Monday, October 11, 2010

Quick Thought on the 11th October

I mentioned in a prior post my suspicion that we are becoming lazier in research, probably on average. Anyone willing and able to fight this will probably be at an informational advantage in whatever field he or she chooses, unless that field disproportionately has people who have avoided that laziness (hard to believe given the prevalence of Google and the like) or that field has practically all information available online and easily findable.

I have a suspicion that in one of the fields I follow that the majority of the people do not reading the filings available for free online at a government commission's website; and furthermore, that those who do mostly focus on certain bits, not all of the reporting and therefore may miss something utterly relevant tucked into a footnote towards the end of a 200 page document. So while those people are at an informational advantage compared to the majority, they themselves are at an informational disadvantage compared to the people who read all of the reporting (though only if those people reading all of the reporting are able to interpret it).

Right now I am slogging through This Time is Different. Unfortunately, this means that I do not have the time to read Christina Romer's paper "The Macroeconomic Changes of Tax Changes" where purportly the argument is that tax cuts result in increases in GDP.

Funnily though, or maybe displaying a case of "whose bread I eat, his song I sing," in a blog post on the White House's site and a farewell speech given upon her resigning from Obama's economic team, she makes the case that extending the high-income tax cuts will provide very little in terms of short term job creation. (This blog entry was from July 28, 2010, so as you can see, I am rather late in my criticism. Then again, since I'm not a blogger for a conservative, libertarian or liberal institution, I have the liberty of writing about what I want when I want. And my small readership hasn't complained...yet.) Romer also notes that:

"Since most postwar tax changes have been broad-based, our evidence indicates that broad-based tax cuts have large effects.  But it’s important to note that our study did not distinguish among tax cuts for different groups and did not focus on high-income earners.  Thus, it provides no basis for doubting the compelling evidence that tax cuts for high-income earners are less effective than broad-based tax cuts focused on the middle class."

Now, here's where the laziness mentioned above comes in play, for me at least. If I read those remarks correctly, Romer says that the paper does not focus on high-income tax cuts. However, since she looks at broad-based tax cuts, presumably the effects include tax cuts for the rich since she did not focus on tax cuts for any one socioeconomic class but any tax cut package. Since I have not had the chance to read Romer's paper, I can only make that remark based on my reading of her blog post, not the primary source, her paper. That, my friends, is intellectual laziness on my part.

May I note this though? If her paper included effects of the tax cuts on the rich, not separating them on, then Romer has no basis to say that tax cuts to the rich will have little effect. Her paper focused on broad tax cuts, which included tax cuts for the rich. It did not do an analysis of which tax cut to which group had the most effect, so how does she have a basis to say that tax cuts for the rich will provide little benefit when her research did not disaggregate the effects? (She quotes the CBO and a Goldman Sachs study,  Goldman Sachs Global ECS US Research, “US Daily:  Extending the Expiring Tax Cuts:  What, How, When and Why (Phillips),” July 26, 2010.) Dare I say that this could be intellectual laziness on her part? Or a case of "whose bread I eat, his song I sing," since this administration seems to love to target the rich whenever possible? I'm not sure if it is a case of laziness, but unless that Goldman study is econometrically rigorous (again, laziness on my part for not reading it), I think it's more a case of "whose bread I eat, his song I sing."

Continuing on the theme of "whose bread I eat, his song I sing," let's run a thought experiment about the CBO. The CBO is a part of the government. In my skeptical mind, that means that it is biased towards anything that increases government revenues (and size), since those revenues form the basis of what can be spent and help fund the CBO's existence. (Sorry, not going to discuss the deficit.) So, while the CBO may not be biased towards either political party, it probably is biased towards government, since that is the system it is part of. And rationally, it would want to increase its revenues wherever possible. Hence, why they think high-income bracket tax cuts will be ineffective. Or maybe it's just that they want that additional revenue, eh?

Finally, let me level another criticism at Romer, again from her lovely blog post:

"Likewise, estimates by the Council of Economic Advisers suggest that spending $10 billion to prevent the layoffs of teachers, firefighters, and police would lead to nearly twice as many jobs as the estimated $30 billion of high-income tax cuts—that’s twice as many jobs for one-third the cost."

