Mark Thoma reminds me of something Art Goldberger taught me: R-squared is over-rated

Mark posts a letter from Stephen Ziliak:

The chief finding of the Soyer-Hogarth experiment is that the expert econometricians themselves—our best number crunchers—make better predictions when only graphical information—such as a scatter plot and theoretical linear regression line—is provided to them. Give them t-statistics and fits of R-squared for the same data and regression model and their forecasting ability declines. Give them only t-statistics and fits of R-squared and predictions fall from bad to worse.

It’s a finding that hits you between the eyes, or should. R-squared, the primary indicator of model fit, and t-statistic, the primary indicator of coefficient fit, are in the leading journals of economics - such as the AER, QJE, JPE, and RES - evidently doing more harm than good.

This reminds me of Art Goldberger's teaching in Econ 612.  After I took that class, he turned his class notes into a book.  From page 177:

From our perspective, R2 has a very modest role in regression analysis, being a measure of the goodness of fit of a sample of LS (least squares) linear regression in a body of data.  Nothing in the CR (classical regression) model requires R2 to be high.  Hence a high R2 is not evidence in favor of the model, and a low R2 is not evidence against it...

...In fact, the most important thing about R2 is that is is not important in the CR model.  The CR model is concerend with parameters in a population, not with the goodness of fit within the sample. 

I also remember Gary Chamberlain was not crazy about t-statistics--he said he didn't want to see any "damn stars" in our papers.  We should care more about confidence intervals than hypothesis tests. 

The practical problem with San Bernardino County using eminent domain to acquire mortgages.

Joe Nocera writes about Housings Last Chance?

As for fair value, since the home has dropped dramatically in value, the mortgage is worth a lot less than its face value. On Wall Street, in fact, traders are buying securitized mortgage bonds at a steep discount — reflecting the true value of the mortgages they’re buying. Yet the homeowner remains saddled with a mortgage that is unrealistically high. The plan calls for the county to buy mortgages at a steep, but fair, discount to its face value, and then to offer the homeowner a new mortgage that reflects much, though not all, of that discount. (Fees and costs would be paid for by the spread.) The money to buy the mortgages would come from investors; indeed, Mortgage Resolution Partners is in the process of raising money...

But if the county effectively originates mortgages that are more valuable than the amount it pays to investors in eminent domain proceedings, it is hard to see how it is paying fair value.  I don't know how this works without the county losing money--the value of what the county buys has to be equal to the value of what it sells, but it will also be taking on a bunch of fees, legal and otherwise.

If the threat of eminent domain gets lenders to modify more loans, the threat could produce a better outcome.  But eminent domain itself...

Tiebout sure knew what he was writing about

I went to my 35th high school reunion last weekend, in La Crosse, Wisconsin.  It was very pleasant--people were very nice.  But a large number of my classmates were puzzled that I enjoyed living in Los Angeles (OK, Pasadena)--they figured that I lived there only because I had to for my job.  I had to assure them that beyond weather, LA had many attributes that I liked, including the fact that it is a large city.  (I also like the mountains, the ocean, the food, the music, and the wide variety of people).

On the other hand, I always figured that people stayed in La Crosse because of family ties.  While it is a lovely place (it is on a stunning spot on the Mississippi River), it is, well, small, and pretty homogeneous.  It was quite clear to me, however, that the people who stayed there generally enjoyed living there. 

And so we sort.  Having lots of variation in cities almost certainly allows people to be happier, as they (often) find or stay in the place that best suits them. 

Let me finish the post with a picture of a sign one will certainly not see in Los Angeles.

I am glad I do not live in a battleground state

Because national election results in California are pretty much a foregone conclusion, we don't get too much in the way of campaign advertising at this time of year.

I am in Wisconsin at the moment, however, and the airways here are already flooded. The quality of the "discourse" is disgusting. I am not surprised at being appalled at the xenophobic anti-Chinese Koch brothers stuff, but I was also appalled at the xenophobic anti-Chinese stuff being run by Tammy Baldwin. She should know better.

Marlon Boarnet on his Wal-mart study

He tells me:

WalMart wages were very difficult to get, but we found nothing to indicate that WalMart adjusted wages by cost of living. So if the article did not make it clear, the comparison is the wage gap if WalMart entered the Bay Area paying their nationally prevailing wage. We focused on benefits because a lot of the compensation gap was benefits, and assuming WalMart would enter Bay Area with their national prevailing benefits policy seemed even more reasonable.

and

On my re-read, we handled [whether comparing Wal-mart wages to SF wages was ok] on p. 441 and footnote 10. 

The part that is amazing to me now, and was evident then, is that half the gap was benefits. I wonder how much of this issue is attenuated with Obamacare?

More to come.

Yes, wages at Wal-mart are awful.

My colleague Marlon Boarnet and co-authors show that they are: in the Bay Area, Wal-mart grocery workers' total compensation is a little over half the compensation of unionized workers (Table 4).  Wal-mart also initially offers grocery costs that are between 8 and 20 percent lower than the stores with which they compete, and lead other stores to reduce their prices by 5 to 13 percent (Table 2).

I am prepared to accept the argument that higher wages are more important than lower grocery prices, although it is not an overwhelming argument to me.  Nevertheless, there are good reason reasons not to like Wal-mart (which is one of the reasons I don't shop there).

On the other hand, it makes no sense to me to bundle a bunch of weak arguments with a strong one, and yet that is what anti-Wal-mart activists often do.  And zoning should be zoning--if one entity has the right to use land for a particular use, others should as well, whether they are liked or not.

 

Some evidence of the decline of civilization

Many years ago, I watched the Today Show on a regular basis.  I hadn't watched it at all for years, however, until a day a few months ago when I was so sick that all I wanted to do was watch TV.  I turned on the Today Show, and it was absolutely awful.

This morning, I read this:

NBC News chief Steve Capus candidly told THR that he thought Curry had not been right for the job in many respects. He said he agreed with interviewer Marisa Guthrie that Curry had faltered in the cooking segments, movie star interviews and fluffy features that make up a large portion of "Today." 

"I think her real passion is built around reporting on international stories," he said. "It’s tough to convey a sincere interest in something if you don’t possess it ... and you could tell with her, you can tell with any anchor, whether they’re into it or not. And I think we’ve now come up with a role that will play to her strengths.”

What I essentially learned, then, is that Ann Curry got fired for being too serious and too smart. I guess the good news is that NPR gets good ratings. (OTOH, the top two on the list are further evidence of civilization's decline).