President Barack Obama’s job approval rating has been steadily declining over the past several months at the same time that the Dow Jones Industrial Average has been steadily increasing. This inverse relationship between job approval and the stock market is not a common pattern with recent presidents. In fact, Gallup’s analysis of job approval ratings and the Dow going back to 1977 finds few sustained periods where the Dow and presidential approval were strongly related (either positively or negatively)…
A wide variety of factors affect presidential job approval ratings, and a similarly wide array of variables are responsible for the rising and falling of the stock market. It is possible that awareness of a president’s popularity could affect how people relate to the stock market, or that awareness of changes in the stock market could affect how people view the president. But clearly, many other factors could influence trends in both measures, including the very real possibility that what appears to be a relationship between the two is merely a coincidence.
You must be logged in to post a comment.
Blowback
Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Registration is currently closed. That means if you're not already registered, you can't comment. We will let you know if and when registration re-opens. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.
Trackbacks/Pings
Trackback URL
Comments
I didn’t necessarily agree with Kramer (sp?) but wasn’t it the disapproval rating?
LastRick on August 28, 2009 at 5:14 PM
It was the disapproval rating .. you’re right.
skree on August 28, 2009 at 5:19 PM
But they didn’t address if it was funny or not.
exception on August 28, 2009 at 5:20 PM
Yeah, the link to Ed’s article is here. We get this line from the original article:
But Gallup is clearly plotting the approval rating versus DJIA.
These two ratings, approval and disapproval, are completely independent from one another. You can see that at RCP: a drop or rise in approval does not necessarily correspond to a reaction in the disapproval. Why? Because sometimes people go from “approve” to “well, I’m not sure”.
I still don’t necessarily agree with Cramer; I looked at the correlation factor back when this original idea posted and I was unimpressed. But Gallup could have strengthened their case by using the correct rating.
LastRick on August 28, 2009 at 5:23 PM
Must be my lying eyes. The trend lines on that graph sure look inversely related to me.
obladioblada on August 28, 2009 at 5:23 PM
Wow, a “modern” polling company that has failed to keep up with the way information gets out and how people use it.
Of course no President in recent history has pushed such anti-business policies as this one, but we have historical evidence to back up our claims!
David
LifeTrek on August 28, 2009 at 5:25 PM
I’m confused. Did someone ever suggest there was?
lorien1973 on August 28, 2009 at 5:26 PM
a pollster getting it wrong? Who’d have thunk it?
gatorboy on August 28, 2009 at 5:26 PM
Cramer’s was Obama disapproval vs. the S&P 500 not the Dow. The S&P 500 is a broader measure than the Dow which just measures 30 industrial stocks. This Gallup piece is therefore dishonest.
levi from queens on August 28, 2009 at 5:26 PM
Yeah like Wall Street isn’t sorry to see HIM FAILING.
anniekc on August 28, 2009 at 5:29 PM
Great catch. Not only did they screw up on the approval/disapproval part, they used “Closing Dow” (DJIA?) instead of S&P. The ol’ double FAIL.
LastRick on August 28, 2009 at 5:32 PM
Hey, Barry did have relations with Dow Jones. Wait till you see the baby. Their going to name her Destiny.
portlandon on August 28, 2009 at 5:32 PM
Obviously these “ratings” are aimed at different targets and timeframes. The selected stock indices (S/P and Dow) look at the future. The political polling is now or yesterday.
Those stock indices look how BIG business earnings will fare down the road. Often the earnings of BIG business (think GE getting a guaranteed lightbulb market) does not correlate with small business (squeezed by governmental actions/taxes that actually BENEFIT the BIG boys)…. Stock markets are not necessarily main street. Euorpe over the last 20 years is proof that the economy can suck and the BIG business stock market can perform okay.
clorensen on August 28, 2009 at 5:36 PM
I think the likelihood of Obamacare failing and the stock market are directly related.
Hoodlumman on August 28, 2009 at 5:37 PM
I haven’t dug out my statistics textbooks, but something in their described methodology stinks. Is it just me, or are they calculating an autocorrelation when they should be running a cross-correlation?
Blacksmith on August 28, 2009 at 5:42 PM
I agree. But you should be able then to slide one set of data along the x-axis to match them up.
And as I type this, I remember what is now a third problem with this article from Gallup: Cramer said to begin only at the minimum in March. If there’s a worst-fit part of these two lines, it’s definitely the time period up to March; after that point, I think they fit better.
Still, just b/c they fit doesn’t infer causation. X doesn’t have to cause Y; X and Y can be caused by Z.
LastRick on August 28, 2009 at 5:44 PM
I agree; DJIA and Obama’s rating are probably correlated. His policies are, in general, bad for the people with money in the markets. So if Obama trying to take money from you, it stands to reason you might be against that.
They definitely should be doing cross-correlation. Why do you think they are doing auto-correlation (the comparison of data to itself)?
