Despite Daily Kos suspending polling earlier this month, the generic congressional ballot picture remains unchanged. Two weeks ago, Republicans led by a simple mean of 0.9%, and currently lead by a simple mean of 1.0%. Here are the polls used in the calculation:
Once again, the numbers stay in the narrow range of Dems +0.7%, to GOP +1.7%, that has been the state of play since mid-January.
While the removal of Daily Kos polls made no difference to the overall result, there is one poll which, if removed from the average, would make a big difference. Removing Rasmussen polls from the averages produces a pro-Democratic shift of 1.9%
Generic Congressional Ballot, June 29th: GOP +0.9% No Rasmussen polls included
By consistently outlying in favor of Republicans, Rasmussen skews polling averages unlike any other. I still don't have a good solution for this as an election forecaster, but it remains worth noting nonetheless.
Or not. The appropriate way to react to polls like these is, as always, to look at the broad context of all polling. Here is why:
There is no One, True Poll There is a tendency among people who follow elections to look for The One, True Pollster that is more accurate than any others. This is folly. Every poll has error. No poll that is always right, nor will there ever be a poll that is always right. The goal it to minimize the error as much as possible.
Polling averages more accurate than any poll: Simple polling averages of all available polls have less than 50% the error of individual polls. Across 144 statewiode, general election campaigns from 2004 to 2010, simple polling averages have produced a total error of just under 1.8% from the final projected margin to the final result. By way of comparison, the average total error from individual polls to the final result was just over 3.6% for the same time period (the NCPP report linked in the previous sentence halves the total error, and refers to it as "candidate error."). No pollster is ever going to beat the average of all other polls by 50% across that many campaigns. Poll averaging works.
The broad view of generic congressional ballot polling shows a toss-up. Here are the 24 generic congressional ballot polls conducted over thepast month and released to public (Zogby Internet polls excluded). When they are sorted by margin, the tightness of the 2010 campaign is clearly revealed:
Rasmussen's pro-GOP outliers aside, this is an extremely tight campaign. Of the 20 non-Rasmussen polls, 11 show the campaign either in a dead heat or with one party up by a single point. Only 4 of the 20 non-Rasmussen polls show a margin of more than 3%. That is a pure toss-up.
Little has changed in five months: Since mid-January, the 30-day, simple mean polling average in the generic congressional ballot has shown little variance. The high point for Democrats was a lead of 0.7%, and the high point for Republicans was a lead of 1.7%. The current Republican lead of 0.9% is thus not only not cause for panic, but it is hardly newsworthy at all. The electoral environment remaains in a five-month period of near statis.
I understand entirely that writing blog posts about polls is both tempting and easy. Not only are polls news, but there is an attractive, empirical side to them that can elide the oft-annoying spin tsunami of the political news cycle. However, when looking at polls, it is always best to look at the entire picture. While doing so makes for a less dramatic view of the electoral environment, it does make for a more accurate one.
This week's update on the national generic ballot has some good news and some bad news for Democrats.
The good news is that the Gallup poll showing Republicans ahead by 6% was indeed an outlier based on polling over a holiday weekend. The 6% lead Gallup showed for Republicans last week has been entirely erased, with the poll reverting to a dead heat.
The bad news for Democrats is that Republicans are actually slightly improving anyway. This week they have regained the overall lead, moving ahead by 0.3% after trailing by 0.1% last week:
In reality, there has been little to no change in the generic ballot over the past five months. Since mid-January, the average has consistently hovered within a narrow range of Dems +0.7% and GOP +1.7%. The median error of the forecasting methodology I use is 1.76%, so a 2.4% variation range over such a long timeframe does not mean that much. Every time the average swings toward one party, it tends to swing right back to the other. Pollster.com shows this trend pretty well, even if we have slightly different numbers:
Neither party has broken out of this pattern for a while, leaving control of the House of Representatives a toss-up. However, if either party were to move ahead by 2% or more, that would be a big deal.
June 2nd generic ballot: Democrats +0.13% Last week: Democrats +0.46%
There is going to be a lot of news today about the llatest Gallup generic congressional ballot poll showing Republicans with their largest lead ever in that poll, 49%--43%. While that certainly is not good news for Democrats, it cannot be emphasized enough that this poll is still just one of twenty-four generic congressional ballot polls to be released this month. In the overall average of those 24 polls, Democrats still hold a slight edge, 44.17%--44.04%:
(Note: Zogby interactive also released generic congressional ballot polls during the past month, but they are not included in these averages due to their horrendous track record)
Some might object to this poll averaging methodology, arguing that including all of the polls over the past month, and also including multiple polls from individual polling firms, will miss developing trends. However, not including multiple polls from the same polling firm, and including only the most recent polls, has been shown to produce less accurate poll averaging results (I would know, since I used that type of less accurate methodology in 2008, and conducted research to find a more accurate method for 2010). To put it a different way, removing older polls and multiple polls from individual pollsters produces results in a less accurate poll methodology that often finds trends which do not exist.
