Measuring Success

May 18, 2009

Comments

Here is an interesting graphic.

Geoff at Innocent Bystanders notes that we are on path for unemployment as if there was no stimulus plan. Problem is there is a stimulus plan. So either the initial estimates of the effect of the recession, absent a stimulus plan, were overly optimistic, or the stimulus plan is providing no benefit at all. I think we know which explanation President Obama and his Administration are going to rely on.

Now, all snarkiness aside as Greg Mankiw notes, with a shifting baseline projection there is absolutely no way to judge President Obama’s policies as a success or failure. Now this isn’t something inherent in Obama Administration, but is a problem with administrations of all Presidents. No President will set up objective standards by which we can judge their policies as successes or failures. And when out campaigning every policy put in place by a given President is always a success, never a failure.

34 responses to “Measuring Success”

odograph

May 18, 2009 at 1:15 pm

Do you have such confidence in economic modeling, Steve? To really buy a “no stimulus” projection as a “fact” in an “all things being equal” parallel universe?

Economists is not that good, and thinking they are makes them doubly foolish.
Drew

May 18, 2009 at 1:22 pm

“So either the initial estimates of the effect of the recession, absent a stimulus plan, were overly optimistic, or the stimulus plan is providing no benefit at all.”

I know you meant to say: “…were overly pessimistic..”

And I would add..”yet.”

Which brings us to a problem: The usual mis-timing of government “corrective” policy.

And an observation: I have a running 15 year debate with a friend about the role of the media in policy and public perception. My view: media are powerful propagandists, and therefore useful idiots of the pols. His view: “the people are at fault, but they usually get it right.”

Test: How many in the general population will understand the true effects, timing and comparisons to Obama’s “baseline” in the face of slobbering media coverage?

I’m thinking 15%.
Steve Verdon

May 18, 2009 at 1:32 pm

Odograph,

First, an objective standard doesn’t have to be based on an economic model. Second, there is always the truth: “We wont be able to measure the success of this policy.” And finally, Presidents run around taking credit when there is no way of telling if it really was their policies that resulted in good outcomes or just the normal course of things. Also keep in mind that several months ago President Obama was aksed precisely this question. Instead of giving the honest answer he gave us the verbal version of this graph.
Drew

May 18, 2009 at 1:35 pm

odo –

The “without” is standard past recession analysis. I share your healthy scepticism, but I would note that the oh, so Obama-sympathetic Menzie Chin has been pounding these baseline stats at Econbrowser.
Steve Verdon

May 18, 2009 at 1:35 pm

“So either the initial estimates of the effect of the recession, absent a stimulus plan, were overly optimistic, or the stimulus plan is providing no benefit at all.”

I know you meant to say: “…were overly pessimistic..”

Uhhmm…errr…no? I was think optimistic in that the projected rates of unemployment were lower than what they are turning out to be. That is, the without the stimulus should be higher…hence the initial forecast was too optimistic.
Dave Schuler

May 18, 2009 at 1:42 pm

Do you have such confidence in economic modeling, Steve? To really buy a “no stimulus” projection as a “fact” in an “all things being equal” parallel universe?

odograph, I’m puzzled by this comment. Last week Vice President Biden announced that the stimulus had created/saved 150,000 jobs. That pronouncement was based on the administration’s econometric model. By reasonably objective measurements we’ve actually lost ten times that number of jobs over the last several months.

Now Steve’s quoted a different econometric model that shows a different outcome.

Since we don’t really know the details of either model we might prefer one model over the other for psychological, political, ideological, etc. reasons. Or we could prefer objective measures (like the ADP figures) which tell us that we’ve lost a lot of jobs since the stimulus was enacted.
Bithead

May 18, 2009 at 1:58 pm

odograph, I’m puzzled by this comment. Last week Vice President Biden announced that the stimulus had created/saved 150,000 jobs. That pronouncement was based on the administration’s econometric model. By reasonably objective measurements we’ve actually lost ten times that number of jobs over the last several months.

Correct. Then too, is the idea of how many jobs would have been created absent the stimulus.

