by
Renard Sexton
@
6:00 AM
Over the last few weeks, we've had multiple discussions about the numbers games in the UK general election. Whether it is looking at
swing voters,
redistricting or
proposed changes to the electoral system, the goal has been to examine the relevance and strength of various factors on the optics and outcome of the upcoming election.
With David Cameron and Gordon Brown alternatively on the offensive -- bullying; class war; he's ruined the economy and the budget; the NHS and granny will both die -- and Nick Clegg and the "fourth" parties trying desperately to make the case for wholesale change, personality and polling have made for a dynamic series of news cycles.
This is not lost on the major pollsters in the UK, upon whose work a significant amount of media back-and-forth is written.
Each time a new poll comes out, as it is often in the US context as well, the key reported information is the top-line party and candidate numbers and how they have changed since previous polls done by the given newspaper/TV station and pollster pair. The poll release formula then dictates that any minute changes be cast in quasi-apocalyptic terms for the party(ies) or candidate(s) on the decline. Finally, talking points from the involved candidates and parties are mentioned, usually with the implication that some recent event caused the shift in numbers.
As the election comes to a close, however, it is a slightly different barometer (rather than newsworthiness) that occupies a pollster's mind. In fact, throughout the election campaign, horserace polling is in some ways simply a series of test-runs that calibrate for the real test: the final numbers. Get them right and you reign supreme until next election; get them wrong and travail in obscurity for the foreseeable future.
Hyperbole aside, the truth is that the economics of being a pollster today dictate an approach not too dissimilar from the one above. Nick Moon of the pollster GfK-NOP explained to me that with "very little money," the political polling industry in Britain today is "very cut-throat." When I spoke with Andrew Cooper of Populus (another pollster) he concurred, adding that "with low entry cost in new internet polling and quite small number of viable clients, there is consolidation bound to happen."
Peter Kellner of the very prolific online pollster YouGov went a step further in explaining how the business model works for them. "YouGov does polling on lots of subject, most of them not politics," he said. With numerous offices across Europe, North America and the Middle East, YouGov brands itself as a "market research agency," with a 250 thousand strong internet panel who can answer just about any question a client may like to ask.
This is where the pollsters really make their money, not on quadrennial sets of dense national polling of party politics, but day-in-day-out market research for companies, governments, political and economic organizations or think tanks.
This has an important impact on the way national polling is done in British politics. "Getting it right" for the major British election pollsters is measured by how close to the national popular vote share numbers you get, looking at the three large parties and an aggregate of all the "fourth parties" in another category.
After the 2005 election, the British polling council released the following regarding the accuracy of the top five UK pollsters:
Unfortunately, as we have looked at repeatedly, the popular vote shares listed tell us only a fraction of the story. Only the Gfk-NOP exit polling on election day itself included an actual seat projection (which, by the way, they projected correctly -- admittedly said Nick Moon of GfkNOP because "two minor biases canceled each other out"). For other election day projections, it was up to John Curtice of the University of Strathclyde and his team at the BBC, rather than the pollsters themselves.
In an election where it is likely to come down to the wire, of course, the national popular vote fraction of the story is simply not enough. Even the inclusion of occasional broad polls of marginal seats, which according to Nick Moon "don't get included in the poll-of-polls" calculations by most media outlet, will not explain the whole story.
But at the end of the day, this is not the point of media-based national polling, says Peter Kellner of YouGov. "We don't include undecided or refusals because we ask how people would vote if the election were held tomorrow," he explained. These polls are intended to show snapshot of the race, he continued, "not the final winner."
Of course, in political polling or otherwise, it is pretty easy to win when you set your own rules.
---
Renard Sexton is FiveThirtyEight's international affairs columnist and is based in Geneva, Switzerland. He can be contacted at sexton538@gmail.com
There's More...
Contract Post
by
Tom Schaller
@
10:50 AM
The March labor numbers are
out from the Bureau of Labor Statistics and, at first glance, they look pretty good: 162,000 net new (nonfarm) jobs, and the national unemployment rate holding steady at 9.7 percent. 538's Hale Stewart will have more on this subject in the near future, but for today let's just set out the good and bad news in the latest jobs data.
The good news, of course, is that the economy added
any jobs. Against the backdrop of the pattern during the past two years, which is depicted above and resembles an inverted bell curve during the year leading up to Obama's inauguration and the year afer his inauguration, 162K new jobs is generally encouraging. And those 162K jobs added in March were the most added since
three years ago, in March 2007. (This month's and last month's data are still preliminary and as-yet uncorrected, so the 162K could be adjusted.) And the 9.7 unemployment holding steady, as it did for the third month in a row, is at least not bad news.
Of course, the economy needs to add about 150,000 jobs just to keep up with growth of the working-age population, which is why 162K is basically a break-even figure. Moreover, about a fourth of the new jobs are temporary, many of them courtesy of the federal government's hiring of Census-takers.
The worse news is the share of unemployed persons who are long-term unemployed, defined as having been unemployed for more than six months (27+ weeks). That percent looked like it had capped out in January 2010 at 41.2 percent, dipped slightly in February to 40.9 percent, but jumped in March to 44.1 percent. These figures are all seasonally adjusted, so they account for payroll drop-offs that occur naturally at the end of summer or as the drop in temperature forces certain types of employment to cease for winter. By comparison, the seasonally-adjusted long-term unemployment share a year ago, in March 2009, was 24.6 percent. (For details, see Table A-12, p. 23, here.)