The methodology I am using in 2010 has found minimal movement--less than 2.3% on net--in the national generic ballot over the past fours months. This is a lot more believable than the idea that the country is swinging 5-6% points in favor of one party in a single week, even if it isn't uncommon for individual polls to show such a trend (Quinnipiac showed such a result for Democrats last week, for example). The entire difference between the 2004 and 2008 presidential elections was less than a 10% net swing--how could one week of news early in the campaign season possibly result in half of the net swing from 2004 to 2008? A swing like that requires a mega-story like a presidential convention or the financial metldown of September 2008, not run of the mill political tit for tat during the spring legislative and primary season.
Shorter version of all this: one generic congressional ballot poll means very little. As always, look at the complete polling picture.
Generic Congressional Ballot, May 24th: Democrats +0.5% Polls from here and here
This is good news for Democrats, but also only represents a swing of only 2.0% in their favor from the largest Republican advantage in my averages. This is really is only a slight improvement, not a dramatic one.
It is also worth noting that Rasmussen generates an overall swing of 1.41% in favor of Republicans. Without Rasmussen, Democrats lead by 1.95% in my average, pushing the average to Dem +0.54% from Dem +1.95%. Pollster.com shows a 0.8% swing in favor of Democrats without Rasmussen.
I don't really know what to do with Rasmussen polling at this point. Rasmussen is clearly outlying from other polling, which is a big problem given that it produces a huge percentage of polling public polling on Senate, Governor and House campaigns. One thing I am not going to do is tweak Rasmussen polling, by leaving it in the averages but adding 6% (or so) to the Democratic candidate. Either I will continue to include Rasmussen in the averages with no tweaking, or it will just be dumped from the averages altogether. The best solution is probably just to produce polling averages both with and without Rasmussen numbers, with the former still receiving the topline credit.
Perhaps more notably, at Gallup's 2010 election preview event today (follow it on Twitter at #gallup2010), Jeff Jones produced a chart that projected seats in the House of Representatives based on the Generic Congressional Ballot. According to Jones, if Democrats received 50% of the two-party vote, as the current average projects, they would win 234 seats and maintain control of the House of Representatives:
Projecting seat totals based on Generic Congressional Ballot
This model from Jones suggests that a swing of 7% in the national two-party Congressional vote could result in a net swing of 64 seats in the House of Representatives. While that seems extremely wide for only a 7% shift, from 2004 to 2008 the Demcoratic share of the national two-party Congressional vote swung almost precisely 7% (from 48.6% to 55.6%), and Democrats gained a net of 54 seats over those two years. So, perhaps Jones is exaggerating a bit, but not by much.
What this means is that even small shifts in national vote preference can make a huge difference in the final result. If Democrats were to improve by 3%, they could keep their majority at close to its current size. By contrast, if Republicans were to improve by 3%, they would be able to take control of the chamber. At a rate of eight seats for every 1% shift, there is a lot on the table even with small fluctuations.
My own numbers don't show quite as dramatic a swing, but Democrats have still gained 1% over the last month (from down 1.3% to down 0.3%), with more likely to come by the end of the week:
The average since April 22nd has Democrats up 0.56%, gaining almost another 1%.
There have been rumored Democratic comebacks before, all of which quickly dissipated. However, this recent rise is the largest Democratic net gain since the September improvement of last year. There are also good reasons to think this improvement will stick this time, as the economy continues slowly to improve (even though it is still nowhere near full recovery, there are indications that the relative direction of the economy is has more impact on electoral outcomes than the its absolute condition.)
At least for 2010, Democrats have likely already hit their low point, and won't see the electoral environment become any worse than its current state. As such, Democrats should maintain a narrow advantage in the generic Congressional ballot from here on out, and thus keep narrow control of the House after 2010. A prediction like that might make me look foolish in November, or even in just a few weeks, but it is fun to go out on a limb every once and a while.
