Arnold Kling  

A Plea for Facts

In Defense of Rationalist Club... Why the Left Should Not Forgiv...

One of the reasons that I am worried about the economy is that I am seeing such a high ratio of sweeping proposals to hard facts.

[UPDATE: At this discussion of the role of the Community Reinvestment Act in the crisis, the people with the facts were the ones arguing that CRA was not a big factor. I think that Russ Roberts made a persuasive case against the non-transparency of using Freddie, Fannie, and regulated banks to allow the government to outsource social spending. But I think that the CRA supporters were effective in pointing out that the default rates were higher for loans originated by institutions not subject to CRA.

From the point of view of CRA advocates, we kind of went full circle. In the early 1970's, poor people took on too much debt from unscrupulous lenders. Then we got CRA in 1977, which forced regulated institutions to participate in poor neighborhoods. Then, starting around 2000, the unscrupulous lenders piled back in. I'm neither endorsing that story nor discounting it. It's worth considering, and it's worth getting some more facts.]

When I wrote my book on health care policy, my first goal was to assemble and organize facts. What are the largest components of health care spending? How much of the rise in health care spending is due to price increases, and how much is due to increased use of medical services? How does medical spending differ by age group and income class?

Even more than the division between left and right, what is visible to me in health care policy is the division between fact-based and fact-free analysis. People will assert that much of the excess costs in medicine come from malpractice suits, regardless of the facts. Or they will assert that drug company advertising and profits are a big issue, regardless of facts.

Every week, there are several proposals for dealing with troubled home mortgages. They are always presented in fact-free form. Instead of reading about someone's clever loan modification proposal, I would love to read a column that tells me what percentage of troubled mortgages are cash-out refis, what percent are owner-occupied purchase loans, and what percent are investor loans.

Instead of reading about new ideas for using the $700 billion slush fund (formerly the program to be buy troubled mortgage securities), I would like to know approximately what percentage of banks is troubled. Both unweighted and weighted by assets. I would like to know what percent of troubled banks would be fine if their portfolios were marked to their prices as of, say, October of 2007. (I am not saying I have a policy in mind based on those historical values, but I just want some facts.)

I would like to know how much of the market values lost on credit default swaps and other securities represents actual losses (i.e., real defaults) and how much of it represents an increase in the probability of losses. If I sold a credit default swap on a company when the estimated default probability was 1 in 10,000 and that probability is now 1 in 100, the market value of my liability is now (perhaps more than) one hundred times greater than it was. On the other hand, there is still a 99 out of a 100 chance that I will come out whole. Does that hypothetical example typify what is going on? Or not?

I find the fact-free policy recommendations pouring onto the op-ed pages to be pure noise. I find the news reports equally empty. Occasionally, the press provides a bit of signal amidst the noise, as in the WSJ piece that led me to write this post on AIG. But even that left me hungering for hard data.

Maybe I am the only one who feels this way, but I would love it if an official from the Fed or the Treasury, instead of announcing the next zig-zag in policy, would give a fact briefing. Tell us what they know. Tell us what they don't know. The current situation has characteristics {X} and we are going to try to move it to {Y}.

It could be that, deep down, they know all this information and have decided that it is either too boring or too sensitive to talk about. But I've probed a few places in Washington (connections are not my strong suit), and nobody has hinted to me that they know the answers to the sorts of questions that I'm asking. I would feel better if it were otherwise.

Comments and Sharing

COMMENTS (14 to date)
David writes:

I think this can be blamed on the implosion of the MSM, honestly. They used to be all about gathering, summarizing, and reporting facts, but they switched to opinion reporting and press release publication, and no one's filled in the gap.

pushmedia1 writes:

Are these raw data available publicly and you desire for the statistics you mention to be calculated? Or is your complaint that these data aren't available, raw or not?

Brian writes:


I can understand your consternation (or is it bewilderment?) with the whole situation. Unfortunately, there doesn't appear to be a source of data (at least publicly available or relatively inexpensive to obtain) one can point to regarding who exactly are the troubled homebuyers, i.e. investors, 2nd homes or primary homes.

However, I have had a discussion with a fairly well-known economist (he shows up quite a bit on the TV circuit, newspapers, etc.) regarding the profile of who is defaulting on mortgages. He has a lot of contacts within the banks and from what he has learned the majority of defaults/foreclosures for investors and 2nd home purchases were at the leading edge of this because these folks tended to use option ARMs and the like. Since mid to late-2007 owners of primary residences have been the bigger problem group.

I know this isn't data one can point to and say "HA", but figured it was better than the politicians' decrying predatory lending and the plight of the homeowner.

Methinks writes:

Amen, Arnold. If you ever get that data, please share it with us.

I think the typical response to such data is "don't dazzle me with science", so the media doesn't report it. The target market doesn't understand it. Since policy makers are judged on how fast they rush to "do something", we are left with shoot now, ask questions later government.

Brad Hutchings writes:

Policy seems like more of a Darwinian world with the underlying facts acting as the natural selection agent among competing ideological species that procreate in the political process. That suggests to me that the most important information we can generate right now is an understanding of why things failed rather than where things actually are.

In the current popular narrative, the unregulated free market caused this mess, and so any solution must involve "effective regulation". This has become the most prominent underlying fact shaping the debate over how to fix things.

