the Republicans run both houses of Congress, but the outcome in November 2006 seems in doubt given the unpopularity both of President Bush and Congress. If the Democrats win one or both houses, the implications for the Bush administration are serious.

The Republicans have controlled the House of Representatives since 1994, but never with a comfortable majority. The Democrats currently need a net pick-up of 15 seats in a 435-member House to win control. On the face of it, this should be easy. President Bush’s approval ratings are in the tank, more or less everywhere.

In ‘generic’ polls as to which party voters would prefer to see in control, double-digit majorities are choosing the Democrats. The Democrats seem highly motivated and have recruited some strong candidates. But despite all this, most commentators think the chances of a change in control are only evens or worse.

There are remarkably few truly competitive House districts, so gaining a net 15 out of perhaps 25 vulnerable Republican seats would be a strong showing. There are several reasons for this. One is gerrymandering. In several large states – Ohio, Texas and Pennsylvania – Republicans have drawn the boundaries for blatant party advantage. In others such as California there have been bipartisan ‘incumbent protection plans’ to minimise the number of marginal seats.

It is quite possible for the Democrats to poll more votes, as they have several times since 1994, but win fewer seats – they need a 5-7 per cent lead nationally to win control.

Winning the Senate, where the Republicans have 55-45 majority is perhaps even tougher. The Democrats need six net gains in order to win control. This is possible, but quite a stretch.

It would be surprising if the highly conservative (and homophobic) Senator Rick Santorum of Pennsylvania survived against Democratic challenger Bob Casey. Conrad Burns of Montana is also very endangered by links with corrupt lobbyists and his own foot-in-mouth tendency.

Three other Republican Senators are lagging in the polls at the moment – John Chafee in Rhode Island (easily the most anti-Bush state – the net Presidential approval rating is -54), Jim Talent in Missouri and Mike DeWine in Ohio. The Republicans have narrow leads in a number of other races, including Nevada, Tennessee and now Virginia, where James Webb, once Reagan’s Navy Secretary, has been gaining ground on tarnished Republican golden boy George Allen. There are currently no Democratic incumbents behind in the polls, although the most endangered seats seem to be in Maryland, New Jersey and Michigan.

The picture may be complicated a little by Independents – Bernie Sanders of Vermont would vote with the Democrats, while if Joe Lieberman defeats the official Democrat Ned Lamont in Connecticut his voting behaviour would be less predictable. It might all come down to which side of the aisle Lieberman chooses.

It is tempting to follow the conventional wisdom and look for modest Democrat gains in both houses in November, without control switching in either. The Iowa electronic markets give the Democrats about a 50 per cent chance for the House but only about 20 per cent for the Senate.

But in the past midterm elections have often seen a strong trend, with most of the close races ending up on the same side and a change in the political atmosphere. In 1986 the Democrats gained the Senate surprisingly easily; in 1994 few anticipated the size of the Republican sweep; in 1998 the Democrats’ resilience delivered a verdict on the attempted impeachment of President Clinton. In 2002 the midterms gave encouragement to the Bush administration.

The 2006 results could be a long way from the conventional wisdom on either side. So far, 2006 looks like a Democratic year, and brave punters may find that there is good value in a bet on the Republicans losing both houses.

Lewis Baston is research officer of the Electoral Reform Society and co-author
of several books on elections.