There are indeed glimmers of hope out there, but the broad indicators do not yet show enough movement toward Democrats to suggest a trend. Over the past month, Democrats have gained about 2% net on Republicans in the national generic congressional ballot, while President Obama has gained about 3% net in job approval. Here are the six month trends from Pollster.com, removing Zogby interactive polls from the mix:
Six month trend, National Generic Congressional Ballot
Six month trend, President Obama job approval
The electoral environment has been pretty stagnant over the past six months. This recent 2-3% improvement for Democrats could be the start of an actually upward trend of Dems, or it could be yet another small, possibly random movement well within the margin of error for polling averages (which is, according to my research, 1.8% total error at 50% confidence, 6.0% total error at 95% confidence).
Even if the electoral environment is static (and it probably is), it does at least feel like there is more to be optimistic about as a Democrat lately. The Republican rise has at last been stopped, the economy probably is going to keep slowly improving, primary challengers are making things interesting for both parties, and legislation, however watered down, is starting to pass. It all lacks the epic feel of the 2008 elections, or of the health insurance reform fight, but it also lacks the apocalyptic feel of the special election for Massachusetts Senate. It seems likely that Democrats have already hit rock bottom, and 2010 will only keep getting better from here on out.
The economy grew 3.2% in the first quarter of the year, the Commerce Department reported Friday, another indication a steady, though modest, recovery has taken hold.
The annualized rate of growth of the gross domestic product -- the nation's total production of goods and services -- was down from the 5.6% rate of the last three months of 2009. But that had been expected as the effect of the federal government's stimulus policies peaked during that period.
"We're still running on the fumes of stimulus in the U.S. economy," said Diane Swonk, chief economist at Mesirow Financial in Chicago. "It's a recovery, but by any standard is still a muted recovery. But we're thankful to have what we've got," given the depth of the recession, she said.
While this seems like decent growth,, it is not helping Democrats. Looking at the two broadest measures of the electoral environment right now, Democrats have actually dropped a net of about 7%:
National House Vote: When I first introduced the national House Ballot monitor on October 29th of last year, Democrats led by 5.3%. Last week, using the exact same methodology, Republicans were ahead by 1.4% (I have been unable to update the average this week due to travel, but it would not have been much different).
President Obama job approval: According to Pollster.com, President Obama's job approval is currently net positive by 0.4%. At the beginning of October, when this period of growth began, his approval rating was a net positive of roughly 7.5%.
If an opposition party is gaining about 1% a month on the governing party during a period of economic expansion, as is currently happening in the United States, that economic expansion is clearly not benefiting all that many people. Whether it is due to continued foreclosures, high unemployment, stagnant real hourly wages, a loss of retirement savings, or all of the above, most people do not feel their economic situation is getting better. The polling situation is all the evidence needed to demonstrate this: if the majority did feel like the economy was really recovering, then Democrats would be recovering, too.
If Democrats do get whitewashed in 2010, as is seeming increasingly likely, it will be because they failed to to improve the objective economic conditions facing the majority of the country. This is not an area of politics where spin plays any meaningful role. People don't need talking heads, or blogs, or press releases, or even Presidents to tell them if their personal economic situation is improving or not. If anything, politicians who claim that the economy is improving when most people feel it is getting worse just look out of touch, ala John McCain claiming that the fundamentals of the economy were strong the day of financial crisis.
The job of a governing party is not to tell people that things are getting better, but to actually make them better. Right now, democrats do not seem to be getting the job done.
The House of Representatives passed health reform one month ago today. At that time, Democrats led the National House Ballot 41.87%--41.65%. Now, one month later, with an entirely new set of polls, all of which have the majority of their interviews conducted post-hcr passage, Republicans lead 44.92--43.44%.
This represents a net 1.7% shift toward Republicans since the passage of health reform. That is a modest, though not insignificant, improvement.
While Democrats did not receive a lasting post-health reform boost in polling, I am sympathetic to the argument that Republicans would have gained even more if health reform had not passed. Admittedly, that argument is based entirely on hypotheticals and speculation, and cannot be proven.
Perhaps even more striking than the net Republican gain is the decline in undecideds. Both Democrats and Republicans have actually seen increases in their raw support--Republicans just gained more. Overall, the number of "undecided" or "other" voters declined by 4.84% over the past month, or nearly 30% of all "undecided" and "other" voters. Not unlike a Presidential election, the health care reform fight pushed people into one camp or the other. It made people pick sides.
Democrat Tom Deutch won easily last night in the special election to replace Robert Wexler in Florida's 19th congressional district. However, the enormous margin is not actually rosy news for Democrats nationwide. Here is why:
Florida 19 Cook PVI: D + 14.8 This is a very blue district. Given the results of the 2004 and 2008 elections, Democrats would be expected to win this seat by 29.6% (twice the PVI) if the national margin was precisely even.