El Presidente writes:

This is where I wonder if Bernanke has internal personal turmoil given his desire to introduce greater transparency to the Fed.

What would people do if they had more information? I wonder how much of the information would stick, how much would be misunderstood or misrepresented, and how much initiative would be sacrificed by policymakers.

It's certainly more democratic to compile and release all of the pertinent information. I thought you were leaning toward anti-democracy. This would seem to suggest you might be conflicted, kinda like Uncle Ben. :-)

Dr. T writes:

El Presidente, one can be anti-democracy while still favoring release of economic data by the government. The federal government appears to release nothing other than policy statements and sweeping generalizations of vaguely defined goals.

Most decision makers would be bothered by the lack of relevent information. Our current decision makers seemingly aren't bothered by the lack of data needed to create information. Very scary.

El Presidente writes:

Dr. T

"El Presidente, one can be anti-democracy while still favoring release of economic data by the government.

I suppose one could. I can't argue with that.

Maniel writes:


Here are three facts: The Social Security Act was signed into law in 1935 by President Franklin Roosevelt; beginning in 1937, payroll taxes at 2% of income were collected and then transferred to retirees; and in 1937, the ratio of workers to retirees was 45 to 1. From these three facts, it is possible to deduce that in 1937 Social Security was beneficial to retirees at relatively low expense to retirees.

What the facts do not address was whether Social Security, as enacted in 1937, was good for the country. I claim that it was not, not simply based on the new facts that show a stunning unfunded liability, a much lower worker to retiree ratio, and a tax rate 6 times the original. I claim that Social Security was not (and never could be) good for the country based on logic: it created a government retirement fund monopoly (first draw on worker money through the FICA tax), an income transfer from working young to non-working old, and the bureaucracy necessary to shuffle the money.

One cheer for facts!

P Samuel writes:

"How far will deleveraging go?" by David Roche has plenty of data on the capital and loans of banks before and after the financial freeze-up. I'm not sure how good the data is or whether Roche uses it properly... but it's data of the kind Mr Kling correctly asks for to improve the quality of these discussions.


phineas writes:

I, too, say "amen" to Arnold's exceedingly fundamental point. And it follows from Arnold's point that he himself should now ignore all the chatter that's out there, and focus on digging up facts for himself. When he says he's "hungering for hard data" and "would love it if an official from the Fed or the Treasury would give a fact briefing" what he's saying is "I would love it if somebody would spoon-feed me, please". A very understandable sentiment. But if he wishes this blog to be more than chatter, he should get digging.

Steve Sailer writes:

Now that the election is over, we can get a better understanding of how both parties are culpable in the diversity recession. Few wanted to put together a comprehensive account of what went wrong because both parties were to blame. The GOP played up Congress passing the CRA way back in 1977, which sounds pretty silly, but at least they could blame it on the Democrats. And the GOP attacked Fannie Mae because lots of Democrats get sinecures there, but Fannie was just part of the store.

Almost everybody has been silent on George W. Bush's 2002-2004 jihad against down payment on home purchases in the name of boosting minority homeownership by 5.5 million households. This led to a debauching of traditional credit standards not just for minorities but for everybody. What use was this in the just finished election. This was total political correctness, but it was perpetrated by the Republican President as part of Karl Rove's plan to make conservative homeowning Republicans out of Hispanics.

Still, the most insane number is that total dollars loaned to Hispanics for home purchases almost _octupled_ between 1999 and 2006 (up 691% according to the federal Home Mortgage Disclosure Act). (Dollars for home purchases going to whites was doubled.) Lending to blacks quintupled. Minorities accounted for half of subprime dollars in 2004-2007, and, I would guesstimate, something like 60-75% of the dollars defaulted.

That's why 50% of the defaults nationwide have been just in four heavily Hispanic states: California, Nevada, Arizona, and Florida.

Brad Hutchings writes:

That's a great update, but it's easy in that case to get caught up in the facts and overlook the question of how CRA, Freddie, Fannie, and overt political pressure worked with loose credit to exacerbate the mess or even just be the straw that broke the camel's back. One plausible way it could do that is to seed demand among usually unqualified borrowers. If your friend buys a house using one of the programs and you think you're in more or less the same financial situation, you might try to buy a house too, but might fall prey to a subprime lender.

The CRA defense strikes me as the opposite of how statistics are compiled for cell phone use and automotive accidents. In the latter case, there's too much information. If a cell phone was in use, the correlation is noted, and in aggregate, when researchers assume causation, talking on the cell phone looks more dangerous than drinking and smoking crack and wearing a blindfold at the same time. In the former case, they just look at the CRA portfolio, and ironically, assume it isn't having positive effect on demand in the communities! In both CRA and cell phone cases, there is probably some percentage that accounts for spillover into the broader mess. It's not 100%, but it isn't 0% either.

TSheCat writes:

Im not an economic genius or carry any degrees other than from the University of Hard Knocks. Just a simple middleclass business owner. But I think, although the Community Reinvestment Act at face value seems a compassionate act toward minorities, it's main purpose was to slow a housing glut and stimulate new housing starts. As more babyboomers retire to nursing homes or assisted living even more homes will be on the market for some time to come. The only answer is to lend to some with less than average credit or to start flattening houses and building something in their stead. Perhaps move people in low income government assisted housing into these homes and replacing current low income apartment complexs with schools or libraries?

Comments for this entry have been closed
Return to top