What are Obama's odds of reelection/defeat?

[Mr.Phips]

Based upon the approval rating this far out, assuming no strange events, my guess:

approval%                            win%

40                                       <  1
41                                           5
42                                         15
43                                         30
44                                         50
45                                         70
46                                         83
47                                         90
48                                         94
49                                         97
50                                         99
51                                       >99

At this point the President would need to have some severe and not-yet-known weakness as a leader or get entangled in something devastating to lose -- and such would obviously cut his approval rating drastically.


 

Where do you get these numbers from?

Gerald Ford had approvals of near 50% in 1976, yet he still lost. 

Paying attention to pbrower is certain way to lose brain cells.

As I said in some other forum, it's about 55-45 against for now primarily because we're most likely headed back into negative GDP growth by years end, which means recession, but these things really don't mean very much at this point.  Talk to me in 6 months.

You really think we are headed for another recession by year's end?  If that is the case, Obama would probably lose even to someone like Gingrich, because that would mean rising unemployment throughout 2012.   I am predicting around 2% growth for 2011 and a slowly falling unemployment rate throughout 2012.  Nobody else seems to be predicting a recession like you are. 

We should see a negative GDP quarter in 4Q 2011 or 1Q 2012 based on current trajectories, input costs, capital spending, inventory replentishment, and the like.  Marty Feldstein said a few months ago that the top of this current cycle was 3Q 2010, which fits in fairly well with this call.

1Q of negative GDP growth is not a recession, as defined by NBER, and I am suspicious that the Fed next year will start trying to pump money into the system again to try and keep the status quo in place (as they did in 1980).  So, I am not calling recession, let me correct my post there.  Don't expect employment to improve, though.

One quarter of negative GDP growth would certainly feel like a recession and would almost certainly result in a couple months of job losses.  The only time a quarter of negative GDP growth didnt result in monthly job losses was 1956. 

[Sam Spade]

Based upon the approval rating this far out, assuming no strange events, my guess:

approval%                            win%

40                                       <  1
41                                           5
42                                         15
43                                         30
44                                         50
45                                         70
46                                         83
47                                         90
48                                         94
49                                         97
50                                         99
51                                       >99

At this point the President would need to have some severe and not-yet-known weakness as a leader or get entangled in something devastating to lose -- and such would obviously cut his approval rating drastically.


 

Where do you get these numbers from?

Gerald Ford had approvals of near 50% in 1976, yet he still lost. 

Paying attention to pbrower is certain way to lose brain cells.

As I said in some other forum, it's about 55-45 against for now primarily because we're most likely headed back into negative GDP growth by years end, which means recession, but these things really don't mean very much at this point.  Talk to me in 6 months.

You really think we are headed for another recession by year's end?  If that is the case, Obama would probably lose even to someone like Gingrich, because that would mean rising unemployment throughout 2012.   I am predicting around 2% growth for 2011 and a slowly falling unemployment rate throughout 2012.  Nobody else seems to be predicting a recession like you are. 

We should see a negative GDP quarter in 4Q 2011 or 1Q 2012 based on current trajectories, input costs, capital spending, inventory replentishment, and the like.  Marty Feldstein said a few months ago that the top of this current cycle was 3Q 2010, which fits in fairly well with this call.

1Q of negative GDP growth is not a recession, as defined by NBER, and I am suspicious that the Fed next year will start trying to pump money into the system again to try and keep the status quo in place (as they did in 1980).  So, I am not calling recession, let me correct my post there.  Don't expect employment to improve, though.

One quarter of negative GDP growth would certainly feel like a recession and would almost certainly result in a couple months of job losses.  The only time a quarter of negative GDP growth didnt result in monthly job losses was 1956. 

If Feldstein is right, then the top in unemployment should normally occur 6-9 months later, which is, well, now.

Consider that in April, the birth-death adjustment was 175K (which is laughable, but the government tends to do this in April-June period), and the actual job gain was 244K.  In other words, I doubt the gain was much over 150K.  March's numbers looked better than that actually.

Which means that...

Anyway, we shall see.  It's hard to make predictions now, as I initially said, so take this all with a grain of salt.  Except for the GDP call.