Meanwhile, involuntary part-time workers jumped from 8.9 million to 9.1 million, and the unemployment rate for African Americans continued to rise. "The overall picture here is not very positive," writes liberal economist Dean Baker. "While we are seeing job growth it is at a very slow pace. The average in the private sector over the last two months was just 66,000 jobs. With the public sector shedding jobs, we will need more rapid job growth just to stay even with the growth in the labor force."
Not surprisingly, House Minority Leader John Boehner pounced, issuing a statement saying, "A 9.7 percent unemployment rate is no cause for celebration, and any politician who takes a victory lap for it is out of touch with the struggles working families and small businesses asking ‘where are the jobs?’ are facing." And, not surprisingly, he pinned the blame on what he believes is a failed stimulus bill. "Today’s private-sector job gains are encouraging but not nearly what President Obama promised when he signed the trillion-dollar ‘stimulus’ into law last year with promises it would keep unemployment below eight percent and create jobs ‘immediately.’ Our economy has lost more than three million jobs since then and unemployment remains near ten percent."
There's More...
Contract Post
by
Nate Silver
@
7:35 PM
Veronique de Rugy has issued a fairly gracious
response to my
critique of her study on the disbursement of stimulus funds, the crux of which was that she had failed to account for a variable (the presence of a state capital) that was extremely important in predicting the allocation of stimulus funds (because much of the money is intermediated by state governments).
Most importantly, she has promised to evaluate some of my concerns and to re-run her analysis. This is terrific -- and she is to be commended for her responsiveness. de Rugy is also to be commended for having released portions of her dataset** on the Mercauts Center website (something which she had done originally). Nevertheless, some further comment on her response -- and the issues in research design that her study raised -- is warranted:
-- I share de Rugy's disappointment with the quality of the data available at recovery.gov. Frankly, I am not sure that testing her hypothesis to a peer-reviewable level of robustness is possible given the middling quality of data and the inherent ambiguity with how particular projects must be assigned to particular congressional districts.
-- de Rugy writes: "The unemployment data for the regressions has in fact been used by congressional district, not by MSA. The confusion comes from the fact that the Excel file on the website includes unemployment by MSA." Good: that particular issue is cleared up, as well as the reason for my confusion.
-- For me, personally, the notion that the allocation of stimulus funds could have reflected a broad-based and widespread effort to benefit districts represented by Democrats seems implausible -- something which is well worth examining but something which should have received especially rigorous scrutiny. This is particularly so given that many of the funds were intermediated by state governments, not all of which are controlled by Democrats, as well as federal agencies that were constrained by formula rules.
There are two other variations that I find less impluasible:
I find it less impausible that the funds could have been directed toward those sorts of districts which tend to vote Democratic (e.g. as measured by PVI or by Obama vote share) -- even after controlling for other demographic variabes -- a possibility that de Rugy raises in her response but which was not the focus of her hypothesis. The difference is that that this could have resulted from a sort of unconscious bias in the design of the stimulus rather than a deliberate conspiracy.
I also find it less implausible that some *particular* projects could have been directed toward those districts that had a Democratic representative who was either especially influential or who a key swing vote in the House. (This is what we call pork.) However, de Rugy ran various tests on the types of Democratic districts that benefited from the stimulus and did not find any relationships with the characteristics of the Democratic members of Congress that tended to represent them.
-- There are a few passages in the response where de Rugy is still taking her initial results a bit too credulously, such as:
"So even after I use his methodology I will find that Democratic districts, other than state capital ones, are getting 30 percent more than Republican ones. That does seem like a possible political bias to me, which would be worth looking into.
How much of a bias? I don’t know. Let’s not forget that my take on the data has always been the following: The regression analysis shows that district’s party representation matters. However, I cannot say how much it matters compared to other factors (such as the formula used by different agencies).
Her results suggest correlation -- but until confounding variables are rigorously tested and rooted out, they are no proof of casuation at *any* order of magnitude. (It is emphatically not just a question of how much bias there was.)
-- de Rugy seems more concerned than I am about my four word-aside ("and possibly deliberately biased") that raises the question of whether her research design could have been the result of any deliberate political bias. She says it wasn't and I take her at her word, particularly given her fairness and transparency in responding to me. But I don't really see raising the possibility that the bias was deliberate as being particularly "inflammatory" -- it was manifestly a *possibility*, given how obvious the design flaw was relative to how smart and capable de Rugy obviously is.
Far more likely, however, is unconscious bias: how hard do you push back and cross-check your assumptions when you initially come up with a research finding that you "like" (or one that you don't like)? This kind of bias, almost by definition, is very hard to avoid, and potentially threatens the work of virtually every social scientist, not just de Rugy.
-- de Rugy is correct that many demographic variables are correlated with one another, which makes model speification more difficut and can lead to potential problems with overfitting. However, these demographic variables are also correlated with the poltical representation in the Congress. Moreover, because the stimulus consists of many different 'layers' (categories of projects), it is quite plausible that many different demographic variables (as well as intercations between two or more such variabes) could come to bear on how funds were ultimately distributed.
This is a sticky (albeit common) problem. The best way to handle it would probably be to make several different specificiations of the model and to publish them explicitly. If you have five different model specifications, all of which have roughly the same explanatory power, publishing only the one that you most like can potentially reflect bias. I have no idea how many other model specifications de Rugy tested and how many of them might have relieved the partisanship variable of its significance: it is not uncommon for a variable to go from *highly* significant to *completely* insignificant when a new variable with which it is correlated is introduced.