As discussed in another thread, I’m reading matters this way: The stimulus as offered, is useless in terms of the stated goal of actually keeping the economy afloat, much less growing it. It is by that measure, irrelevant, at least in the short term. In the long term, there’s the mass of added debt to consider… it cannot be without consequence.
legion

May 18, 2009 at 2:46 pm

It seems to support a concern I’ve had for a long time now… both businesses and individual consumers are so scared s*itless about their own individual prospects in this economy that, when they do get some small additional economic boost, they simply aren’t spending the way most economic models predict. They (both consumers and businesses) are hoarding whatever they can get their hands on to survive the current drought. When people start feeling a little more secure that things – at the very least – aren’t going to keep getting _worse_, they’ll slowly begin to use what funds they have to actually stimulate the economy. but we’re not there yet…
sam

May 18, 2009 at 2:52 pm

Now, all snarkiness aside as Greg Mankiw notes, with a shifting baseline projection there is absolutely no way to judge President Obama’s policies as a success or failure.

Just a technical question. Wouldn’t the rate of decrease or increase in the unemployment figure count as evidence for or against the policy? I would think that rate would be a measure independent of any baseline chosen.
steve

May 18, 2009 at 2:56 pm

May I refer you to Calculated Risk? Great economics blog. Unemployment is a lagging indicator. FTR, I fail to see the reason for the stock market rally, other than sideline money wanting in. This is a global financial crisis. I expect things to get worse yet. Also, how much of the stimulus has actually gone out?

Steve
odograph

May 18, 2009 at 2:57 pm

The common ground I have with Steve is that Presidents are largely/usually disconnected from economic growth and job creation, and that their claims are politics/BS as usual.

We can get into some really weird selection effects there though, because one party might be selected in downturns more than the other, leading to the sampling error that they are there during the natural recovery.

That all said, a pencil drawing of “jobs without stimulus” is exactly the kind of BS that led to economics failing us.
odograph

May 18, 2009 at 3:01 pm

Just a technical question. Wouldn’t the rate of decrease or increase in the unemployment figure count as evidence for or against the policy? I would think that rate would be a measure independent of any baseline chosen.

See Sam, for this to be a science the experiment would actually have to be run with ‘all things being equal.’

The fatal flaw in all economic modeling is that those experiments can never be run, and worse yet, in their absence too many economics simply assume ‘all things being equal.’

It is frankly crazy, and goes beyond crazy to insane in its uniform acceptance.
Rick DeMent

May 18, 2009 at 3:07 pm

I think we know which explanation President Obama and his Administration are going to rely on.

Yeah just like the Republicans said that the 90’s boom would have been sooo much better if Clinton hadn’t raised taxes. or how much worse the mild recession would have been without Bush II tax cuts.

Just pick whatever numbers you want to look at and go with those because there’s no way to tell.

It’s like people saying that Reagen’s tax cuts fueled the 80’s boom. Sure if you discount the oil collapse, the huge influx of woman in the workforce, and the hit that Volker decided to take by allowing interest rates to rise. but hey it’s a great story. It’s bull hockey but …
sam

May 18, 2009 at 3:12 pm

See Sam, for this to be a science the experiment would actually have to be run with ‘all things being equal.’

Hmmm. Glad I didn’t throw in the ceteris paribus clause as I first intended….
Drew

May 18, 2009 at 3:20 pm

Steve –

I see. I was referring to the fact that most pundits I have read had projected far higher unemployment rates – Depression style unemployment rates, Obama “catastrophe” style unemployment rates – than are actually occuring, sans a stimulus package that has yet had time to have a material effect.
Steve Verdon

May 18, 2009 at 3:28 pm

May I refer you to Calculated Risk? Great economics blog. Unemployment is a lagging indicator.

This is a joke right? So….the graph from the Obama administration is f*cked because Larry and Tim forgot to adjust for the lags in their econometric models? Really? If this is Barry’s best argument, stop reading him NOW.

Also, how much of the stimulus has actually gone out?

Ahhh, much better. According to reports I…well…reported on here, damn little.