[Mr.Phips]

Based upon the approval rating this far out, assuming no strange events, my guess:

approval%                            win%

40                                       <  1
41                                           5
42                                         15
43                                         30
44                                         50
45                                         70
46                                         83
47                                         90
48                                         94
49                                         97
50                                         99
51                                       >99

At this point the President would need to have some severe and not-yet-known weakness as a leader or get entangled in something devastating to lose -- and such would obviously cut his approval rating drastically.


 

Where do you get these numbers from?

Gerald Ford had approvals of near 50% in 1976, yet he still lost. 

Paying attention to pbrower is certain way to lose brain cells.

As I said in some other forum, it's about 55-45 against for now primarily because we're most likely headed back into negative GDP growth by years end, which means recession, but these things really don't mean very much at this point.  Talk to me in 6 months.

You really think we are headed for another recession by year's end?  If that is the case, Obama would probably lose even to someone like Gingrich, because that would mean rising unemployment throughout 2012.   I am predicting around 2% growth for 2011 and a slowly falling unemployment rate throughout 2012.  Nobody else seems to be predicting a recession like you are. 

We should see a negative GDP quarter in 4Q 2011 or 1Q 2012 based on current trajectories, input costs, capital spending, inventory replentishment, and the like.  Marty Feldstein said a few months ago that the top of this current cycle was 3Q 2010, which fits in fairly well with this call.

1Q of negative GDP growth is not a recession, as defined by NBER, and I am suspicious that the Fed next year will start trying to pump money into the system again to try and keep the status quo in place (as they did in 1980).  So, I am not calling recession, let me correct my post there.  Don't expect employment to improve, though.

One quarter of negative GDP growth would certainly feel like a recession and would almost certainly result in a couple months of job losses.  The only time a quarter of negative GDP growth didnt result in monthly job losses was 1956. 

If Feldstein is right, then the top in unemployment should normally occur 6-9 months later, which is, well, now.

Consider that in April, the birth-death adjustment was 175K (which is laughable, but the government tends to do this in April-June period), and the actual job gain was 244K.  In other words, I doubt the gain was much over 150K.  March's numbers looked better than that actually.

Which means that...

Anyway, we shall see.  It's hard to make predictions now, as I initially said, so take this all with a grain of salt.  Except for the GDP call.

The only reason I am skeptical of your GDP call is because the yield curve usually inverts about 12-14 months before GDP growth turns negative.  Now, I dont know if the situation is different now because the fed is holding the yield curve positive through its zero bound rates, but this would be the first time we had a quarter of negative GDP growth without the yield curve first inverting.

[Eraserhead]

100% chance.

Oh please.

What? I'd agree with him that there's a 100% chance that Obama will be reelected or defeated in his bid for reelection!

[pbrower2a]

Based upon the approval rating this far out, assuming no strange events, my guess:

approval%                            win%

40                                       <  1
41                                           5
42                                         15
43                                         30
44                                         50
45                                         70
46                                         83
47                                         90
48                                         94
49                                         97
50                                         99
51                                       >99

At this point the President would need to have some severe and not-yet-known weakness as a leader or get entangled in something devastating to lose -- and such would obviously cut his approval rating drastically.


 

Where do you get these numbers from?

Gerald Ford had approvals of near 50% in 1976, yet he still lost.  

I recall a piece by Nate Silver in which he analyzed the chances of a Governor or Senator winning re-election based upon his approval rate six months before the election. The break-even point was 44%, and the chance of re-election dropped to near-zero at 40% (which suggests that the Governor or Senator has problems other than bad luck) and near 100% at 50%.  Such is before any campaigning begins, and incumbency usually has an advantage. It can't save a low-achiever (let us say Rick Santorum in 2006 or John Corzine in 2009). A Governor or Senator usually gains about 6% with an active campaign; being Governor or Senator usually means that one knows how to campaign effectively. This might not apply to someone appointed to office.