-- In general, there are a lot of things that de Rugy could have done -- both in terms of her research design and in terms of her presentation -- to give one more confidence that she had rigorously cross-checked and scrutinized her design, assumptions, and findings. If this were a chickenscratch, reasonably well-caveated, back-of-the-envelope blog post, I would have been more gentle in my critique. However, given that she has used her study to testify before the Congress, I believe it proper to hold her to a fairly high standard. That she is now willing to revisit her assumptions speaks highly of her, but until she does, her original study should be given no deference.
** Although its a relatively minor point, she did not, in fact, release the entirety of her dataset -- such as the economic variables she used in her regressions.
There's More...
Contract Post
by
Nate Silver
@
11:04 AM
A study purporting to find a connection between stimulus spending and the partisanship of a district suffers from an obvious flaw. But in so doing, it provides an example of why it's important to retain some common sense -- and some sense of context -- when conducting a statistical analysis.
The
study, by Veronique de Rugy of George Mason University and the National Review, claims that congressional districts which elected a Democrat to the Congress received a larger amount of stimulus finds by a margin which is statistically significant even after controlling for certain other effects like the unemployment rate. However, the study does not control for at least one other variable that is overwhelmingly important in determining the dispensation of stimulus funds.
The variable in question is in fact pretty obvious if you simply look at the districts that have received the largest amount of stimulus money, according to de Rugy's
dataset.
The district that received the largest amount of stimulus funding in the 4th Quarter of 2009, according to de Rugy's tally, is California's 5th Congressional District. Is there anything notable about the 5th Congressional? Well, it is home to the state capital, Sacramento. Let's keep that in mind.
Next on the list is New York's 21st Congressional District. The largest city in the 21st is the state capital of New York, Albany.
Third is the 21st Congressional District of Texas. It contains parts of Texas' state capital, the wonderful city of Austin. (Another district that contains parts of Austin -- the 25th -- ranks 14th on de Rugy's list.)
At this point, it ought to be pretty obvious what is going on. The three districts receiving the largest amount of stimulus funds are home to the capitals of the three largest states -- New York, California, and Texas. Let's pause for a moment and make a bold prediction. I'll bet you that the district that ranks 4th on the list will contain the capital of the 4th largest state, Florida.
Bingo. Up 4th on the list is Florida's 2nd Congressional, home to Tallahassee.
Fifth is Pennsylvania's 17th, which hosts the state capital, Harrisburg.
The sixth through tenth districts contain the capital cities of other large states: Ohio, Georgia, Michigan, Illinois and New Jersey, respectively. They are followed by districts that include the state capitals of Indiana, Tennessee, Virginia -- then another part of Austin, Texas -- then Arizona, Missouri, North Carolina and Wisconsin. Finally, in 19th place is South Carolina's 3rd Congressional District, which does not host a state capital. (Ironically, it has elected a Republican -- J. Gresham Barrett -- to the Congress).
This, of course, makes perfect sense. A lot of stimulus funds are distributed to state agencies, which are then responsible for allocating and administering the funds to the presumed benefit of citizens throughout the state. These state agencies, of course, are usually located in or near the state capital.
In fact, the differences are pretty overwhelming. There are 78 congressional districts that contain all, or part, of a state's capital city. Collectively, they received $118 billion in the fourth quarter. The 357 districts that are not home to a state capital received only $48 billion, however. On a per-district basis, the districts with state capitals received 11 times more funding. The ratio would be higher still if we excluded districts that included only outlying areas of state capital cities that do not host any state governmental institutions.
The other piece of the puzzle, of course, is that state capitals are much more likely to elect Democrats to Congress for a variety of reasons. They are, by definition, urban (although some smaller state capitals like Montpelier stretch the definition). They are, by definition, home to large numbers of governmental employees, who may be more sympathetic to bigger government. They tend to be highly educated and often are home to large state universities.
That de Rugy has testified before Congress on the basis of her evidence, and never paused to consider why the top five congressional districts on her list overlap with Sacramento, Albany, Austin, Tallahassee and Harrisburg, is mind-boggling. The presence of a state capital is the overwhelmingly dominant factor it predicting the dispensation of stimulus funds. This could have been discerned in literally five minutes if she had bothered to look at the apparent outliers in her dataset and considered whether they had anything in common -- a practice that should be among the first things that any researcher does when evaluating any dataset.
By the way -- if you throw out the districts that are home to state capitals, those which elected Democratic members to Congress still rank higher, receiving 31 percent more stimulus funds, on average, than those which elected Republicans. So, perhaps there is hope for her analysis yet. At that point, it would become important to consider other variables such as the economic conditions within each district. I'm not going to do her work for her, but I would suggest to de Rugy that she consider the following recommendations to correct other flaws with her research design:
First, she should look at unemployment rates at a district-by-district level, which are available through the American Community Survey, rather than at a metropolitan area level as she has done. Unemployment rates are often much higher in poor, downtrodden, inner cores of cities (which are also much more likely to elect Democrats to Congress), as opposed to suburbs and outlying areas.
Next, I would suggest that she look at a more robust array of demographic variables, such as the urban-rural distribution, the poverty rate (as opposed to just average income), the population, the number of seniors and children, and perhaps the racial composition. Were she have to considered these variables initially (particularly the urban/rural distribution), they may have nullified her conclusion, even without accounting for the presence of state capitals.
But my bet is that this is all a bunch of noise resulting from an incomplete -- and possibly deliberately biased -- research design. If de Rugy follows my recommendations -- excludes state capitals, accounts for a broader array of demographic variables, and evaluates unemployment rates at the district level -- and still finds a statistically significant positive relationship between the distribution of stimulus funds and whether the district elected a Democrat to Congress, I will buy her and three of her colleagues lunch anywhere in Washington or New York City. And if such a study is published in a credible, peer-reviewed journal, I will buy them dinner as well.
