jakke

Jan 03 2012 ∞

Seeing as the diverse and representative voters of Iowa get to pick the next Republican candidate for US President today, I figured it was a reasonable time to update the national uniform swing projections.

As always, these predict the election by taking national-level polling data, assuming that any shift in party support from the 2008 election comes from a uniform shift everywhere in the country, and using this information and the polling margin of error to generate win probabilities. The colours on the map indicate the probability of the win, not the expected vote share - because getting 50%+1 of the votes generally means a win. (Except in Nebraska and Maine. Which I’ve accounted for here.) You can see the predictions from two weeks ago here, for comparison.

This time I’ve added a fourth map, which tries to predict what happens if Ron Paul runs as a third-party candidate against Romney and Obama. (This honestly seems pretty likely to me, especially if he’s still polling well late into the spring.) Under the national uniform swing assumption, it’s obviously pretty difficult for Paul to win any seats whatsoever, but the national uniform swing assumption might not be too accurate.

Anyway, you can and should click through to see the full-sized maps - but for now, here are some highlights:

  • The race between Romney and Obama would be close to a tossup, and Romney would actually be a very narrow favourite to win the popular vote. However, much of this popular vote advantage would get wasted by winning the states he won by a landslide, while Obama would be winning his states by a narrower margin - so, advantage Obama. Any slight deviation towards Romney from national uniform swing in half a dozen states would make Romney the favoured candidate. 
  • Not only is Romney really competitive against Obama, but he’s basically the only candidate who is competitive against Obama. Paul has basically no chance of winning, and Gingrich is even worse. At this point it’s difficult to keep making the excuse that voters just don’t know enough about Gingrich or Paul, because they’ve both had heaps of media coverage. (And to be realistic - neither is a candidate that becomes more appealing with increased scrutiny.)
  • In a three-way matchup, Obama has such a clear advantage that it wouldn’t even be sporting anymore. He’d be competitive in states like Montana and Georgia. For all the talk of Paul attracting Democratic and independent voters during the primary season, virtually everyone who says they’d vote for him in a three-way election comes from the Republican Party. This actually makes sense, because honestly his positions on abortion, taxes, gun control, and other core Republican issues are entirely within the party mainstream, and the vast majority of voters really don’t care about the Federal Reserve or whatever.

Anyway, obviously all of this can change hugely, because the actual election isn’t for another ten months. But if Romney starts winning Iowa, New Hampshire, and so on, then Democrat types should start getting very worried that this is going to be a really contentious election.

36 notes

Dec 21 2011 ∞

Here are some updated national uniform swing predictions. You’ll definitely want to click on those maps to enlarge.

National uniform swing means taking national-level polling data, and then assuming that any shifts at the national level are happening uniformly across all states, and then using this to predict the election. I use the margin of error on the polls to generate margins of error here. The colours on this map represent the probability of Obama or the Republican nominee winning, not the vote shares - because even if Obama is getting 50% + 1 votes in a given state, he’s still winning that state’s electoral votes. (Except Nebraska and Maine. More on that later.)

These predictions are drawn from the last two weeks of national polling, which has included fifteen polls with at least one of these three candidates from nine different polling agencies. It accounts for the new district allocations from the most recent census, as well as the electoral college allocation systems used in Nebraska and Maine. (The only individual district that was particularly competitive was Nebraska’s 2nd, which Obama won last time. Due to redistricting, I wouldn’t take the district-by-district predictions seriously, though.)

Note that of these three candidates only Romney is even possibly competitive, and even then Obama has a big advantage unless Romney starts winning additional Midwestern states. Even though Newt Gingrich gets the serious-thinker treatment while Ron Paul gets the conspiracy-theorist treatment, Gingrich does substantially worse than Paul. Given current polling, it is virtually impossible for Gingrich to win. He’d have to start polling ahead in states like Pennsylvania or Ohio while not losing ground in Georgia or Montana. A Gingrich nomination for the Republicans would basically guarantee an Obama win, even more so than a Paul nomination.

Obviously, all of this could change. The election is still ten and a half months away, and campaigning, economic conditions, foreign affairs, and voters’ increasing knowledge about the candidates all make a big difference. But right now, I can’t see any way that Gingrich could possibly become President.