Now, why Governor or Senator? Because with few exceptions, US Presidents used to be Governors or Senators. Counter-examples in the past century include George H.W. Bush, Gerald Ford, Richard Nixon, Dwight Eisenhower, and Herbert Hoover. There is no polling data for Herbert Hoover, but the literary evidence suggests that he was extremely unpopular by 1932. Eisenhower was in on his record as a heroic general, and he vastly underperformed in the polls what his approval ratings suggested -- he did little active campaigning and didn't need to.  Ford was appointed VP after having no record of ever running for a statewide office; he had one of the most incompetent campaigns for re-election ever and lost to a weak challenger. The elder Bush had never held a statewide office and, despite some superb achievements of foreign policy, couldn't offer a clear vision of what he would do in a second term.  

Luck can be events for which the President has little control.  Harry Truman probably had approval ratings closely linked to the latitude of the front line in Korea, and won re-election when the front line seemed securely  north of the 38th Parallel.  When UN forces were confined to the Pusan redoubt and seemingly on the brink of expulsion, Truman was unelectable. Approval ratings for President Truman swung wildly, and by 1952 he was unelectable during a Commie advance. The economy is even more blatant. It's hard to see how any President can dictate how the stock market or unemployment stats go. None can ever prevent the collapse of a speculative boom and the ugly consequences that follow. To be sure, Jimmy Carter had made much of the stagflation during the Ford Presidency, but he could do nothing to stop it -- especially as late-wave Baby Boom kids entered the workforce in huge numbers and saw few good economic opportunities but faced rising costs of housing and gasoline.  But Carter got caught with the Iranian hostage crisis that dragged on with no resolution.

All in all, a Presidential campaign comprises fifty statewide races of various levels of importance (Texas matters much more than Wyoming), one district-wide race (DC, even though that is a given), and five decisions on some Congressional districts (Republicans aren't going to split Nebraska and win or split Maine and lose). Some politicians are unusually-adept campaigners; some aren't. No level of campaigning could have rescued Hoover or Carter, and it's hard to see how Eisenhower could have lost. To be sure, what works in Massachusetts in recent years and what works in Indiana are vastly different.

I'm guessing on the shape of the curve, and I don't have a perfect fit for the numbers. The best that I can do is to guess. It's possible that an incumbent who has some bad luck (a weak economy in his state) at some point that improves might see some gains and win. It is also likely that some politician with very low approval ratings six months before the election has problems beyond any resolution in campaigning -- like a cloud of rumors of corruption, excessive and inappropriate partisanship, or an aloof character. Some politicians are simply poor fits for their bailiwicks and got elected due to freakish circumstances that might not be repeated.  Whether challengers are unusually strong or weak is a matter of individual luck.

Maybe the relevant question is how likely the President is to win a State if his approval is at a certain point. If the President has an approval rating of 38% in Texas, then he probably should not even try to campaign there. If his approval rating in Massachusetts is 60%, then don't expect him to appear there to try to get 70% of the vote instead of 55% of the vote.  Recently I saw an approval rating of 44% for the President in Missouri, suggesting that the state projects to vote much as it did in 2008. That suggests that he is in fairly good shape nationwide. But why would he campaign in Missouri if he doesn't need the state? Very simply, the state has a US Senator up for re-election , and his presence might make the difference in Missouri -- and for control of the US Senate.  

All in all I am estimating. There's just not enough data for a precise fit. It looks much like a titration curve for an acid-base reaction as an analogue. If you know any college-level chemistry, you have a fair idea of what happens to pH of a solution of 1.0 M HCl as you add  a solution of 1.0 M NaOH to it. As the level of NaOH put in is much less than the level of HCl, the pH budges little, and after the NaOH overwhelms the HCl of the original solution, the pH gets stuck at a high level. But in the area around neutralization -- when the relevant quantities of HCl and NaOH are roughly equal, then the pH level rises abruptly in a very narrow range. I just don't have the equation of the curve, and my data points are more to fit the style than anything else.

(If you know nothing about college chemistry, then you probably never got the opportunity to see the reaction first-hand, especially since those chemicals are too dangerous for most people to ever experiment with. Just trust me on that). In that analogy, "losing" is the analogue of a low pH or an acid solution and "winning" is the analogue of a high pH or an alkaline solution. Neutrality is a pH of 7 or an exact split of the relevant vote.  

