Wednesday, January 28, 2009

Oil Prices

A former student writes in:

Oil prices were high in the 1970s, and we had a recession. Oil prices are now a third of what they were this summer... and the economy seems to be suffering. I remember learning in my undergrad that energy prices play a significant part in the economy overall. But, I thought the cheaper the energy...the better off the economy. So, what's going on now? Would we somewhat better off if gas prices increased?

Changes in oil prices can either cause, or be caused by, broader changes in the economy. That is, the causality can go both ways.

In the 1970s, oil prices rose in the US because of the Arab Oil Embargo. This caused a recession in the United States. The reason for this is that oil was an input into many firms' production processes. When input prices rise, firms produce less. This makes GDP (gross domestic product --- the sum value of all goods and services produced) go down. It also makes employment go down. Think of a trucking firm that hauls fewer loads because diesel is more expensive; this means fewer truck drivers will be employed.

In today's US economy, oil prices are much less important. We have fewer manufacturing jobs and more service jobs. Manufacturing jobs tend to be ones where "energy" is combined with "labor" to make "output". These are the jobs that go away when oil prices rise. In service industries (like mine), the price of oil has only a very minor impact on costs. So these jobs don't go away when oil prices go up.

This is why the changes in oil prices we saw from Jan 2007 to Aug 2008 didn't cause a severe recession.

Our current troubles are driven largely by the problems in the financial sector. We were already in a recession by Sept 2008, but the collapse of Lehman Brothers and subsequent problems in the credit markets have led to a serious economic slowdown.

Consumers in the US are spending a lot less money as a result. And this means less demand for "stuff." Because there's less demand for stuff, firms that manufacture things are cutting back. They're cutting back on output, which means using less labor and less energy. Unemployment in China has risen, due to reduced demand for workers by Chinese firms. Because these firms don't want as much energy, the demand for oil has fallen, and this is what has caused the recent drop in the price of oil. Think about all the TVs that aren't getting bought because US consumers feel poor. Those TVs would have been shipped from China on boats, and boats run on oil. This is a drop in the demand for oil, and this is why prices have fallen.

So, in the 1970s high oil prices caused the economy to struggle. Now, a struggling economy has caused oil prices to fall.

Here's a link to a recent NPR story that hits on the interconnectedness of the global economy that draws out some of these themes.

Thursday, January 22, 2009

Winter Break Reading

I'll get back to the personnel economics of U budget cuts tomorrow maybe, but some winter reading for today.

Two of the books I read over the winter break were related, and they were:
Wallace Stegner's Beyond the Hundredth Meridian: John Wesley Powell and the Second Opening of the West
Mark Stein's How the States Got Their Shapes

I certainly wasn't expecting to see a connection, but one of Powell's proposals (in the late 1800s) involved the economics of political boundaries, which is the topic of the second book.

Powell recognized, long before most anyone else, that water was the crucial resource in the American West. This is unlike the Eastern US, where there's plenty of water falling from the sky ("rain", I guess they call it) onto any piece of land. Here, we don't get too much of this "rain" stuff. For most agricultural purposes, land is useless without water. Powell argued that the homestead laws used to settle places like Iowa just weren't going to work in the West. It just wasn't going to work to give individuals 160 acres --- the standard grant under the Homestead Act --- with no thought about where water was going to come from.

One of Powell's water-centric proposals was made to the Montana territorial legislature. He advised them to organize counties according to drainage basins. If all residents on a particular creek or river are members of the same county, then water usage decisions can be delegated to that county and resolved locally. Essentially, Powell was arguing that political boundaries ought to be drawn to allow for localized decision-making regarding scarce resources. He was ignored, of course, and this means that upstream counties and downstream counties fight over water rights. One problem with this political organization is that both upstream and downstream counties act in a completely self-interested manner, and issues are either resolved by the state (which has comparatively limited access to information on local conditions) or through protracted negotiations. Witness the problems of the Colorado River Compact --- which is an agreement of seven Colorado-River-basin states --- for examples of the exact problems Powell was hoping Montana could avoid.

This year is the 100th anniversary of Stegner's birth, so it's a good year to read him. I'm not a big fan of fiction --- why read something made up when the real world is so interesting? --- but he wrote lots of great non-fiction stuff about the West (and about Mormonism in particular).

Mark Stein's book is about how Powell's advice --- design political units to minimize across-jurisdiction conflict --- was sometimes heeded and sometimes ignored as the Congress drew up state boundaries. Under "heeded", have a look at Washington and Idaho. Gold was discovered in what-was-then the Washington Territory north of Boise in 1860. Miners flocked to the area. Farmers in the Puget Sound area did not want miners in Boise making laws for them (and vice versa), so Congress split the territory in two.