41 notes

Dec 19 2011 ∞
Okay so America’s favourite charming stoner racist great-uncle is apparently leading in Iowa, according to Public Policy Polling. While obviously it’s not productive to read too much into any individual poll, PPP polls extensively in Iowa, and in the past they’ve produced excellent results, and these shifts in poll results (especially the drop for Gingrich) are hugely statistically significant. I’m guessing the drop for Gingrich comes from the fact that as primary voters learn more about him (and his multiplicity of wives) they are less impressed by his ability to return the United States to some mythical glorious pre-Mad Man era patriarchal Christian paradise. This has kind of been a pretty steady pattern throughout the campaign process: voters identify a candidate who seems promisingly conservative and personable, they learn more about the candidate’s past and priorities through increased media scrutiny, and they realize the candidate is undesirable. It’s happened for Bachmann, Perry, Cain, and now Gingrich. Anyway, apparently they’re going through the same process for Paul, at least in Iowa. Media coverage of the Paul campaign tends to depict it as some kind of wacky sideshow, but his local-level campaigning has apparently been really effective in Iowa. I think when he focuses on restricting abortions, removing restrictions on gun ownership, and removing social spending, it makes him really attractive to Republican primary voters - as they’re probably more motivated by these issues than esoteric concerns about the gold standard. And I’m not sure there’s enough time left for primary voters to lose interest in Paul before the Iowa caucuses. What happens if Ron Paul wins the Iowa caucuses? Probably not much, honestly. Winning the Iowa caucuses has been a noisy indicator in Iowa, especially for the Republicans. In 2008 Huckabee won, and in 1996 and 1988 Bob Dole won. None of them ended up being the Republican nominee, obviously. In 2008 John McCain came in fourth, behind some guy called Fred Thompson, and still won the nomination overall. And I can’t imagine Fox News being like “oh hey Ron Paul won let’s all throw our support behind him” like they might for, say, Perry. Anyway, this has kind of motivated me to put up more 2012 US general election prediction maps after I’m done with this last paper. I think I’ll do Obama vs Romney, Gingrich, and Paul. (An unsurprising preview: Newtmania is not currently sweeping the nation.)

Okay so America’s favourite charming stoner racist great-uncle is apparently leading in Iowa, according to Public Policy Polling. While obviously it’s not productive to read too much into any individual poll, PPP polls extensively in Iowa, and in the past they’ve produced excellent results, and these shifts in poll results (especially the drop for Gingrich) are hugely statistically significant.

I’m guessing the drop for Gingrich comes from the fact that as primary voters learn more about him (and his multiplicity of wives) they are less impressed by his ability to return the United States to some mythical glorious pre-Mad Man era patriarchal Christian paradise. This has kind of been a pretty steady pattern throughout the campaign process: voters identify a candidate who seems promisingly conservative and personable, they learn more about the candidate’s past and priorities through increased media scrutiny, and they realize the candidate is undesirable. It’s happened for Bachmann, Perry, Cain, and now Gingrich.

Anyway, apparently they’re going through the same process for Paul, at least in Iowa. Media coverage of the Paul campaign tends to depict it as some kind of wacky sideshow, but his local-level campaigning has apparently been really effective in Iowa. I think when he focuses on restricting abortions, removing restrictions on gun ownership, and removing social spending, it makes him really attractive to Republican primary voters - as they’re probably more motivated by these issues than esoteric concerns about the gold standard. And I’m not sure there’s enough time left for primary voters to lose interest in Paul before the Iowa caucuses.

What happens if Ron Paul wins the Iowa caucuses? Probably not much, honestly. Winning the Iowa caucuses has been a noisy indicator in Iowa, especially for the Republicans. In 2008 Huckabee won, and in 1996 and 1988 Bob Dole won. None of them ended up being the Republican nominee, obviously. In 2008 John McCain came in fourth, behind some guy called Fred Thompson, and still won the nomination overall. And I can’t imagine Fox News being like “oh hey Ron Paul won let’s all throw our support behind him” like they might for, say, Perry.

Anyway, this has kind of motivated me to put up more 2012 US general election prediction maps after I’m done with this last paper. I think I’ll do Obama vs Romney, Gingrich, and Paul. (An unsurprising preview: Newtmania is not currently sweeping the nation.)

17 notes

Nov 27 2011 ∞

U.S. Workers’ Sliding Share of Pay Poses Spending Risk: Economy

Gross domestic income, or the money earned by the people, businesses and government agencies whose purchases go into calculating growth, rose at an average 2.8 percent annual rate from April through September after climbing 4.3 percent in the previous six months, Commerce Department data on Nov. 22 showed. Employee compensation last quarter accounted for its smallest share since 1955.

In contrast, the portion accruing to corporate profits was the biggest since 1950, showing companies are hoarding cash as concern grows that a European country will default on its debt and that deficit-reduction gridlock in Washington will continue. Without more pay and a pickup in hiring, households may ring in 2012 by making their own budget cuts.