Florida 19 Result: D +26.9 Duetch won last night by 26.9%, or 62.1%--35.2%. However, Kerry won this district 66%--33%, and Obama won 65%--34%. So, Duetch slightly underperformed Democratic presidential candidates in this district.
Perhaps the district is trending redder, as Obama underperforming Kerry would suggest. Perhaps this is still a good result for an open seat campaign. Or, perhaps this suggests that Republicans have a slight lead in the National House Ballot.
I am going with the latter. This is because my latest numbers on the National House Ballot, updated today, suggest that Republicans have taken a small lead:
Even if all of the various "questionable" polls are removed (Economist because it is Internet, Daily Kos because it isn't exactly a generic ballot, plus all of the Republican leaning polls), the result doesn't change at all. Republicans still lead by 1.0%:
Tonight, and for the next month, the National House Ballot goes international. Until the election on May 6th, I will feature the latest polling on the UK House of Commons alongside the national ballot polling here in the United States.
Even though the Tories are favored to win a plurality, Labour and the Liberal Democrats combined still hold a commanding edge on the Tories. Were that the center-left still clearly ahead in the United States as well.
Unless there is a noticeable shift in the share of the vote for Conservatives (either up or down), it's going to be anarchy coalition forming time in the UK. The margin is close enough that a Lib-Lab coalition would likely need the left-wing Respect, and / or some combination of the various nationalist parties, in order to form a government.
Fourth and fifth parties are really on the rise in the UK, receiving a record 9.3% of the vote in 2005. They might form a part, however small, of Her Majesty's Government in 2010.
The margin in this campaign has not shifted more than 1.5% in either direction in four months. That is remarkable stability compared to the epic pitched battle over health reform, which saw several wild swings of momentum since early December. Once again, that makes me think that health care will not be a major factor in this election.
With the arrival of new generic ballot polls from CNN, Gallup and Rasmussen, the National House Ballot is now almost precisely tied.
One bright spot for Democrats is that the median of these polls is more favorable than the mean. Fourteen of the twenty-six polls show Democrats ahead, by a median of 1.5%. The disastrous results for Democrats in Rasmussen polls throw off the mean a bit. The mode actually shows Democrats ahead by 3, with a Republican lead of 3 being the second most common result. A numerologist might make something of that.
The other good news for Democrats is that, even though virtually all polls have switched over the registered voters from "all adults," Democrats have not lost any ground in two months. We might be at the low point of Democratic electoral fortunes right now, with an improving employment situation ready to start sending it upward. Then again, switchover to likely voters in September could send Democratic chances down even further. Much of that depends on if the enthusiasm gap narrows or not.
Gallup is now releasing a regular tracking poll every Tuesday, just like Rasmussen. The other two weekly tracking polls, Daily Kos and the Economist usually release late Thursday or early Friday. As such, the National House ballot will now be updated on the front page twice a week, on Tuesday and Friday afternoon.
A seat-by-seat House forecast will be updated every Tuesday and Friday, along with the National House ballot, starting on April 20th.
In the first week following the completion of this round of health reform legislation, Democrats are holding steady in the National House Ballot. Across the four tracking polls--Daily Kos, Economist, Gallup and Rasmussen--Democrats have actually improved slightly. The previous four polls from those organizations showed Democrats trailing by 1.25%, but the most recent polls from each of those organizations show them ahead by 1.25%.
The continuing Democratic advantage is significant since polls of "all adults" have almost entirely washed out of the system. CNN, Daily Kos, and the Economist are among the many organizations that have moved away from polls of all adults in favor of polls of registered voters. In theory, such a polling shift should have been beneficial to Republicans, since Republican voting groups have a higher voter registration rate than Democratic voting groups.
Republicans are probably ahead right now in the National House Ballot among the people who will actually turnout to vote in 2010. However, it is a narrow lead, and there are still over seven months until the election.
After the House passed the Senate health insurance reform bill last night, the National House Ballot stands almost exactly where it did at the beginning of the month. The Democratic advantage is 0.2%, compared to 0.3% 18 days ago.
There are a lot of varying opinions on whether passing this bill will be good, politically speaking, for Democrats. I say we just measure where the political situation stands today, and compare it to the political situation over the next month.
It won't be a perfect test, as there are other factors in these polls, and in the national political scene, than just health reform. Additionally, while we have passed the climax, this particular health reform legislative fight isn't quite over yet, either.
Still, right now, Democrats are narrowly ahead. Let's see if that holds.