by Ruy Teixeira
(cross-posted at www.washingtonmonthly.com)
Broadly speaking, there are two approaches to looking at the outlook for this year’s Congressional elections. One is the “macro” approach, where one looks at a variety of national indicators to gauge the mood of the electorate and how that’s likely to affect the incumbent and challenging parties. The other approach is the “micro” approach, which assesses how each individual House and Senate race is likely to turn out, and aggregates up from that level to assess the likely gains and losses of the two parties.
The two methods can tell different stories and, indeed, this spring that’s just what they did. The macro story suggested that the GOP was in terrible shape and likely to get swamped by the Democrats in November. Indeed, by these macro-indicators, as Charlie Cook pointed out at the time, the GOP was at least as badly off as the Democrats were at that point in the 1994 election cycle.
The micro story was different, however. Looking at individual races, it was hard to see where the Democrats could pick up enough seats to take back the House, while the Senate looked almost impossible.
But that was then. This is now and now the macro and micro data are aligning and pointing in the same direction: big trouble for the Republicans and a good chance that they could lose not only the House—which looks better than 50-50 at this point—but also the Senate.
Let’s review the relevant data, starting with the macro indicators.
Right Direction/Wrong Track
Right before 1994 election, the NBC News/Wall Street Journal (NBC/WSJ) poll had this critical indicator of the public mood at 55 percent wrong track/27 percent right direction. Today, the same poll has this indicator at 58 percent wrong track/29 percent right direction.
Generic Congressional Contest
In polls concluded this week, Democrats averaged a 14 point lead among registered voters in the generic congressional contest. Charles Franklin’s model-based trend estimate looks to be about a 12 point lead for the Democrats, judging from the chart on his site. Even assuming the generic question overestimates Democratic support by 5 points (Charlie Cook’s rule of thumb and the average difference between Gallup’s final poll among registered voters and the actual election result), that still gives the Democrats an average lead of 7-9 points.
The Democrats are also running even larger leads among independents in the generic Congressional ballot–typically 6-7 points higher than their overall lead. Thus, if the Democrats’ “true” overall lead is now about 8 points, then their true lead among independents is probably 14 or 15 points.
With that in mind, consider the following. As far back as I can get data (1982), the Democrats have never had a lead among independents larger than 4 points in an actual election, a level they managed to achieve in both 1986 and 1990. Indeed, since 1990, they’ve lost independents in every congressional election: by 14 points in 1994; by 4 points in 1998; and by 2 points in 2002. So, even leaving questions of relative partisan turnout aside (and I suspect the Democrats will do better, not worse, in this respect in 2006), the implications of a strong Democratic lead among independents in this year’s election, if it holds, are huge.
Generic congressional data also tend to show substantial shifts away from the GOP among a wide range of Republican-leaning groups, including some of their strongest base groups. For relevant data, see the October 3 Democracy Corps memo “Key Targets for November” and Friday’s Washington Post article, “GOP’s hold on Evangelicals Weakening”.
Presidential Job Approval
According to the Hotline, the average approval rating for Bush in this week’s polls is 38 percent approval/56 percent disapproval (Charles Franklin’s trend-based estimate has his current rating a couple of points higher). By comparison, Clinton’s average approval rating right before the 1994 election was 46 percent/45 percent.
Congressional Job Approval
The Hotline’s weekly poll average for Congressional job approval is now 28 percent, with 65 percent disapproval. Right before the 1994 election, Congress’ job approval stood at 24 percent (according to the NBC/WSJ poll). This indicator is not just bad for the incumbent GOP in general, but there are reasons to believe this is a key indicator of potentially large seat swings. As a Gallup report on Congressional job approval and the election notes:
During recent midterm election years, low congressional approval ratings have been associated with greater shifts in the partisan composition of the U.S. House of Representatives. In the five elections since 1974 in which Congress’ approval rating was below 40%, the average net change in U.S. House seats from one party to the other was 29. In the three midterm elections in which congressional approval ratings were above 40%, the average change was five seats….
The fact that both congressional and presidential approval ratings are low does not bode well for the Republican Party. The current situation is similar to the political environment in 1978 and 1994, when Democrats controlled both the legislative and executive branches — which were both unpopular. Those elections resulted in net losses for the Democratic Party of 11 and 53 seats, respectively.
Party Favorability and Preferences
According to a Gallup report based on data collected before the Foley scandal, Republicans are now running a considerable favorability deficit. The public rates them 42 percent favorable/53 percent unfavorable, compared to a 54 percent favorable/40 unfavorable rating for Democrats.
The latest Pew poll finds the Democrats preferred 55-27 on “more concerned with people like me”, 48-28 on “can bring about changes the country needs”, 44-34 on better managing the federal government and 41-27 on governing in a more honest and ethical way. And the public believes, by 41-27, that the GOP is more influenced by lobbyists and special interests.
On issues, the latest Ipsos-AP poll reports the following. Registered voters prefer Democrats over Republicans by 58-27 on health care, 53-31 on Social Security, 52-27 on gas prices, 51-36 on the economy, 50-37 on taxes, 48-38 on Iraq, 44-35 on same-sex marriage, 44-36 on immigration and 41-25 on political corruption. Most amazingly, Democrats are even preferred by 43-41 on terrorism and by 43-41 on protecting the US. (Note: the just-released Newsweek poll also finds the Democrats ahead—this time by 44 percent to 37 percent– on which party is trusted more to fight the war on terror.)
The Micro Situation
As these data suggest, there is precious little in the macro indicators that suggest anything other than a bad election for the GOP. But macro indicators don’t determine elections, voters in individual races do. And it is here that the big changes have taken place. In the spring, one could look race by race and it would be hard to see where the Democrats could make the necessary pickups to translate macro sentiment into a victorious election. But now you can.