“Businesses are very cautious so they’re not hiring and they’re not distributing their profits to consumers as they had in past expansions,” said Michelle Meyer, a senior U.S. economist at Bank of America Corp. in New York. “With slow wage and salary growth, consumer spending will be on a sluggish trajectory.”

[…]

A pickup in spending in the third quarter, a 2.3 percent gain at an annual rate after a 0.7 increase in the previous three months, was the “the worst possible outcome,” Tom Porcelli, chief U.S. economist at RBC Capital Markets Corp. in New York, said in a Nov. 22 interview on Bloomberg Television, since it was “entirely because of a draw down in savings.”

The savings rate dropped to 3.8 percent in the third quarter, the lowest level since the last three months of 2007, which marked the end of the previous economic expansion.

Okay, so this is a big problem. Nothing slows down economic growth more than a decrease in the share of gains going to wages, and this kind of situation (where the economy overall is growing but wages and employment aren’t) is basically unprecedented in post-World War Two recession recovery. Usually, when the economy improves, wages pick up too - and right now, this isn’t happening at all. As I’ve noted previously, the same thing is happening in Canada.

What’s going on here? It seems like companies are just really fearful that banks are going to collapse in Europe and trigger another big crash, so they’re not making any investment in hiring new workers or offering higher salaries. At the same time, workers are all so convinced that the economy is terrible right now that they’re too afraid to ask their employers for raises or look for higher-paying jobs. The result is a really bad equilibrium where the economy continues to grow but all that growth is realized on the capital side of things and wages don’t really move.

Why is this a bad thing? Well, obviously it’s seriously suboptimal for low-wage employees who are having trouble paying for necessities in the face or rising food and fuel costs. It’s also totally unsustainable, because in wealthy industrialized countries like the US consumer spending is a huge part of the economy, and without consumers spending it’s not clear how the economy can return to steady growth and full employment, ever. On another level, it indicates that the current theoretical models that give wages rising with economy growth are not as robust as we thought they were (and there is already serious work underway to improve them).

Really with news like this it’s kind of amazing that more people aren’t out demonstrating in the streets. It’s kind of a tribute to how much the Occupy movement has appealed mostly to college students from relatively wealthy backgrounds that it’s stained any activism in this regard as some kind of economically privileged liberal pastime.

6 notes

Nov 21 2011 ∞
“ U.S. credit rating companies have made it clear that, while a failure of the supercommittee might not lead to a credit downgrade, undoing the automatic cuts probably would. Moody’s said the lack of a supercommittee agreement wouldn’t on its own cause the U.S. to lose its top credit ranking because the August debt-ceiling deal includes $1.2 trillion in automatic cuts.
— Over the next few days you’re likely to see moves to undo the automatic spending cuts that will be triggered by the supercommittee’s failure to reach any agreement. Half the cuts would come from defense spending and since many entire Congressional districts are kept economically viable by defense spending this would be considered political unacceptable. Expect investors to push back.

(Source: bloomberg.com)

Nov 15 2011 ∞

Republican Candidates in Four-Way Dead Heat

A Bloomberg News poll shows Cain at 20 percent, Paul at 19 percent, Romney at 18 percent and Gingrich at 17 percent among the likely attendees with the caucuses that start the nominating contests seven weeks away.

Economic issues such as jobs, taxes and government spending are driving voter sentiment, rather than such social issues as abortion and gay marriage, the poll finds. Only about a quarter of likely caucus-goers say social or constitutional issues are more important to them, compared with 71 percent who say fiscal concerns.

The poll reflects the race’s fluidity, with 60 percent of respondents saying they still could be persuaded to back someone other than their top choice, and 10 percent undecided. Paul’s support is more solidified than his rivals, while Cain’s is softer. All of the major contenders have issue challenges to address.

Yup, Ron Paul is basically tied for the lead - and counting only the 40% who are sure whom they’re going to support, he leads Romney 32% to 25%. Will Ron Paul win the Iowa caucuses and become a frontrunner for the Republican nomination?

Ugh, I just felt an eldritch chill when I typed that. Almost definitely not, though. Given the bizarre nature of the Iowa caucus process, where sometimes supporters of one candidate are required to switch to supporting another candidate in a very public process, it’s very difficult in situations with multiple candidates in a near tie to predict who’s going to win overall. However, it seems likely that most Cain, Romney, and Gingrich supporters would have another of these three candidates rather than Ron Paul as their second choice, so it’s way more likely that one of them will win the most delegates.