There's More...
Contract Post
by
Tom Schaller
@
9:48 AM
What to make of the president's decision yesterday to permit offshore drilling? A coldly rational decision that will reduce our dependency on foreign oil from politically unstable parts of the globe? A sell-out of his liberal base? A shrewd bit of wedge politics? A political bouquet to--or electoral boxing in of--a certain former Alaska governor?
Short answer: Probably a bit of all of the above. (Although on the last, it should be noted that Alaska's Bristol Bay was specifically cited by the administration as an area to be
protected from drilling.)
For starters, let's get this out of the way: Offshore drilling polls pretty darn well. A Rasmussen poll from late December
pegs general support at 68 percent nationally--although they note that many Americans, demonstrating their federalist sympathies, would support allowing individual states to exert veto power over drilling off their shores. The number was almost identical a year earlier in autumn 2008. Those from states along the eastern seaboard are less enthusiastic, but a
slight plurality is still supportive. Hell, even a
slight majority of Californians give drilling the green light.
Here's the key excerpt from Obama's remarks yesterday morning at Andrews Air Force Base announcing the policy as part of a new comprehensive energy plan. After talking about energy-saving initiatives and demand, he transitioned to the issue of supply:
"But we have to do more. We need to make continued investments in clean coal technologies and advanced biofuels. A few weeks ago, I announced loan guarantees to break ground on America’s first new nuclear facility in three decades, a project that will create thousands of jobs. And in the short term, as we transition to cleaner energy sources, we’ve still got to make some tough decisions about opening new offshore areas for oil and gas development in ways that protect communities and protect coastlines.
This is not a decision that I’ve made lightly. It’s one that Ken [Salazar] and I -- as well as Carol Browner, my energy advisor, and others in my administration -- looked at closely for more than a year. But the bottom line is this: Given our energy needs, in order to sustain economic growth and produce jobs, and keep our businesses competitive, we are going to need to harness traditional sources of fuel even as we ramp up production of new sources of renewable, homegrown energy."
In many ways, this is a classic Obama split-the-difference-with-a-tilt-to-the-left play straight from his Audacity of Hope playbook. Trying to preempt the kind of obstruction he experienced with healthcare reform as energy reform moves forward, the emphasis on drilling is an opening bid to win over some Republicans. The Post's Juliet Eilperin and Anne Kornblut correctly assess the move this morning as follows: "Some conservative critics questioned whether the policy will have any real impact on energy production, while liberals decried the risks to the environment. But the White House's key audience -- undecided senators who will determine whether a climate bill succeeds on Capitol Hill this year -- suggested that the move had helped revive the legislation's prospects."
Now, my initial reflex is, Haven't we been here before? Didn't the president learn from the drawn-out healthcare fight that the extended hand simply gets bitten, and that any attempts to cobble together anything beyond a Democrats-plus-maybe-one-Maine-senator "bipartisan" coalition is mostly a waste of his and our time? Yes and no.
Yes, because in the end, of course, it's gonna take mostly Democratic votes to pass any major energy bill. But no, because, well, energy is just not politically the same as health care. Let me explain what I mean.
In the period after Obama's win and before his first 100 days were over, one of the predictions I made in several public presentations--and got completely wrong--was that Obama would have to do energy before healthcare. My reasoning was akin to what we saw with George W. Bush tackling education early in his first term: You win on a stolen issue from the other party's traditional agenda. And, generally speaking, energy is more of a Republican issue, or certainly more Republican of an issue than healthcare. In other words, you win an away game to build confidence and momentum (political capital) to then win a much tougher home game. I'm mixing my metaphors here, but you get the point. And there's an obvious historic analog here with Bill Clinton and welfare reform, which many later concluded was a victory he could have used to then go for healthcare reform second, instead of the failed order in which he tackled them. (We'll never know.)
If offshore drilling--along with nuclear power investments, which should surprise nobody given Obama's background and Illinois roots, not to mention his clear if under-emphasized signals during the campaign about his support for nuclear--can win some converts, it's good politics; the poll numbers confirm that. Whether that politics is enough to win GOP converts, we'll see. As to whether it's good policy, well, I'll leave it to the readers to argue about below.
There's More...
Contract Post
by
Nate Silver
@
10:43 AM
It's a fun story: the idea that Republicans, perhaps because of the discouragement of people like Michelle Bachmann and Ron Paul, are slacking in returning their Census forms. But, save for some
anecdotal evidence in a few locales, it's just not clear that it has the virtue of being true.
Here, taken from the Census Bureau's
website, is the percentage of households who have mailed back their Census form in each state. Nationwide, 50 percent of forms have been mailed back to date. Thirteen of 21 red states (states which voted Republican in each of the last three elections) are above that threshold, however, and their average is slightly higher than the nation's at 51 percent.
There could still be evidence of a partisan shift in participation if it happened that red states were ordinarily quicker to turn in their Census forms and had fallen off their usual pace. But that does not appear to be the case either. In 2000, the red states were actually a bit behind the curve in terms of participation -- 70 percent of households turned in their forms without further prompting -- versus 72 percent of households nationwide. So, the red states have gone from slightly below-average to slightly above, rather than the other way around.
This wouldn't preclude the presence of highly localized effects in participation. Texas, where Paul is from, is indeed a bit low, although that's true of both red and blue areas of the state. Minnesota's participation rate of 56 percent is quite a bit above average, although lower than that of each of the states which it borders.