While there is a lot of data available in a lot of different places on House races, Chris Bowers of MyDD provides a useful summary of competitive races tiered by likelihood of going Democratic and including the latest polling data, where available. This provides the raw material for thinking about how races might fall and lead to the net gain of 15 seats Democrats need to take back the House. The key thing to keep in mind is that the races near the top of Bowers’ chart appear highly likely to go Democratic (including, of course, a new entrant to this category, Mark Foley’s FL-16 seat). These races alone should take the Democrats within a handful of seats of retaking the House. After that, less probable races have to fall the Democrats’ way, but there are enough of these that average performance in these districts should put the Democrats over the top (i.e., if two races are 50-50 for the Democrats, those odds say that, on average, the Democrats should pick up one of these two seats). And historical experience suggests that in a “wave” election like this one, the party favored by the wave—the Democrats this year—may do far better than average in races that now appear 50-50.
Turning to the Senate, the math here is simpler. The Democrats must take all five of the most vulnerable GOP seats (Missouri, Montana, Ohio, Pennsylvania and Rhode Island) plus one of two other seats that have been considered less vulnerable but still competitive (Tennessee and Virginia) plus lose none of their own. Alternatively, they could lose one of their own (the obvious candidate here is New Jersey), but then they’d have to win all seven of the GOP seats just mentioned. This is a tall order and last spring it seemed virtually impossible; it was not clear how strong Democrats would be in the top five races and the Tennessee and Virginia races looked like real outside shots.
Now things look different. The Democrats must still run the table in the manner described but the outside shots now look quite plausible and chances in the top five look good to very good. Here are the last-five-poll averages from the very useful site, Pollster.com, run by Mark Blumenthal and Charles Franklin: Missouri (McCaskill-Talent), 44-45; Montana (Tester-Burns), 49-43; Ohio (Brown-DeWine), 45-42; Pennsylvania (Casey-Santorum), 49-39; Rhode Island (Whitehouse-Chafee), 44-40; Tennessee (Ford-Corker), 45-43; and Virginia (Webb-Allen), 42-48. In addition, in New Jersey (far and away the Republicans’ best chance for a seat pickup), Democrat Bob Menendez now leads Republican Tom Kean by 43-41. So, in a wave election, all the raw materials are there for these seats to all or almost all break in the Democrats’ direction– an outcome with plenty of historical precedent—leading to a switch in control of the Senate. That doesn’t mean it will happen (chances still look poorer than in the House for a switch of control and are probably less than 50-50) but it easily could happen, something most observers would not have said earlier in the year.
What Lies Ahead
One month ‘til election day! What we would all like to know, of course, is whether this situation is liable to get better, get worse or stay the same for the GOP. On the stay the same or get worse side of the argument, start with the fact that this is a heavily nationalized election, which is a big disadvantage for an unpopular incumbent administration and Congress. To cite some representative data, the latest Pew poll found voters saying national issues, rather than local issues, were most important to their vote by a 51-23 margin. And 39 percent said they are thinking of their vote for Congress as a vote against President Bush. Analogous figures going back to 1982 show that this level of anti-president voting has never been surpassed—indeed, there are no figures before 2006 that are even close.
The Foley scandal should, if nothing else, keep the spotlight shining on the failures of the Bush administration and GOP Congress. Changing the subject back to local issues, already difficult, has just become even harder.
But it could be much worse than that. Two of my favorite political observers, Charlie Cook and Chuck Todd (editor of the Hotline), termed it respectively a possible “inflection point” or “tipping point” in the campaign, creating serious momentum toward the Democrats as we move toward election day. Already, we know that almost everybody (78 percent in the latest Time poll) has heard of the Foley scandal and that they strongly believe a GOP cover-up is going on (64-16 in the same poll).
But it could take awhile for these effects to be fully felt. In the Pew poll, which concluded on October 4, there was no difference in the Democratic lead (13 points) in the generic Congressional contest before and after the Foley news broke. A more extensive review of recent data by Mark Blumenthal also finds no recent change.
On the other hand, an October 7 story in The New York Times suggests that the Foley scandal is already tipping some races where corruption or related issues have been important in the Democrats’ direction. And the just-released Newsweek poll does have Bush’s approval rating down to 33 percent, a new low in that poll. So we shall see.
But the biggest problem for the GOP remains Iraq. Even before the Foley scandal broke, the string of Iraq-related bad news and revelations (the loss of Anbar province in Iraq, the NIE conclusion that the Iraq war has made the war on terrorism harder, the Woodward and Powell books and their documentation of Bush administration failures) had halted some modest momentum in the GOP’s direction. Now Iraq is increasing in importance to voters’ Congressional vote intentions—and is clearly the top voting issue—even as pessimism on Iraq deepens. In the new Newsweek poll, 64 percent believe the US is losing ground in Iraq and 66 percent say the war in Iraq has not made the country safer from terrorism.
On the get better for the GOP side of the argument, there are limited possibilities. One, of course, is some unforeseen event that allows the GOP to change the subject. Not much one can say about this other than it could possibly happen.
Then there is the vaunted GOP turnout machine (but polls have generally shown Democrats more enthusiastic about voting this year and the Foley scandal seems likely to have a further negative effect on GOP voting enthusiasm) and their ability to spend a lot of money in the last days of the campaign. This may be their last and only hope of avoiding a very bad election. The Democrats, however, will not be standing idly by while the GOP tries to muscle their way out of bad situation, so it should be a very interesting last several weeks.
State of the Race: The Macro and the Micro
by Ruy Teixeira