Tuesday, April 13, 2004
The Problem With The Middle Class Misery Index
The John Kerry campaign released a cool new economics statistic: the Middle Class Misery Index.
Essentially, the index is a composite of seven different indicators: median family income, college tuition, health costs, gasoline cost, bankruptcies, the homeownership rate, and private-sector job growth.
Check out the link for the graph.
Now, those of us who remember a time before ubiquitous Internet access also remember another metric called "The Misery Index." The Misery Index was simply the sum of the unemployment rate and and the inflation rate.
I think the "Middle Class Misery Index" is an interesting and potentially useful shorthand for describing economic cycles and their effects on the middle class. But it's not useful this year.
Here's why: 1.) There's nothing in the index that's predictive of future trends. Every indicator in the Middle Class Misery Index is a trailing indicator--describing bad times which are behind us.
2.) Gasoline is a commodity. Indeed, it's a very volatile commodity, and for that reason, economists often exclude the price of fuel when computing inflation figures. Gasoline prices are only useful when adjusted for inflation or measured against some other contemporary benchmark, such as wages. Adjusted for inflation, gas prices are still far below their $2.94/gallon peak, in 1980.
3.) Note the absence of broad inflation figures and the unemployment rate itself in the Middle Class Misery index. If they were important guages of middle class misery to begin with (a questionable assumption in the case of inflation figures, to the extent the middle class A. a heavily leveraged and B. maintains the largest share of its net worth in its home--which rises with inflation, unlike bonds, which get killed). Here's where we ought not to let political campaigns dominate our economic thinking.
The reason the original Misery Index figures are excluded is because the unemployment figures for the last recession have been comparatively moderate (compared to, say, the 1981-82 recession, in which unemployment reached nearly 11%), and inflation has been extremely low by historical standards. What's more--and here's a forward looking indicator here: inflation expectations are near multi-decade lows, using long-term interest rates as a proxy. (Nobody's going to deliberately lend money at rates below the expected inflation rate).
If the misery index were a pot full of pasta, then, it wouldn't stick to the wall. The Kerry campaign had to cook the data a little more, throw out the CPI, focus our attention on fuel prices only during the last three years (during a war in the middle east, no less. Hey, THAT doesn't bias anything, does it?) and cherry pick two narrow fields which have outpaced inflation: college costs and health care premiums, no matter how you slice them. (They outpaced inflation under Clinton, too, by the way. They don't have much to do with presidential policies. Their cost could be hidden from the taxpayer through subsidies and they'd STILL outpace inflation).
Cooking the data is called 'data mining.' Which is, in statistics, a dangerous practice. If you're cooking data to arrive at the truth, that's one thing. But usually people cook data to arrive at the conclusion that best serves their interests. And that's what's been done here.
I think there's a lot of potential use in the Middle Class Misery Index, if they fix its built-in biases. But right now, it tells me a lot more about the economic cycle than it does about presidential policy, and it tells me a lot more about the people who designed it than it tells me about the economy at large.
Splash, out
Jason
Hat tip: Mudville Gazette
Update: Gregg Easterbrook weighs in here.
Kerry's index can make you giggle because, in order to be manipulated such that George W. Bush has "the worst record of any president ever," indicators must be chosen that give a great economic rating to Jimmy Carter. Check the Kerry campaign's graph, halfway down the page. When were times best by this index? At the end of the Clinton administration, and in 1978. Can you find one single person in the United States who would want a time-machine ride to the economic conditions of 1978?
He brings up some other points, too. Good read.
Essentially, the index is a composite of seven different indicators: median family income, college tuition, health costs, gasoline cost, bankruptcies, the homeownership rate, and private-sector job growth.
Check out the link for the graph.
Now, those of us who remember a time before ubiquitous Internet access also remember another metric called "The Misery Index." The Misery Index was simply the sum of the unemployment rate and and the inflation rate.
I think the "Middle Class Misery Index" is an interesting and potentially useful shorthand for describing economic cycles and their effects on the middle class. But it's not useful this year.
Here's why: 1.) There's nothing in the index that's predictive of future trends. Every indicator in the Middle Class Misery Index is a trailing indicator--describing bad times which are behind us.
2.) Gasoline is a commodity. Indeed, it's a very volatile commodity, and for that reason, economists often exclude the price of fuel when computing inflation figures. Gasoline prices are only useful when adjusted for inflation or measured against some other contemporary benchmark, such as wages. Adjusted for inflation, gas prices are still far below their $2.94/gallon peak, in 1980.
3.) Note the absence of broad inflation figures and the unemployment rate itself in the Middle Class Misery index. If they were important guages of middle class misery to begin with (a questionable assumption in the case of inflation figures, to the extent the middle class A. a heavily leveraged and B. maintains the largest share of its net worth in its home--which rises with inflation, unlike bonds, which get killed). Here's where we ought not to let political campaigns dominate our economic thinking.
The reason the original Misery Index figures are excluded is because the unemployment figures for the last recession have been comparatively moderate (compared to, say, the 1981-82 recession, in which unemployment reached nearly 11%), and inflation has been extremely low by historical standards. What's more--and here's a forward looking indicator here: inflation expectations are near multi-decade lows, using long-term interest rates as a proxy. (Nobody's going to deliberately lend money at rates below the expected inflation rate).
If the misery index were a pot full of pasta, then, it wouldn't stick to the wall. The Kerry campaign had to cook the data a little more, throw out the CPI, focus our attention on fuel prices only during the last three years (during a war in the middle east, no less. Hey, THAT doesn't bias anything, does it?) and cherry pick two narrow fields which have outpaced inflation: college costs and health care premiums, no matter how you slice them. (They outpaced inflation under Clinton, too, by the way. They don't have much to do with presidential policies. Their cost could be hidden from the taxpayer through subsidies and they'd STILL outpace inflation).
Cooking the data is called 'data mining.' Which is, in statistics, a dangerous practice. If you're cooking data to arrive at the truth, that's one thing. But usually people cook data to arrive at the conclusion that best serves their interests. And that's what's been done here.
I think there's a lot of potential use in the Middle Class Misery Index, if they fix its built-in biases. But right now, it tells me a lot more about the economic cycle than it does about presidential policy, and it tells me a lot more about the people who designed it than it tells me about the economy at large.
Splash, out
Jason
Hat tip: Mudville Gazette
Update: Gregg Easterbrook weighs in here.
Kerry's index can make you giggle because, in order to be manipulated such that George W. Bush has "the worst record of any president ever," indicators must be chosen that give a great economic rating to Jimmy Carter. Check the Kerry campaign's graph, halfway down the page. When were times best by this index? At the end of the Clinton administration, and in 1978. Can you find one single person in the United States who would want a time-machine ride to the economic conditions of 1978?
He brings up some other points, too. Good read.
Comments:
Post a Comment