On the whole, however, this claim isn't supported by the evidence. And perhaps we shouldn't be surprised. Consider the consequence of failing to mail back your Census form: you'll have a government bureaucrat come knocking on your door, a prospect that ought to be truly terrifying to the residents of Bachmannistan.
There's More...
Contract Post
by
Nate Silver
@
12:37 PM
Color me a bit confused. Did the same person who once wrote a book entitled
The Neglected Voter: White Men and the Democratic Dilemma also just write an editorial entitled
Note to Liberal Elite: It's Not About Race?
Anyway, you come here for data-based argumentation, so let me give you some of that first -- and then we can come back to Real Clear Politics' David Paul Kuhn, the author of both those pieces. This is Obama's approval rating by racial group since the start of his term, as broken down by
Gallup:
Obama's approval rating among white voters has indeed fallen off by around 20 points since his tenure began. But, it's also fallen off by nearly as much among Hispanics. Opinions among black voters, meanwhile, have been more or less unchanged.
All of which goes to prove ... whatever you want it to prove, I suppose. But it's not clear that the decline in Obama's approval ratings have been especially steep among white voters -- once you're controlling for other variables like income and ideology. White voters' votes tend to be the swingiest to begin with -- followed by Hispanics, and distantly trailed by blacks. If you compared, say, white moderates/independents to Hispanic moderates/independents or Asian moderates/independents, you might find about the same erosion of support among all three groups (or you might not -- perhaps Gallup will see fit to publish this data).
Another reason the decline looks so steep is because Obama got a fairly large bounce in his approval ratings between the election itself and his inauguration -- something which used to be routine during the Eisenhower/Kennedy era, but which hasn't been recently. Most of that bounce -- which ironically very well may have had to do with the historical significance of Obama's election -- came from white voters, as we can see if we add the election day exit polling to our approval ratings chart:
Obama's approval decline among white voters does not look so steep if you compare it to his election day results -- nor for that matter to the election results of past Democratic candidates. Since 1972, the Democrat's share of the white vote in Presidential elections has been: 31 percent, 47 percent, 36, 35, 40, 39, 42, 41, and 43 percent (Obama). Something around 39 or 40 percent of the white electorate -- where Obama's approval ratings seem to be now -- is pretty much right in that ballpark.
So, I'm not totally unsympathetic to the argument made by Kuhn. To say that opposition to health care reform is all about race would be a hasty generalization. Surely the fact that people are stressed out about the economy must warrant a mention in the discussion. Likewise, that the concerns that the electorate developed during the Reagan Era about the proper role of government have never entirely worn off.
At the same time, whenever this subject comes up -- as it has recently because of Frank Rich's editorial in the New York Times this Sunday -- it always seems as though the comeback is to respond to hasty generalizations with even hastier ones. Kuhn takes a relatively nuanced argument that Rich makes and characterizes it thusly:
If Rich is correct, and opposition to the healthcare overhaul concerns race, then the roughly 125 million white adults who do not approve of the legislation are racists.
Rich, of course, never said anything like that. He column was focused more on what he calls "right-wing extremism" than on the feelings of the electorate at large. He comes closer to saying that the anger he sees embodied by the tea-party movement has much to do with race (a point on which I'm ambivalent -- and which Kuhn, ironically, seems almost willing to concede). But, Kuhn's formulation of Rich's argument is, of course, a straw man.
There's More...
Contract Post
by
Nate Silver
@
10:05 AM
I don't know that I'd make
too much of this
Mason-Dixon poll in Florida which has very bad numbers for Democrats across the board -- Mason-Dixon, although a strong pollster, has had a pretty significant red-leaning house effect for the past couple of cycles and there aren't any recent trendlines to look at on the health care or Bill Nelson numbers.
But, it would stand to reason that Obama's 2012 strategy may not be able to bank on the Sunshine State. Seniors don't like the health care plan and vote in great numbers in Florida. And there are also a lot of Jewish voters in Florida who might not be pleased with Obama's somewhat hawkish stance toward Israeli settlements. This isn't rocket science.
Yes, completely different issues may be on the table by the time that '12 rolls around (actuarially speaking, for instance, there's a decent chance of a shock in U.S.-Cuba policy caused by Castro's death). And it's a little early to be worrying too much about the electoral math. But considering how tenuous Obama's hold over Florida was in the first place -- it wasn't really a swing state in 2008 so much as a come-along-for-the-ride state -- just a point or two that pushes it further to the right of the country as a whole could deprive it of the critical role it has played in past election cycles. Obama could perfectly easily still win it -- but it seems doubtful that he'd win it before an Ohio or a Colorado or some other combination of states that has him well past the 270 mark.
There's More...
Contract Post
by
Renard Sexton
@
7:15 AM
As we approach the closing weeks of the UK campaign, it has become increasingly clear that the campaign will be hard fought to the bitter end.
Labour party insiders told me that while "even six months ago most in the party felt that there was no question that this would be a bad loss," there is new energy in the Labour campaign among the party faithful. At the same time, there remains the powerful refrain from swing voters that change is badly needed. A 30-something Conservative party campaigner explained that, "though [she] voted for Tony Blair in 1997 and 2001, it could not be clearer" that Labour must be "swept out".
I spent five days in London talking numbers and personalities with several of the main UK pollsters last week, finding a great deal of agreement on overall trends of the race, but substantial differences of opinion on the likely outcome.
The generally accepted state of the race is largely dictated by YouGov's internet polling operation, which through its high quantity of polling (commissioned mainly by the The Sun and the Sunday Times) tends to dominate the various "uniform national swing" calculators out there.
