December 9, 2010

"Approval By Numbers: How Obama Compares To Past Presidents."

A chart graphing the popularity of 11 Presidents over a 2-year time line.

Click forward to see individual comparisons, such as Bush 43 and Obama. The closest similarity is to Jimmy Carter.

64 comments:

kent said...

The closest similarity is to Jimmy Carter.

As well it should be.

Original Mike said...

Poor Jerry Ford.

former law student said...

The closest similarity is to Jimmy Carter.

Right. Obama's been consistently more popular than Reagan was.

garage mahal said...

Obama more popular than Clinton or Reagan. Carter was more popular than Reagan at the same time. That's not how we remember it now.

former law student said...

Whatever happened to that Reagan guy, anyways?

Scott M said...

Obviously a racist graph and incorrect at that. President Obama's color is solid black.

Original Mike said...

He saved the world, and your ass, FLS.

Original Mike said...

Obama wants to be more like Ike.

Toad Trend said...

I see a graph like this having limited value.

Triangle Man said...

Carter? The closest trend is Reagan, as mentioned in the article. The take home message is that approval ratings are meaningless.

Original Mike said...

Reagan started at 50%. Obama started at 70%.

Phil 314 said...

Gosh, we sure loved our Democratic presidents in the 60's. At least until

former law student said...

He saved the world, and your ass, FLS.

He single-handedly reversed our progress on decreasing our dependence on foreign oil and on reducing the national debt. But he did preside over our evergrowing trade imbalance, setting the stage for our current total dependence on Red China.

On the other hand we did whoop Granada's ass, making the world safe for offshore schools of medicine.

Scott M said...

Whatever happened to that Reagan guy, anyways?

He's dead, Jim.

David said...

You can make a case that Reagan is just as close a comparison.

Interesting that there are less swings in Obama's numbers. Just a slow steady decline.

alan markus said...

Now I know where that global warming hockey stick went. (see Bush 43/Obama Comparison).

David said...

On the other hand, until recently, Obama has had an adoring press, while the press was not friendly to Reagan.

Good Old Ike, man. People liked him for many good reasons. Mostly because they trusted him.

Revenant said...

Obama's been consistently more popular than Reagan was.

Comfort yourself with that thought.

coketown said...

The silver lining is the inevitable post-presidency bounce that most presidents get. Maybe he should skip 2012 and try again in 2016 when we've had a chance to recover. You know: lick our asses to get the bad taste out of our mouths.

Paul said...

Hahaha... Jimmy Carter.

Well I tell you that is the BEST case Obama can have with the way things are going.

I suspect it will, as River said in Serenity, "get worse, much much worse". And the Reavers will eat up Obama's term.

David said...

[Reagan] single-handedly reversed our progress on decreasing our dependence on foreign oil and on reducing the national debt. But he did preside over our evergrowing trade imbalance, setting the stage for our current total dependence on Red China.

The longer your posts, FLS, the more the bullshit.

1. We were not making progress in reducing dependence on foreign oil. But we did import somewhat less because the economy sucked under Carter.

2. Reagan did not single handedly do anything. No President does or can.

3. Our trade imbalance and the growth of power and influence of China are a result of big wave historical forces, many of which were greatly amplified by the American victory in WW II. We can no more stop the rise of Asia and the rest of the developing world than we can regulate the rise of the sun. We have to live with it and manage it.

4. We are not dependent on China, but interdependent with it. Big, crucial difference.

5. A pox on all their houses (Reagan included) regarding the growth of the national debt. But at least Reagan (and Volcker) gave us an economy that could sustain a reasonable level of debt. Now the debt has grown out of control (poxes all around) and the economy may not be strong enough to withstand the numerous unattractive options we have.

On the other hand, Obama said yesterday that the tax cuts would create "millions of jobs." Who does that sound like?

pm317 said...

He is in free fall!

veni vidi vici said...

Tends to demonstrate that the first two years' approval ratings don't mean much of anything. In one slide they say Barry's nearly identical to Reagan, then the next they say nearly identical to Carter. That's enough to make it unnecessary to read further.

kent said...

Interesting that there are less swings in Obama's numbers. Just a slow steady decline.

Yup. And it ain't gettin' any better for him in the immediate future, either:

Most Americans Say They're Worse Off Since Obama Took Office, Poll Shows

The American electorate, a-snooze, tends to awaken slowly, and only in fits and starts; but, once it inevitably does...

ricpic said...

Obama is sui generis bad.

Ankur said...