That all said, a pencil drawing of “jobs without stimulus” is exactly the kind of BS that led to economics failing us.

1. Economics failed? I don’t think there were too many economists making the decisions that lead to this problem.

2. You haven’t read that book have you? I haven’t, but based on the descriptions on Amazon it highlights how people (i.e. not just economists) fail to understand probability and evaluate risk accurately. It is indeed probably at the very root of our current problems, but it isn’t like economists are ignoring this (the Ellsburg Paradox, the Allais Paradox, the work of guys the Kahneman, etc.)

The fatal flaw in all economic modeling is that those experiments can never be run, and worse yet, in their absence too many economics simply assume ‘all things being equal.’

Actually…oh nevermind.

Yeah just like the Republicans said that the 90’s boom would have been sooo much better if Clinton hadn’t raised taxes. or how much worse the mild recession would have been without Bush II tax cuts.

Let me guess Rick you got to this point in the post and —–> to the comments to chastise me for being so unfairly partisan in not writing,

Now this isn’t something inherent in Obama Administration, but is a problem with administrations of all Presidents. No President will set up objective standards by which we can judge their policies as successes or failures. And when out campaigning every policy put in place by a given President is always a success, never a failure.

Oh, wait….
sam

May 18, 2009 at 3:34 pm

[The book] highlights how people (i.e. not just economists) fail to understand probability and evaluate risk accurately. It is indeed probably at the very root of our current problems….

Indeed, see Recipe for Disaster: The Formula That Killed Wall Street

Of course, the really interesting question is how a group of very intelligent people could allow themselves to become bewitched by an equation.
odograph

May 18, 2009 at 3:35 pm

Steve, for a while I amused myself by collecting articles about the failure of economics.

I stopped not because they stopped coming, but because they seemed done to death.
odograph

May 18, 2009 at 3:37 pm

Look at that! The last article in the “backlash” series begins (Arnold Kling speaking):

The problem is that macro time series are nonstationary and have low statistical power. Go back to the table on labor force composition I gave earlier in this post. There is no way to estimate a model in a way that is robust to such dramatic structural changes.

Low statistical power … but still good enough for a Verdon post.
anjin-san

May 18, 2009 at 4:19 pm

I’m reading matters this way: The stimulus as offered, is useless in terms of the stated goal of actually keeping the economy afloat, much less growing it.

Well of course you are Michelle. Ummm, Errr I mean Bithead…
Steve Verdon

May 18, 2009 at 4:24 pm

sam,

Of course, the really interesting question is how a group of very intelligent people could allow themselves to become bewitched by an equation.

Well not really. I think it was the Allais Paradox that showed even top experts in their field will make “irrational” choices.

If you go to the link and look under experiment 1 you’ll see that most people like gamble 1A over 1B, and that is usually because 1B has a 1% chance of winning nothing. Then under the second experiment they prefer 2B over 2A. In the first case that 1% of nothing moves people away from the better gamble, but a similar risk but with a bigger payout does not, hence a contradiction. Even top experts in this subject matter can be fooled at times.

Odograph,

It may come as a shock, but economics isn’t about holding your hand in everything you do. The problems with making choices under uncertianty are well known and have been explained. That you think otherwise is your problem.

And your beef is with macro-economics, not econoimcs in general. Take for example your article point to Will Wilkinson’s complaint about expectation reformulation…the rational way is via Bayes theorem. Problem is that that way is not how many people think. People routinely think P(A|B) = P(B|A). So there is a theory, but it doesn’t work in general. People use short cuts and/or fail to understand how probabilities work. There are alternatives, but this is microeconomics stuff and it doesn’t get into the macro literature. For example, there is prospect theory. Again the problem is how does it work in aggregate for an entire economy with actors utilizing prospect theory to make choices under uncertainty.

Now if you want to say there is a failure of macroeconoimcs, thats probably right. It is a field that has lots of work still to be done. However, areas like microeconomics are pretty well established. Demand curves slope downwards with respect to price, upwards with respect to income (for normal goods) and so forth.