LastRick on August 28, 2009 at 5:48 PM
It’s the Marxist Variable that is causing the inverse relationship
runner on August 28, 2009 at 5:50 PM
But you can’t slide them all to one side and put them all on the same X-axis; the data are all compared to the average value for their respective POTUS’ term-of-office (so for Carter and Bush One, you’re averaging over 4 years, not 8!)
And yes, that’s why I thought they were doing the auto-correlation instead of cross; I see nothing in their methodology that says they ever compared the popularity (nevermind the UNpopularity they should’ve used to validate Cramer’s claim!) against the closing-value on the DJIA. The whole thing looks like a godawful unruly mess, that some “poll analyst” fresh out of J-school tried to baffle their editor with BS, and it worked.
Blacksmith on August 28, 2009 at 6:03 PM
To elaborate, from the Gallup story:
So what you’ve got is [popularity]/[POTUS-term average of popularity] plotted against time, and [DJIA closing]/[POTUS-term average of DJIA closing] also plotted against time – but nowhere is there the analysis of the correlation between them!
Blacksmith on August 28, 2009 at 6:08 PM
There are no coincidences. What’s bad for Obama is good for the USA and the markets.
BottomLine5 on August 28, 2009 at 6:19 PM
We were talking about completely different things. I understand you now; you were talking about comparing each president to one another. I meant (by sliding) that the S&P 500 for Obama’s term could be slid on the x-axis to compare to Obama’s approval rating. Is that what you meant by “autocorrelation”? If so, I think that’s an incorrect interpretation of the term. Autocorrelation, I understand it, means (for ex.) comparing the S&P 500 to itself to see if there are any recurring patterns (i.e. sinusoidal).
LastRick on August 28, 2009 at 6:29 PM
No, I meant that you can’t line up all these graphs on a shared axis Carter-Reagan-Bush-Clinton-Bush-Obama even if you wanted to, because the calculations are both averaged only over the terms for each given POTUS (so you’ll get a discontinuous set). And yes, they are autocorrelating given your definition; the approval numbers are computed against the average of the approval numbers (ie, against itself) and the DJIA is computed against the average of the DJIA (same deal). So if you’ve got a wide variance in the stock market (Reagan and Clinton and Bush) then you’ll have that nice sinusoidal pattern as the extended terms give you time for that fluctuation. For GHWB, there was some dynamism but the limited timespan means we don’t see quite so strong a sinewave. Then look at Carter where not only was there less time, but there was very little dynamism – and his DJIA is practically flatlined at 1.0 (ie, the average for his term). That’s the very definition of a stagnant economy right there.
Blacksmith on August 28, 2009 at 6:54 PM
And no, I suppose “sinusoid” would be a bit strong of a descriptor for Reagan and Clinton; both started their terms with relatively weak economies that expanded across the duration of their terms, with some softness toward the end; not exactly a sinusoid, but about as close as you’ll get in the markets I suppose.
Blacksmith on August 28, 2009 at 6:58 PM
QFT
MirCat on August 28, 2009 at 7:28 PM
I know this is late, but I’m revising my stance. I’ve been crunching numbers this evening and here’s what I’ve found if anyone’s interested. I’ll lay out my methodology so you can follow it.
1) I took the Gallup data from their website; you can export it as csv to Excel.
2) S&P 500 data was here. I think this is the one Cramer is talking about, but I’m no stock guy, just numbers.
3) I adjusted text to value on the S&P data; began the data at March-9, normalized this day as 1.0 for both the disapproval rating and S&P close. Chose Aug-27 as the end.
4) Some data had to be interpolated (weekends, for example, when there’s no market or if a poll wasn’t taken that day). This produced an 172×2 array, which Excel will correlate for you (remember, no causation, just correlation).
The correlation coefficient for the whole series is 0.78, which is astoundingly high (anything b/w 0.5 and 1.0); there’s a slight improvement to around 0.8 if you only consider after June 19th or so. This coincides with Gallup’s own admission (”Even for Obama, the current pattern is only recent”) that as of late, the trend is very clear. I was trying to think, what happened around June 19th that could make this happen? My first guess: The Iranian elections. Was that the beginning of real trouble for Obama? Any other ideas? (Or where I might have miscalculated?)
Once again, I don’t think Obama’s slide is causing the gains on the market. But the numbers show a very strong correlation which means investors are, at a minimum, happy Obama is having trouble in the polls. Which should be a surprise to no one.
LastRick on August 28, 2009 at 9:13 PM
If you haven’t seen the Wall Street Journal’s “How the Presidents Stack Up” interactive chart- check it out. Click on the presidents (Harry Truman, 1945 through present day) that you want to compare and/or the time frame – there are all kinds of options for comparison.
NightmareOnKStreet on August 29, 2009 at 12:36 AM