Note that this poll was conducted on behalf of Bloomberg by Ann Selzer, who’s possibly the most careful, thorough, and accurate pollster operating in the US right now. Glad to see Bloomberg’s hired her.

EDIT: Nontrivial correction here - as jasencomstock points out, the support-switching thing only happens with Democratic caucuses in Iowa. Republicans aren’t expected to do it. So Ron Paul isn’t so disadvataged at all. Time to stock up on doubloons?

8 notes

Sep 17 2011 ∞

Here are this month’s US 2012 presidential election national uniform swing projections. Basically, it works like this: I take all the polling data on Republican vs Obama matchups, assume that any change from the 2008 results is uniform across all states (this is a big assumption), and then use the margin of error on the polling data to generate probabilities for each candidate’s likely outcomes. Here are the results, for polling data available up to 17 September 2011.

A note on the maps here, because last time I didn’t make this entirely clear: the colours represent the probability of the state going to each candidate. They’re not directly related to each candidates’ share of the vote. The rationale here is that states are generally winner-take-all, so even if a candidate is consistently losing by 5% they’re still losing the state. (A couple states aren’t winner-take-all. More on that in a bit.)

So here’s how things are looking for the candidates right now: 

  • The frontrunners in the Republican race (Romney and Perry) are both relatively uncompetitive in a general-election setting under current polling. Obama’s approval rating may be falling, but the general public still sees him as a superior candidate to either of these two. Obviously that could change over time as public perception of Obama changes and knowledge of the Republican candidates increases.
  • Bachmann would stand no chance in the general-election setting. None whatsoever. 
  • Ron Paul surprisingly doesn’t do that much worse than Romney and Perry. Probably this is because his positions on micro-level domestic policy (e.g., taxes, guns, abortion) are right in keeping with the mainstream Republican opinion, and he’s probably at sufficiently low name recognition that many poll respondents consider him to be a generic Republican. In some bizarre universe where he became the candidate, his performance would likely be far worse than polls currently predict.

Considering a Perry or Romney run for a moment (since those are the most likely outcomes), we can look at which states are the most on-edge between going for Obama or for the Republican - that is, which states would be the swing states under current polling. Here are the top five, from most to least swingy:

  • Ohio
  • Florida
  • Nebraska 2nd district (the part with Omaha)
  • Indiana
  • Virginia

This list really shouldn’t be too surprising to anyone. Obama carried all of these last time, but it’s not entirely clear he will do so again. Ohio and Florida would likely be very tight, and Indiana would most likely be in the Republican column right now.

As the real political nerds among you probably know, there might be a few changes in the way that states allocate their electoral votes between now and 2012. Nebraska currently allocates its electoral votes on a district-by-district basis, and they might be changing back to winner-take-all. Last election Obama took the district containing Omaha, but that’s the only competitive part of the state. On the other hand, Pennsylvania might be moving in the opposite direction - switching from winner-take-all to district-by-district. This could have a huge effect, as Pennsylvania has a few heavily Democratic districts in urban areas and more Republican-leaning ones in rural areas. These two changes combined would probably shift things five or ten electoral votes in the Republican direction; if either one passes, I’ll update the model accordingly. For now, the model is allocating district-by-district in Maine and Nebraska and winner-take-all elsewhere.

Here’s last month’s prediction, if you’re interested. As always, if you have any questions, comments, or concerns, let me know.

51 notes

Aug 18 2011 ∞

As requested, here are the election prediction maps for the 2012 US Presidential election. Click on the maps to enlarge them.

Because it’s not clear whom the Republican candidate is going to be, I’ve mapped things out for Rick Perry, Mitt Romney, Michelle Bachmann, and Ron Paul. I’m using all the two-way matchup polling data I could track down from the last two months. This is hopefully going to be a semi-regular feature around here, depending on the availability of polling data and also on how the nomination race progresses. 

The model I’m using here is called national uniform swing. Basically, it works like this: I take national level polling data, and then assume that all the shifts in national numbers since 2008 correspond to uniform shifts across the entire country. (I realize this is a big and dubious assumption, but in practice a national uniform swing model does a pretty decent job of predicting changes. Once the Republican candidate is determined then it will be more worthwhile to look at state-level data.) Anyway, after that I use the margins of error on the polling data to generate win probabilities for each state, and then translate that into electoral college votes. Nebraska and Maine’s special system for allocating electoral votes is included here, but I haven’t drawn the results on the map for the sake of clarity.