Low statistical power … but still good enough for a Verdon post.

See, you are talking out of your ass. It is macro theory you have your beef with.
odograph

May 18, 2009 at 4:53 pm

And your beef is with macro-economics, not econoimcs in general.

Spoken by a nimrod who just gave me a “unemployment without stimulus” chart.
G.A.Phillips

May 18, 2009 at 5:26 pm

Geoff at Innocent Bystanders notes that we are on path for unemployment as if there was no stimulus plan. Problem is there is a stimulus plan. So either the initial estimates of the effect of the recession, absent a stimulus plan, were overly optimistic, or the stimulus plan is providing no benefit at all.

Um, were in a recession since the donkey’s took Congress, then Zero the 666 gets in office and makes plans to spend 13,000,000,000,000 dollars on government jobs, buying banks, and giving car companies to the unions that drug them under.

You do the math.
odograph

May 18, 2009 at 5:27 pm

Demand curves slope downwards with respect to price, upwards with respect to income (for normal goods) and so forth.

Only if you assume “all things being equal.”

You can do some nice demand curves for gasoline, and then whoops “Arab Oil Embargo.” Nimrod economists say “that’s different, we get a do-over.” Taleb says no, either you have a prediction that works or you don’t. A predictive model with “do-overs” whenever you call them is infantile self-abuse.
odograph

May 18, 2009 at 5:30 pm

Um, were in a recession since the donkey’s took Congress, then Zero the 666 gets in office and makes plans to spend 13,000,000,000,000 dollars on government jobs, buying banks, and giving car companies to the unions that drug them under.

If you believe NBER, this recession started in December 2007, which does not actually coincide with the November 2006 election.
Eric Florack

May 18, 2009 at 5:59 pm

I’m reading matters this way: The stimulus as offered, is useless in terms of the stated goal of actually keeping the economy afloat, much less growing it.

Well of course you are Michelle. Ummm, Errr I mean Bithead…

I await proof from you that the situation is anything but that.
Steve Verdon

May 18, 2009 at 6:10 pm

Odograph,

Don’t post while drunk.

If you believe NBER, this recession started in December 2007, which does not actually coincide with the November 2006 election.

Thus spoke the nimrod. Jesus Odograph, you really are pathetic.
Drew

May 18, 2009 at 6:19 pm

“You can do some nice demand curves for gasoline, and then whoops “Arab Oil Embargo.” Nimrod economists say “that’s different, we get a do-over.” Taleb says no, either you have a prediction that works or you don’t. A predictive model with “do-overs” whenever you call them is infantile self-abuse.”

I think my speculation that a nice load of ‘lumbo had arrived is looking better and better…..
G.A.Phillips

May 18, 2009 at 6:39 pm

If you believe NBER

funny thing about me is that I trust my own sensory receptors and I receptor a lot, lol.
odograph

May 18, 2009 at 7:29 pm

Steve, I can be fairly moderate, but when you are wrong and abusive at the same time, it brings something out in me.

You wrote:

Now if you want to say there is a failure of macroeconoimcs, thats probably right.

This is actually AS you defend your chart of unemployment without stimulus.

From wikipedia:

Macroeconomists study aggregated indicators such as GDP, unemployment rates, and price indices to understand how the whole economy functions.

You poor, dumb, SOB. Why don’t you insult me again to make yourself even more of an ass.
odograph

May 18, 2009 at 7:31 pm

Drew, you can hear Taleb talking about economists, oil, predictions, and do-overs, in this mp3.
anjin-san

May 18, 2009 at 10:17 pm

I await proof from you that the situation is anything but that.

Don’t see you offering any proof or your position. Put up or shut up.
steve

May 19, 2009 at 5:06 am

You probably will not look back at this, but could you please attribute your charts? You list no source for this and there was none given at the link. One of the commenters at the link said you could find it if you went through older posts, but I think that if you wish to write and be read, the writer should do that. Even on my no name blog I list the sources for all charts and graphs.

Steve
skinbad

May 19, 2009 at 3:30 pm

The original chart is here.