The 'average' incumbent politician gains about 6%, although that is no sure thing (George Allen, former Senator from Virginia, who lost for non-random reasons and the random chance of facing an unusually-strong opponent as a challenger) can have an odd conjunction of folly and bad luck. If the President should have a personal scandal or endure either another economic meltdown or a diplomatic/military debacle on his watch, then his ability as a political campaigner might prove irrelevant. President Obama seems in reasonably-good shape for winning re-election, as his nationwide approvals curve is hard to distinguish from those of Ronald Reagan or Bill Clinton, Presidents who seem good analogues fore reasons other than my desire that they be good analogues. But he can still lose due to events that nobody can now see or predict.

[memphis]

100% chance.

Oh please.

What? I'd agree with him that there's a 100% chance that Obama will be reelected or defeated in his bid for reelection!

No. Obama is not immortal. There is not a 100% chance he will be either reelected or defeated next year.

[sentinel]

100% chance.

Oh please.

What? I'd agree with him that there's a 100% chance that Obama will be reelected or defeated in his bid for reelection!

No. Obama is not immortal. There is not a 100% chance he will be either reelected or defeated next year.

The day after the election there is a 100% chance that he won or lost the night before. It's kinda like when the weatherman says its going to rain on the day of your event --when it starts raining there is a 100% chance that it is raining.

[pbrower2a]

100% chance.

Oh please.

What? I'd agree with him that there's a 100% chance that Obama will be reelected or defeated in his bid for reelection!

No. Obama is not immortal. There is not a 100% chance he will be either reelected or defeated next year.

The day after the election there is a 100% chance that he won or lost the night before. It's kinda like when the weatherman says its going to rain on the day of your event --when it starts raining there is a 100% chance that it is raining.

Prime example: back in 2008 I once predicted that the election of Barack Obama was effectively at the point at which he had 264 electoral votes locked up with several states really deciding the election -- Colorado, Florida, Missouri, Ohio, and Virginia -- as independent events, so to speak. (Indiana and North Carolina were possible, but Obama wasn't going to win either without winning one of the five states above). If he won one of them and  he had one chance in two of winning either one, then his chance of winning was effectively 31 in 32.  That's a 96.875% chance of winning.  Not 100%. The assumption is that with two months to go, politicians just don't lose leads larger than 8%. One month to go? Politicians don't lose 7% leads, which put Colorado and Virginia out of contention.

[Small Business Owner of Any Repute]

The assumption is that with two months to go, politicians just don't lose leads larger than 8%. One month to go? Politicians don't lose 7% leads, which put Colorado and Virginia out of contention.

What? Fast and hard narrowing happens all the time. Roy Barnes was leading by 11 the week before the election. Assuming a fast narrowing can't happen leads to "dumb" political analysis.

[opebo]

Assuming they don't manage to fake some hotel-room rape his chances at this point are about 55% of being reeleted.

[Likely Voter]

I think the current intrade price for Obama to win of 60.9 is just about the right %

[I spent the winter writing songs about getting better]

I think what sirnick was trying to say is that there is a 100% chance that Obama will either be re-elected or defeated.

Which is not entirely true, it is possible he ends up not running for re-election after all or resigns or dies whatever, not likely, but certainly above a 0% chance. So probably closer to a 99% chance.

[DS0816]

The assumption is that with two months to go, politicians just don't lose leads larger than 8%. One month to go? Politicians don't lose 7% leads, which put Colorado and Virginia out of contention.

What? Fast and hard narrowing happens all the time. Roy Barnes was leading by 11 the week before the election. Assuming a fast narrowing can't happen leads to "dumb" political analysis.

Midterms and presidentials aren't the same.

[Eraserhead]

I think what sirnick was trying to say is that there is a 100% chance that Obama will either be re-elected or defeated.

Which is not entirely true, it is possible he ends up not running for re-election after all or resigns or dies whatever, not likely, but certainly above a 0% chance. So probably closer to a 99% chance.

Fair enough, sir.

I will put my odds down at 64% - reelected, 35% - defeated, 1% - other.

[Lou34679]

I could be wrong, but unless there's a double dip, shouldn't unemployment at least  7.9% by fall of 2012? I know it's a slow recovery, but even at this pace, I don't know how it's not under 8% by the fall of 2012, unless there's a double dip.