If you do a statistical variance model using the raw data, you will find that actually, the closest similarlity is with Ronald Reagan. Remember, when you're doing a variance model, you square the differences - and the squares of negative number is a positive number.

http://www.talkingpointsmemo.com/gallery/2010/12/approval-by-numbers-how-obama-compares-to-past-presidents.php?img=5

Ankur said...

It is interesting that someone above mentioned Volcker's work in the Reagan presidency in positive terms. And I agree completely.

I wonder how you feel about Volcker now, though. Do you still think he is a great economist? Or does your opinion of his economic ideas depend on which president he is working for?

former law student said...

If you do a statistical variance model

Law profs can be all but innumerate -- it doesn't seem to hinder them.

Unknown said...

The Reagan guy's thing in Grenada stopped the advance of Communism in this hemisphere cold in its tracks (know you always forget that, fls). He also brought down the Warsaw Bloc and created the longest peacetime boom in history.

Mere details, I know.

PS TPM is trying to whistle in the graveyard. Nobody else on the graph had the problems The Zero has at this point (and mostly of his own making).

Original Mike said...

"If you do a statistical variance model using the raw data, you will find that actually, the closest similarlity is with Ronald Reagan.Remember, when you're doing a variance model, you square the differences - and the squares of negative number is a positive number."

Great example of not knowing the right statisitical test to use to analyze your data.

Ankur said...

Really, Original Mike? If you were analyzing to time series graphs, how would you compare them, if not by taking a point by point variance, and calculating the r-squared?

Ankur said...

to=two

kent said...

Law profs can be all but innumerate -- it doesn't seem to hinder them.

Heck, if you're a Democratic senatorial backbencher with a suitably "exotic" name/pedigree, and an energetically complicit MSM behind you: it can net you a sweet four rent-free years in the Lincoln Bedroom. Easy.

The Drill SGT said...

The Reagan-Obama chart may be the best fit, but the one that really has to leave a mark is the Nixon -Obama chart :)

Original Mike said...

A variance test will give you the same answers for two graphs. One that goes monotomically up, and another that goes monotomically down by the same amount. Those two graphs are about as different as you can get (in the case of Presidential popularity, one guy started low and finished high, the other guy did the opposite), yet the variance result comes out the same. In this case, do you think variance is a good tool to characterize the two graphs?

Ankur said...

Wrong. A graph that goes monotically up versus a graph that will go monotonically down will not have low variance.

The data points on either extremes will be different enough that squaring them will blow up the variance nicely.

Do you know what variance analysis is? You square the delta-y (difference between Y values in two graphs for the same value of x). And you add up all the delta-ys for all the data points. Then you normalize, based on the number of data points you used.

Pray explain how a monotonically increasing graph would show low variance vis-a-vis a monotonically decreasing graph in the same time period.

Ankur said...

And the "starting point" as you point out is one data point in a 24 point time series. 22 points of that time series are in line with each other. The first 2 are not.

Now, one might argue that in the case of presidential popularity - data points 1 and 2 deserve more weight than the remaining 22 data points. And that would be a valid argument.

But that is not a statistical numbers based argument. That is sociology. And sociology is not math.

Original Mike said...

I use thm every day, Ankur. In fact I teach noise analysis. I can construct a realistic example, pertinent to the Presidential popularity graphs in question, that contradicts your 2nd paragraph. And I can construct a pathological case that makes you correct.

Variance test are used to quantify deviations about a mean (i.e. noise). They are not very useful to characterize trends.

Ankur said...

We are not talking about variance around a mean here. We are talking about variance of two Ys around each other.

Drawing a trendline for both charts shows very similar slopes and intercepts as well - precisely because math doesn't weigh Points 1 and 2 more than points 3 through 24.

I am sure however, that you will question that analysis as well. So here is my question for you: What statistical model would YOU use to compare two time series graphs? Show/tell me - and maybe I will learn something.

Ankur said...

If, however, your point is that that monthly popularity indices are, in and of themselves, an erroneous indicator for presidential performance - then I couldn't agree more. That is evidenced by the look at the close similarity of Obama with Reagan, Carter, Nixon and Clinton.

You can question the ab initio usefulness of the measure itself - and I would agree. But if that measure is being used, and given credence, then questioning the perfectly fine mathematical analysis that follows simply indicates ideology. Math is non ideological.

Original Mike said...

"precisely because math doesn't weigh Points 1 and 2 more than points 3 through 24."

And that's a limitation of the math. You don't think the starting and ending points of Presidential popularity are important when comparing the popularity of two Presidents?

I mean, the best thing to do is just LOOK at them. Mathematical analysis is only useful when it illuminates things, not when it obscures them.

former law student said...