While I have nothing to say on the police, I must note that a paragraph like this seems to be terribly short-run biased. Actually, all economic discussions seem to be very short-term biased, not medium-term biased on the necessary structural reforms to government spending to prevent the day when my generation pays the bill for the excesses and irresponsibility of prior generations if we continue on this path. In this case, Romer's assumption is that it is good to preserve any job possible, the kind of logic that leads to a bailout of GM rather than asking the hard question of if those jobs are truly necessary, if it is good resource allocation to bailout an ailing motor company with structural woes (labor unions) rather than cyclical woes.

That same question applies to teachers and firefighters (and probably police). The unfortunate thing about teachers is that seniority overrules performance, the exact opposite of a rational employment system. And so when cuts need to be made, the teachers fired are likely the ones you want to keep, probably being young, ununionized, and not disenchanted with education yet; the ones retained are the ones you want to fire. So is it the right thing to spend money to save teacher jobs who would otherwise be fired? Keeping the teachers who would otherwise be fired, as long as they are the outperformers, is not bad; except that patches like that delay any necessary structural reform. Better would be to demand structural reform before handing out some money to teachers. But instead such $10 billion (part for the teachers) inefficiently allocates scarce government resources inefficiently by supporting a system that needs to be revamped if it is to actually serve the customers it purports to serve - the students. Shame on an economist for not thinking about that - or again, is it a case of "whose bread I eat, his song I sing?"

And while I have not seen anything statistical on firefighters, let me provide an anecdote from Merseyside (read it in The Economist), with a sentiment from the fire chief there that reflects part of my gloomy opinion about government in general - "the trouble with the public sector is bone-idle staff":

In 1999-2000 there were 2,140 fires in the Merseyside area and 15 fire-related deaths; last year (2009-10), there were 1,299 and 8. Meanwhile, the number of traditional fire officers has fallen from 1,400 to 850, saving money. [Better results with less staff and money - what's not to like?]

Mr McGuirk saw that speedy response wasn’t enough: prevention was the key.[...]

All this involved cutting the number of fire officers, who, Mr McGuirk realised, were underemployed for long periods during their shifts. Anyway, fewer fires required fewer rescuers. Although no one was made redundant involuntarily, in 2006 the fire-brigade union called a strike. Protesters dubbed the fire chief “McJerk”; 2,000 of them walked through Liverpool carrying banners with slogans such as “I hate McGuirk”.  [When you're part of a system, even if that system is part of a problem, the fact that you're in that system and that it puts the bread on your table means that you will fight for it no matter that it may need reform to fix the problem.]

Ironically, it was soon clear that the 200 officers who stayed at work could run the service at full capacity. [Emphasis mine]“I told the local press they would never notice there was a strike,” says Mr McGuirk. “It’s not my job to be popular, it’s to deliver.” The strike was defeated in a month.

Dare I say that the $10 billion comment from Romer represents intellectual laziness? A distressingly short-sighted viewpoint from an economist, who should look longer term than the next year, two or three? Someone failing to realize that preservation of a job is not the same thing as efficiently allocating resources where needed for the most return? Someone failing to question whether those jobs are truly needed, or whether those departments can run more efficiently with less or whether something structural, such as teachers unions, are the problem with dealing with budget shortfalls for education (if you could fire your worst performers during a cyclical downturn, instead of having to retain them and fire your best performers who are not unionised, wouldn't that make more budgetary sense than just doling out money to keep all the jobs and perpetuate a system in need of reform)?

My ending conclusion is that a constant focus on the short term means that you will always be in the short term, losing sight of the long view and avoiding problems that will come in the medium term until it is too late. Short term pain seems like it will be long term gain, in this case; or as Ben Franklin put it, "An ounce of prevention is worth a pound of cure." But a focus on the short term at the cost of the medium and long term means that we will continue to emphasis that pound of cure (and therefore waste resources) over that ounce of prevention.

N.B. Politics is something that is quite incendiary. Economics is something that can bore many people. While I have a small readership, no doubt I have said something that may have incensed or bored you. Please bear in mind that these are my opinions, and that I can be quite arrogant. If you disagree with me, either politely comment or impolitely send me an angry email - I would rather have any comments with inflamatory remarks, including any four letter words, to be in my mailbox. Besides, you'll have a higher chance of me reading it than if you put it in a comment. And if you're commenting, didn't you want me to read it in the first place? Hence why it is logical for you to send inflamatory remarks (and maybe all of them) to my inbox rather than posting a comment.