As such, the 2 point reduction in Tory support during March that YouGov polling found has been echoed widely in the media. At the same time, newcomers Angus Reid (a Canadian firm) have suggested that things are much worse for Labour than many suspect, and they project a far better performance from the Liberal Democrats and smaller parties like the BNP, UKIP and the regional/nationalist parties.
Of course, these national horserace numbers only tell us so much about the state of the race. Here are three factors (among several) have not yet been accounted for:
1. UK pollsters deal with undecided voters quite differently than US polling outfits: Some pollsters exclude undecided or refused voters entirely, while others count them based on their previous voting affiliation at a discounted rate (often 50%). Because the race is so close, how undecided voters break will be a key factor, making projections quite difficult.
2. UK pollsters account for voter likelihood mainly through self-rating of intention and previous voting history: Most pollsters ask respondants to rank their chance of voting from 1-10, and then exclude those respondants who do not reach a particular threshold. Some only take 10s, other 8-10, and so forth. Some also ask whether the respondant voted in the last election. In some cases this has the effect of a quite strong enthusiasm screen, as we have discussed on 538 several times before.
Given the relatively high voter turnout in UK elections, (e.g. 71.4 percent in 1997, 59.4 percent in 2001 & 61.4 percent in 2005), the downside to overly-aggressive likely voter and undecided screens could be quite significant.
3. Swing voters could (and likely do) behave differently: Marginal seats will of course play the kingmakers this time around. Within them, a relatively small number of swing voters will make the difference one way or the other. March polling from Ipsos published in the Telegraph declared that "in 56 Labour-held seats that the Tories need a swing of between 5% and 9% to win... Labour leads the Tories by four points among those certain to vote."
However, of the 1007 respondants, 440 were screened out as not definite to vote (only 10 out of 10 voter intention was kept), leaving 567 voters. Voting sub-group weights then screened out an additional 34 people, mainly to account for an under-representation of the 18-34 group, and over-representation of the 55-64 and 65+ groups. An additional 40 undecided voters (8 percent), and 12 "refused" voters (2 percent) were excluded from the final 479 respondants upon which the 4 point lead was based.
On top of this, 30 percent of respondants certain to vote said that they may change their vote before the election, along with 50 percent of all respondants.
(A small aside: if the marginal seats included in this study range from L+5 to L+9, with a back of the envelope average of L+7, they have just a L+4 lead overall, this could be used to suggest an overall Conservative swing in LAB-CON marginals of +3, which would flip 40 seats.)
Based on these factors, we here at 538 hope to build a slightly improved mousetrap over the next couple weeks in order to project marginal and safe seats with at least one level of further precision.
At the same time, there will be narrative coverage of several interesting marginal seats, and qualititative comparison of the campaign culture, numbers game and political system in the UK with other systems our audience is familiar with.
---
Renard Sexton is FiveThirtyEight's international affairs columnist and is based in Geneva, Switzerland. He can be contacted at sexton538@gmail.com
Additional information on FiveThirtyEight's coverage of the United Kingdom 2010 General Election can be found here.
There's More...
Contract Post
by
Nate Silver
@
6:34 PM
Michigan State men's basketball coach Tom Izzo, whose Spartans today
made their 6th Final Four in the 15 years under his tenure, is rightly regarded as something of a magician when it comes to post-season basketball. But just how good is he? Very, very good, it turns out.
Michigan State has played in 13 tournaments under Izzo. They've reached the Final Four six times: thrice as a 1-seed, once as a 2-seed, and twice -- including this year -- as a 5-seed. They've also failed to make the Final Four as a 4-seed, a 5-seed, a 6-seed, a 9-seed, a 10-seed, and twice as a 7-seed.
What are the odds of a team with these seedings advancing to the Final Four six times? Very, very low. We can do a good job of modelling a team's probability of making the Final Four by using a
logistic regression model based on the square root of its seeding:
Number 1 seeds, for instance, theoretically have a 43 percent chance of making the Final 4, which closely matches the 42 percent success rate that they've actually achieved. So, even though the Spartans have had a #1 seed three times under Izzo, and a 1-seed is "supposed" to win its bracket, it's to his credit that Michigan State has converted all three times.
A 2-seed, meanwhile, which was good enough for the Spartans to make the NCAA Championship game last year, has only about a one-in-five (22% empirical, 23% theoretical) chance of making the Final Four. And a 5-seed, which the Spartans have used to advance to the Final Four twice, has just a 1-in-20 chance.
Meanwhile, Michigan State doesn't really have any bad misses. They failed to make the Final Four as a 4-seed in 1998 -- Izzo's first tournament -- but 4-seeds have only an 8 percent chance of doing so. They also missed as a 5-seed in 2008. Overall, however, a team with these seedings should only have made about two Final Fours -- 1.8, technically, if you don't mind the decimal -- and not the six that the Spartans have actually participated in.
A simple way to analyze how good Izzo has been is to add an 'Izzo' dummy variable to our logistic regression model. If we do that, it is highly statistically significant at a 99.99 percent certainty threshold (z = 3.63). Translation: it's very unlikely that the Spartans have just gotten lucky; Izzo has probably exhibited some real skill at captaining his team in the tournament.
A couple of caveats. First, since Izzo has only been a head coach for Michigan State, it's plausible that we're giving him credit when there's something about the entire program that instead deserves the praise. But since the head coach is so instrumental to building a college basketball team, and since the player personnel rotates at least once every four years and often more rapidly, I'm not sure that's the same concern it might be in another sport. Every player who has ever played four years under Izzo has made a Final Four.