He also brought down the Warsaw Bloc

Really? I'd like to introduce you to a man I like to call Lech Walesa.

But Austrian Foreign Minister Alois Mock, and his Hungarian counterpart Gyula Horn, literally tore open the Iron Curtain, at the Pan-European picnic in 1989.

stopped the advance of Communism in this hemisphere cold in its tracks

By giving arms to the Ayatollah's Iran.

created the longest peacetime boom in history.

Sorry, the Clinton boom lasted two years longer.

Phil 314 said...

Uh oh, a statistician fight. That can never end well.

Kirby Olson said...

Bush 2 is by far the highest, and the best.

kent said...

Uh oh, a statistician fight

... otherwise known as a "pi fight."

[::rimshot::]

Scott M said...

pi fight

Hysterical.

Ankur said...

Sure it is. But what I "think intuitively" is irrelevant when you are doing numeric analysis.

You are trying to compare two sets of NUMBERS. Either suggest a different and valid numerical model to compare/contrast them - or acknowledge that numbers might not be the appropriate way to look at this anyway.

"Visually" looking at charts might appeal to the layman, but has zero mathematical validity. So, I apologize, but I cannot give credence to your exhortation to "just look" - because looking, by definition, involves a value judgment and is prone to ideological bias. You and I might see different things when we "look".

Anyway - to summarize what I am saying - I see two ways out of this conundrum:

a) Suggestion of a different mathematical model which might 'illuminate rather than obscure'

or

b) Acknowledging that numerical analysis is not the way to go at all, in which case one shouldn't be looking at charts based on numbers anyway.

Ankur said...

Hahaha. Kent wins the thread.

garage mahal said...

In fact I teach noise analysis

Nerd!

MadisonMan said...

I'd say the closest similarity is Reagan. There is a difference right at the beginning, but after that they track very nicely.

Original Mike said...

With respect (and I realy mean that, Ankur) I see this in my incoming students; substituting mathematical analysis for, in my field, physical anaylsis (i.e. explaining what is going on in their data).

Yes, Kent wins the thread.

MadisonMan said...

There are lots of Ws in that plot, by the way. Subliminal?

Ankur said...

And how do you educate them? by asking them to "look at the graphs"? or my suggesting alternative methods of numerical analysis?

Again, we are dealing with numbers here. Not feelings. Don't you think there should be a numbers based way of analyzing this?

If you're 'explaining what's going on with the data' without using math, then what you are doing is conjecture. But if you are explaining using a different mathematical model, then educate me. And I do mean that respectfully as well.

If you don't think math is the right tool to analyze these charts, just say it! That is a perfectly valid opinion - but it is still an opinion.

Ankur said...

Also...when I said above "If you're 'explaining what's going on with the data' without using math, then what you are doing is conjecture" - ONLY refers to the charts we are discussing here. Not to your field of work.

Original Mike said...

"If you don't think math is the right tool to analyze these charts, just say it!"

I thought I had.

Ankur said...

And, by the way, the reason I keep harping on about math is simply this: People have found ways to politicize every branch of human knowledge...including Physics. Physics, for chrissakes!

Math seems to be the last neutral vestige, free of opinion and bias.

GMay said...

A thalydomide baby pi fight?

Kent, you're on fire today.

Original Mike said...

"Math seems to be the last neutral vestige, free of opinion and bias."

Math maybe, but certainly not statistics.

Fen said...

"Reagan created the longest peacetime boom in history."

FLS: Sorry, the Clinton boom lasted two years longer.

Ha. That was still Reagan's boom. Clinton got it in the form of a Peace Dividend.

Anonymous said...

He single-handedly reversed our progress on decreasing our dependence on foreign oil and on reducing the national debt.

I think you should go on pretending the Democrats didn't control the house for the entire Reagan presidency.

Fen said...

Surprised Clinton's line is so low.

I remember him being much more popular.

David said...

"I wonder how you feel about Volcker now, though. Do you still think he is a great economist?"

I think he was a great chairman of the Federal Reserve. Bernanke is a great economist, but how good a chair he is remains to be seen.

As for now, there are two ways to look at Volcker:

1. Irrelevant old man hanging on to his last shred of former glory.

2. Wise old man biding his time and waiting to pick up the pieces.

I can't decide between the two.

Volcker can decide. That's why he was great.

Revenant said...

There is no such thing as "foreign oil". There's just "oil".

former law student said...

There is no such thing as "foreign oil". There's just "oil".

Similarly, there is no such thing as a "current account deficit." There is just "deficit."