N.B. #2 Psychologically speaking, confirmation bias exists. And most people are susceptible to it. Broadly speaking, once I've come to a conclusion about it, I will tend to seek out data that confirms my view rather than data that disagrees with my view. This is probably one of the reasons for the joke "Science advances by the funeral." Replace science with "society" and you probably still are accurate in describing societal mores. (Another post for another day, about J.S. Mill's quote about conservatives.) Anyway, I know I am susceptible to confirmation bias (I also suspect that to be true of writers for think tanks and newspaper blogs of all political stripes). And hence, I promise myself and anyone reading this that I will not let this post lie like this. I've noted above that I've not had a chance to read Romer's paper, the Goldman Sachs or CBO studies. Since those potentially contain views and data disagreeing with my opinion, it is intellectually incumbent upon me to read those when I can and further decide if I am wrong or if the opposition's arguments have enough flaws to render their conclusions suspect. So once I'm done with This Time is Different, I shall read the literature mentioned above and revisit this post (though I'll probably preserve this post as is for my own humor in seeing the potential evolution of the argument).

Sunday, October 10, 2010

Thoughts so far on "This Time is Different"

There are two financial history/economics books which I have found very rewarding to read, even if a tiring slog to get through. One is Kindleberger's Manias, Panics and Crashes, particularly the discussion of credit creation and contraction with the Minsky model (the financial instability hypothesis).

The other, though I'm still working through it, is This Time is Different. Part of the reason it is a slog is the very reason is it so rewarding - the loads of quantitative data and segways discovered, such as India's share of world GDP declining between 1913 and 1990 (though why that table didn't have 2007 figures is unknown to me). Debt intolerance though seems to be a main part of the book, the notion that certain governments have higher debt intolerance than others, and hence while one government can withstand a certain amount of debt to GDP, another can not withstand that same amount. Mental-model wise, it is very interesting to discover this in the quantitative data; and even more interesting to connect it qualitatively to the narrative of history in Walter Russell Mead's God and Gold, which offers a pretty good explanation of the rise of Anglo-Saxon institutions that have resulted in those countries having very strong financial systems that are debt tolerant (it may not seem like we have a strong financial system, on an absolute basis, based on the past crisis, but relatively speaking we do compared to non Anglo-Saxon countries). One of Reinhart and Rogoff's tables show that none of the colonial (as opposed to imperialist era holdings like India, South Africa and Zimbabwe) Anglo-Saxon spinoffs (Canada, Australia, New Zealand, and the United States) have explicitly defaulted on debt - we're very debt tolerant, a blessing for people needing to buy fixed income securities. (Covert defaults, including inflation, is another story, though we still fare pretty well on a relative basis.) Mead's analysis provides a good explanation of that quantitative anomaly revealed on Reinhart and Rogoff's analysis.

From a research perspective, This Time is Different reveals another curiosity - due to the lack of transparency regarding domestic debt of countries, most economic analysis of default and inflation episodes do not incorporate a discussion of domestic debt. As a result, Reinhart and Rogoff find that most analysis is seriously deficient in explanatory power of why a government would choose to default or inflate because most researchers are looking at a very incomplete picture (only external debt, which usually is less than half of the total debt of the average nation) with incorrect assumptions (that domestic debt is usually paid off in full, at face value, with no overt or covert defaults). The simple reason for this deficiency is partially that governments do not want to reveal their knickers (unfortunate because government ultimately should be accountable to their citizens) and partially that researchers might be becoming lazier. (For more on this, see Farnam Street's blog article: Does the Internet Make You Smarter or Dumber?)

And if we take a moment to engage in a thought experiment, we may be becoming even lazier as we become used to instantaneous access to information via our computers, mobile search and more. If that is the case, then any information that is not easy to find via online sources at all or online only if digging deeply (i.e page 200 of a 201 page document) will progressively become more and more ignored. Any decisions we make then will be based on even more incomplete and erroneous analysis. This is important - for example, if I remember correctly (key because I do not have the book in front of me at the moment), Reinhart and Rogoff go so far as to say that institutions advising on debt restructuring for nations are giving inadequate recommendations because of the propensity of research to improperly ignore domestic debt levels and the lack of transparency around a nation's domestic debt.