Secondly, this does not necessarily prove the existence of "clutch coaching". To do so would require a more rigorous study that evaluated the tournament performance of all different sorts of coaches. FWIW, I also ran this analysis for Duke's Mike Krzyzewski, and found that Duke's tournament performance under his tenure fell just short of the 95 percent significance threshold.
Finally, analyses like this are always subject to cherry-picking. Although the odds of any one particular coach doing as well as Izzo in the tournament are very low, the chances that some coach somewhere will have a record like this are considerably higher.
But for the time being, you can enjoy this for what it is. And so long as Izzo remains their coach, you should probably subtract a seed or two from Michigan State the next time that you're filling out your bracket.
Note: The author grew up in East Lansing, Michigan.
There's More...
Contract Post
by
Tom Schaller
@
12:32 PM
With the diversions of healthcare reform vote mostly behind us, let's return to
something I wrote last week about the possible effects of the decline in the so-called "Obama surge voters" and how it might affect the two parties' chances this fall. In that post, I computed and reported the absolute and relative turnout changes in all 50 states between 2006 and 2008.
I suppose there are two ways for Democrats to look at the states that witnessed a huge voter turnout surge in 2008. The
optimistic view is that a significant enough share of these newly-registered or mobilized Obama surge voters will turnout for the 2010 midterms, buffering expected Democratic losses; the
pessimistic view is that, precisely because Democrats in these states benefited most from the surge, these are the very states where the off-year, drop-off effect will be most electorally painful. For what it's worth, I suspect the latter will be closer to the truth come November: That is, one would rather be a Democrat in Minnesota or South Dakota, where there was already high turnout prior to Obama's political rise, than a Democrat in Nevada or North Carolina, where Democrats might be dealing with a severe drop-off effect in the midterms absent Obama on the ballot.
In any case, here is a quick survey of Senate and governors races (House analysis, which is bit more cumbersome, to follow in a future post):
Senate. Let's start with Nate's ratings of Senate races, left. Of the top 10 states most likely to flip--in order: North Dakota, Delaware, Arkansas, Indiana, Nevada, Pennsylvania, Colorado, New Hampshire, Missouri and Ohio--four of them (DE, IN, NV and NH) are among the 21 states that witnessed above-national-average relative turnout increase between 2006 and 2008. Because four of these 10 races are among the roughly 40 percent of states above the national mean, we can say the 10 most competitive Senate race states are neither under- nor over-represented among the big turnout jump states of 2008; put another way, any retention or decline of Obama surge voters will probably not have a disproportionate effect on the overall slate of Senate contests and likelihood of Republicans flipping the Senate this November.
But consider Indiana, which saw a bigger turnout boost in 2008 which helped make it the state with the largest net partisan swing between the 2004 and 2008 presidential cycles. The Hoosier State in 2008 experienced massive turnout increases in cities like Gary, located in Lake County, which is one-quarter African American. Lake County's 2008 turnout jumped 42 percent in absolute terms and an eye-popping 150 percent in relative terms, from just 28 percent in 2006 to a stunning 70 percent in 2008--a rate of change that dwarfs Indiana's 22.7 absolute and 62 percent relative changes statewide, and which in turn ranks it among the top dozen states nationally in terms of the 2006-to-2008 jump. Anytime a centrist with deep family roots like Evan Bayh retires suddenly in a generally red state, it's a problem for Democrats. But to have that happen in Indiana right now compounds the Democrats' problems.
A final point in regard to Nevada and Harry Reid's electoral fate: If there are places that Obama is likely to help or hurt by appearances on behalf of Senate Democrats, Nevada is the kind of state where in general it helps to have the president at your side. Little wonder, then, that Reid invited Obama out to Nevada last month to stump on his behalf.
Governors. Of the 18 governors races that the Cook Political Report rates as toss-ups as of today, March 28--the page gets updated as rankings change, so depending upon when you hit the link it may or may not be 18 or the same 18--only six are also among the 21 states that experienced above-national-average relative turnout increase. They are California, Nevada, Texas, Florida, Georgia and Obama's home state of Illinois. If one takes the pessimistic view of the retention/surge scenario, that means the Democrats are lucky there aren't more large-surge states among the competitive races.
But the first thing that jumps out at me from looking at these six is that five of them are among the nation's 11 most-populous states, which means the gubernatorial outcomes in these five could have very important implications for the drawing of new House maps (and, of course, state legislative maps). Many of these six states also have both significant black and Hispanic populations, in some cases in the double-digits for each minority population group. That means we will get to see if the retention or decline of minority Obama surge voters is the same or different between these two subgroups of surge voters.
As for specific gubernatorial races, the obvious one to watch is in Obama's home state of Illinois. A Democratic loss there would be similar to the Scott Brown victory for Republicans in Massachusetts--a symbolically powerful "message" to the White House. But I'll also be curious to see what happens to turnout in Florida. Florida's result (and to a lesser degree Nevada's) could be especially telling because it is the only large presidential swing state among the six--and it's also a state with a huge, white senior/retiree population, too. If there is a racial and/or generational cleavage in partisan voting this November, Florida is one place where its effects should be demonstrable.
There's More...
Contract Post
by
Nate Silver
@
9:55 AM
For some time, I've been part of the doom-and-gloom brigade when it comes to Democrats' fortunes at the midterm elections this November. As early as last August, on a panel at the Netroots Nation conference, I said that I expected a loss of 20 to 50 Democratic seats, which necessarily implied that a loss of their majority was quite possible. I've since revised the low end of that estimate
downward, to a loss of 20 to 60 seats.