Under current polling, note that Romney is the only viable candidate. The election is still a long ways away, and obviously lots of things could happen to shift the polling (e.g., a big new war, large-scale bank failures, Rapture, raptors) and also other candidates like Huckabee could conceivably enter the race - but right now, Romney is the only Republican candidate who has any chance of winning. Bachmann and especially Perry still have relatively low name recognition, so that could change as more people learn about them, but right now Romney is by far the most competitive candidate the Republicans could nominate.

Interestingly, with Romney as a candidate it’s entirely possible that the electoral college returns a tie, with 269 electors for each of Obama and Romney. You can read what ilyagerner has to say about how that would be handled. It would not be pretty.

Also, many states are entirely locked out of consideration in the presidential election, because they have a greater than 99% chance of sticking with the same party as last time. Only the following states (or districts) went with the same party as last time with probability less than 99% under any scenario:

  • Arizona
  • Florida
  • Georgia
  • Indiana
  • Iowa
  • Mississippi
  • Missouri
  • Montana
  • Nebraska (and its first and second districts)
  • Nevada
  • New Hampshire
  • North Carolina
  • North Dakota
  • Ohio
  • Pennsylvania
  • South Carolina
  • South Dakota
  • West Virginia

Don’t see your state on that list? Then there’s basically no Republican candidate who could make your state at all competitive. And many of these states only come into play if Romney is the Republican candidate. Notably missing are four of the five most populous states, which collectively account for about 100 million people.

If you have any questions or feedback (especially on the data visualization, because that’s something I’m really trying to work on), please let me know.

332 notes

Aug 15 2011 ∞
Here are all the states for which exit polling was conducted for the Republican primary in 2008. You can click through to see what Nate Silver has to say about the whole process and why it matters that Iowa is really super conservative and New Hampshire is somewhat moderate and so on. But what I’m unclear on is why this happens - why is Iowa SO much more conservative than any other primary base? I mean, yeah demographics in a state with a high proportion of rural white evangelical Christians, but there are other nearby states with similar demographics and all, and Iowa still has half the proportion of moderates of the next most conservative state. What’s going on here?

Here are all the states for which exit polling was conducted for the Republican primary in 2008. You can click through to see what Nate Silver has to say about the whole process and why it matters that Iowa is really super conservative and New Hampshire is somewhat moderate and so on.

But what I’m unclear on is why this happens - why is Iowa SO much more conservative than any other primary base? I mean, yeah demographics in a state with a high proportion of rural white evangelical Christians, but there are other nearby states with similar demographics and all, and Iowa still has half the proportion of moderates of the next most conservative state. What’s going on here?

6 notes

Aug 12 2011 ∞

How JPMorgan Made Jefferson County Toxic

JPMorgan Chase & Co. (JPM)’s Charles LeCroy said the key to landing bond deals in Jefferson County, Alabama, was finding out whom to pay off. In one example, that meant a $2.6 million payment to Bill Blount, a local banker and longtime friend of County Commissioner Larry Langford.

“It’s a lot of money, but in the end it’s worth it on a billion-dollar deal,” LeCroy told a colleague in 2003, according to a complaint filed by the Securities and Exchange Commission.

That’s because in the $2.9-trillion market for state and local government debt, where 80 percent of all financings are negotiated in private, conflicts of interest prevail. While Langford and Blount are in jail, LeCroy is fighting an SEC action. JPMorgan, which provided most of the toxic debt that devastated Jefferson County, has suffered no loss of business as the nation’s third-largest underwriter of municipal bonds, according to data compiled by Bloomberg.

Just 21 months ago, JPMorgan agreed to a $722 million SEC settlement to end a case over secret payments to friends of Jefferson County commissioners. The financings arranged by JPMorgan, a package of floating-rate debt and derivatives, exposed taxpayers to the 2008 credit crisis and dealt a blow that may lead the county to approve the biggest U.S. municipal bankruptcy as soon as today.

Wow. This is impressively evil on the part of JP Morgan, although really it’s the commissioners’ fault too. Even the ones who weren’t massively corrupt were totally incompetent. I’m not even going to try to explain how the deal worked, but basically they tried to finance a new sewer system using an elaborate system of securities and derivatives. Along they way they managed to pay JP Morgan, Bear Stearns, Bank of America, and Lehmann around $120 million in fees, which is pretty ridiculous on a $3 billion project.

Anyway, now the county is going bankrupt and will therefore be unable to finance normal operations for decades, and JP Morgan is doing just fine. That seems to be par for the course on how these things go lately.

5 notes

Page 1 of 6