Combining the two books, I distinctly remember Kindleberger talking about capital levels throughout the 19th and 20th century at banks in the United States. Reinhart and Rogoff have a time-series analysis, I believe, of banking crises in the U.S. I think it would be quite interesting to overlay that analysis with the trend of bank capital levels from the 1800s to the present. I think it would make the case that whatever reform we're getting through Dodd-Frank and Basel III, it will only mitigate (at best ) future crises, not prevent them.

I'm including systemic risk in that assumption, by the way - I highly doubt that systemic risk buildup will be prevented by the regulators created and empowered in the legislation, since I wonder if they (or anyone) will know what the next systemic risk is. They might be fighting the last war when the next crisis comes. I'm not asserting that the systemic risk of the prior crisis was hard to discern, since investors like Jeremy Grantham, Mike Burry and others happened upon that risk, analysed it, and profited from it; I'm asserting that I find it hard to believe that the regulators will notice the systemic risk, given the prior track record, such as this remark from Bernanke in Oct 2005: "House prices have risen by nearly 25 percent over the past two years. Although speculative activity has increased in some areas, at a national level these price increases largely reflect strong economic fundamentals." 

After all, if you recall, the Bank of Scotland (now part of Lloyds Banking Group) on average had an equity capital ratio of 25% during the 1700s (i.e. every $4 of loans made by the Bank of Scotland were fund by $3 of deposits or bonds sold to investors and $1 of the Bank of Scotland's money). This, shall we say, is a rather high common equity ratio, especially compared to banks today where you may see 3% common equity ratios (a lot of European banks back in 2008). And yet, when the Royal Bank of Scotland was competing against the Bank of Scotland in the early 1700s, it pretty much caused a liquidity crisis for the Bank of Scotland by buying Bank of Scotland notes and then marching into a Bank of Scotland branch to redeem those notes for payment (in gold or coinage, I think). The Bank of Scotland was forced to call loans early and suspend payments in coinage from March 1728 to September 1728. Bear in mind that there was nothing wrong with the Bank of Scotland's book of loans. They just had an inadequate supply of coinage to cover redemption of all of their currency notes in circulation (a situation the Royal Bank of Scotland probably was in too). A high equity capital ratio did not prevent a liquidity crisis, a crisis of confidence. Even a 100% equity capital ratio (meaning that the bank is not a bank, but an investment vehicle) would not prevent a liquidity crisis if the assets are not easily salable. And legislation would probably not prevent a liquidity crisis in the future. (I realize that such a liquidity crisis for a bank, as opposed to a capital markets dependent financial company, would not happen now because of the Federal Reserve, the Bank of England, and the European Central Bank each offering banks access to liquidity at need, but I think the example still relevant if only to illustrate that there are more to financial crises than just bad loans eating away at a bank's capital. Also, I felt like bragging about what I learned when touring the old Bank of Scotland headquarters in Edinburgh last summer.)

As for a quality crisis (while I do not remember the exact nomenclature, what I mean by this is that the quality of assets/loans on the bank's books are a lot more suspect than stated), a higher capital ratio means that there is more of a buffer provided by the bank's money to absorb losses on loan writedowns before impacting depositors. But if a bank is stupidly underwriting, no capital ratio will ultimately prevent it from going under. Legislation does not prevent stupidity, herding, or a race to the lowest common denominator (and lowest quality) in lending. And when banks did not have federal backstops, there were still banking crises despite the incentive of the bank owners to ensure that the bank underwrote well because if it made a bad decision, they lost their net worth. So why does anyone think that the current legislation will prevent a future crisis? Any rational person willing to take the time (i.e. not be lazy) and look at the historical record will find that to be a delusion.

I realize this post is a bit incoherent - I hope not only to come back and edit it, but to provide some data, such as that time-series analysis of banking crises with capital levels if only to underscore that Basel III capital levels will not prevent a future crisis. In the mean time, I hope you enjoy these off the cuff thoughts (and email me with any thoughts on the deficiencies).