I'm not sure that there's yet been enough time to assess whether the Democrats' passage of health care reform seven days ago could mitigate -- or broaden -- their losses. Most polls suggest that the health care reform bill itself has become somewhat more popular since passage. But President Obama's approval ratings are little moved, and there has thus far been little new polling on the generic ballot or perceptions of the Democratic congress. Moreover, any changes in the polling may prove to be temporary.
Still, there is one set of numbers that potentially contain relatively good news for the Democrats. These concern the
enthusiasm gap, which may be lessening. Daily Kos / Research polling has
found that while Republican voters remain exceptionally engaged by the midterm election cycle, Democrats are becoming increasingly engaged as well. Rasmussen, meanwhile, has
found that about 5-7 percent of voters nationwide have gone from being
somewhat approving of to Obama to
strongly approving of him -- and almost all of the movement is accounted for by Democrats. (Obama's
disapproval -- including his strongly disapprove numbers -- are little changed in the poll).
At the very least, there does not appear to be any fresh animus toward the Democrats for their actually having passed health care, whereas there probably is some fresh enthusiasm from their base. As such, I would probably revise my estimate of the Democrats' losses just slightly, from a loss of 20-60 seats to a loss of 15-55. (Think of these numbers as representing perhaps the 10th and 90th percentiles, respectively; there remains some chance that the Democrats could lose more than 55 seats, or fewer than 15.)
But since so much attention has been focused on the potentially catastrophic losses for the Democrats, let's pause for a moment to consider their upside case. How many seats could they lose while still having the midterms be a "win" for them?
Obviously, there are some overly literal ways to interpret this question. One could say that so long as the Democrats lost any seats at all, it would still be a "loss". Or, one could say that so long as they preserve their majorities by one seat, it would still be a "win".
A better way to interpret this question might be: how many seats can the Democrats lose while still having the chance to advance the key components of their agenda? In the Senate, I have argued, the number is probably about 3, as a 56-seat Democratic majority would allow them to formulate a 60-seat Democratic + RINO coalition with Scott Brown, Olympia Snowe, Susan Collins, and Delaware's Mike Castle. (If the Democrats become more willing to use the reconciliation procedure after their victory on health care, of course, the math become somewhat fuzzier.) But what about in the House?
I took each Democrat-held House seat and assigned it points based on the race ratings from the four major forecasters: Cook, CQ, Rothenberg and Sabato. One point was awarded for a characterization of likely Democratic, 2 for lean Democratic, 3 for toss-up, 4 for lean Republican, and so forth. I then summed the ratings between the four forecasters and sorted the races from most to least vulnerable, randomly ordering the Democrats in the case of ties.
The idea is to assess how the ideological position of the median member of the House changes with an increasing loss of seats. Assuming that there were no changes in the composition of the Congress, for instance -- no loss of seats, including in upcoming special elections, and that any retiring members were replaced by ideologically-identical cousins from the same party -- the median DW-Nominate score of the House would be -.186 on a scale that runs from -1 (very liberal) to +1 (very conservative). This represents, in essence, the status quo.
From there, however, I began eliminating the Democrats one at a time in order of their vulnerability and replacing them with Republicans with an average Republican ideology score (+.635). The first Democratic-held seat to go, for instance, is that held by the retiring Bart Gordon in Tennessee's 6th congressional district, which most forecasters expect to go to the Republicans. This technically makes no difference in the position of the median vote, however, as Gordon was already (slightly) to the right of the median member of Congress. The second-most vulnerable seat is LA-3, which Charlie Melancon is vacating to run for Senate. A Republican win there would also not move the median, however, since Melancon was likewise to the right of the median Congressman.
Eventually, however, some non-Blue Dogs inevitably begin to lose their seats, and the losses begin to add up. Needless to say, there is a very substantial shift with the loss of the 40th seat, when control of the majority changes hands. (There are hardly any moderate Republicans to form a buffer zone, as no current House Republican has a DW-NOMINATE score of lower than +.259, and few truly Republican moderates are likely to survive past the primary in the current political climate.)
There is also something of an inflection point, however, that begins with the loss of about the 20th or the 23rd seat. Up to that point, the median vote in the House does not change very much, since most of the losses will be suffered by Blue Dogs who are not reliably voting with the Democratic majority in the first place. From that point forward, however, the Democrats quickly run out of Blue Dogs, and mainline members of their majority start to be defeated. The loss of the 21st through the 30th seats, for instance, moves the median vote almost twice as much as the loss of the first 20, and as the losses begin to accumulate into the 30s, the Democrats' hold on power becomes very tenuous even if they nominally still control the House.
If the Democrats do not want their agenda to be substantially injured, then, they probably have an implicit goal of not losing more than about 20-23 House seats, and not more than about 3-4 Senate seats. Coincidentally, this is pretty close to the average loss suffered at the midterms by the incumbent party since World War II:
Whether a loss of, say, 19 House seats and 3 Senate seats would be regarded as a "win" by the media is hard to say -- and would depend in part on what expectations had been established in the days and weeks immediately preceding the elections. Given where expectations are right now, however, I suspect that most Democrats would go to bed feeling pretty happy about that outcome and Republicans feeling as though they'd blown an opportunity. Indeed, if the narrative around them also improved -- such that they felt emboldened to push their agenda or Republicans became more wary of obstructionism -- their overall position would arguably even be somewhat stronger. For the time being, however, this remains something of a best-case scenario for them.
There's More...
Contract Post