Under "ignored", look at Wyoming. It's four degrees of latitude high, and seven degrees of longitude wide. Same with Colorado. Congress was pretty much ignoring conditions on the ground, and just making similar-sized states out of blank spaces on the map.

From Stein's book, I also learned that Utah is smaller than originally planned. Much of eastern Nevada was part of Utah --- until the mining really got going there. Congress didn't trust the LDS settlers of Utah, so they lopped off big chunks of land and put them in Nevada (which Congress figured wouldn't be LDS-controlled) instead.

West Wendover was nearly ours!

Wednesday, January 21, 2009

U Budget Cuts

As you probably know, the U's budget has been cut, and probably will be cut again. I'm on the U's "Salaries and Annuities" committee, and I attended a meeting yesterday with the committee and some high-up administrators.

Faculty salary cuts are being considered, depending on the size of the cut coming from the legislative session.

But what a great meeting!

The reason this was a great meeting is that I study labor markets, employment relationships, and how to structure decisions in organizations. And this meeting had all these issues.

(As an aside: Pretty much everything that has happened since September 17 has been bad for my portfolio and salary --- just like everyone else's. But for me it's been interesting professionally. In expressing my interest, I risk coming off like I'm enjoying this. I'm not --- and on top of that I know that a lot of people are hurting from job loss. I'd for sure undo it if I could.)

The U is facing a bunch of interesting decisions right now, and I'll probably blog about them over the next week or so. One decision is this: If the U needs to cut faculty salaries by X%, how should it do so? Here are a couple of options:
(1) Reduce every professor's salary by X%.
(2) Tell each college that its total salary amount has to fall by X%, but let the deans and department chairs determine how to achieve that goal.

Fundamentally, this question is one about delegation of decisions. It's about where we should put decision-making responsibility in this organization. Option (1) is central administration making the decision. Option (2) is central administration delegating the decision to the deans and department chairs, subject to some parameters. It's exactly the kind of issue we consider in Fin 6250.

So what are the economics of delegation?

As with everything in life, delegation involves tradeoffs. There are benefits and costs associated with delegation, and it's important to weigh them carefully.

On the benefit side, delegation often means you're better able to make use of local information. What does that mean in this context? As one example, deans and department chairs know best about the labor market conditions in their particular fields. Central administration doesn't know (at least, not as well as the deans and department chairs) which of the finance professors are highly sought after by other universities, and which are not. In cutting salaries, you're providing opportunities for rival employers to pick off your best faculty. Choosing option (2) allows deans and department chairs to be a bit strategic about any salary cuts. It allows them to make smaller cuts in cases where labor market competition is intense, and larger cuts where the competition isn't so intense. This is potentially beneficial, in that it might allow the U to do a better job of retaining faculty.

There are a couple of potential costs of delegation. The first is agency costs. Deans and department chairs might have different preferences from central administration. It's possible --- at least in principle --- that deans might not allocate scarce salary dollars in the way that central administration thinks is best for the university as a whole. The best example of this comes from thinking about across-department and across-school externalities. Suppose there's a math professor who spends a lot of time working with faculty and students from the school of engineering. The math department might not value this person so much, and might decide to focus salary resources elsewhere. If this person left, it might hurt the school of engineering, but the math department might not take that into account.

A second potential cost comes from coordination problems, but I'm hard pressed to come up with an example of this one in the context of university salary cuts. In Chapter 3 of the forthcoming fifth edition of my book Economics of Strategy, I write about coordination in military organizations. Delegating decisions to lower-level commanders in military operations is often not a good idea, because actions of different units need to be coordinated. Attacks might work best if all units attack at the same moment, and as a result delegating decisions might achieve the best outcome. One good example of coordination in for-profit firms comes from marketing campaigns --- it's sometimes best if a firm's various products are marketed using a consistent message. Delegating decisions to product-level marketing groups might fail to deliver that consistent message, and so it can be good to have centralized oversight of such choices.

Sunday, January 18, 2009

External Perceptions of Utah

I was in Portland over the weekend taking my kids to see the grandparents. And this was the top story in the (Portland) Sunday Oregonian.

It doesn't show in the online version, but the subheading of the article caught my eye: "The (Portland) metro area is less diverse than most -- even Salt Lake City."

Whiter than --- gasp! --- Salt Lake City!?!

I point this out here just to illustrate one example of external perceptions of Utah; and how those perceptions are not always on the mark. The Oregonian's headline writer is clearly shocked --- shocked! --- that Portland could possibly be whiter than Salt Lake. But as the article recounts, census data shows that SLC is not the least diverse city in America.

As I've noted here maybe once or twice, we've seen substantial Hispanic inmigration over the past decade. I imagine the data will show this trend slowing due to the recession. But Utah is looking more and more like the rest of the Western US. Perceptions will catch up. Someday.

Thursday, January 15, 2009

Private Clubs

The Utah Legislature is considering lifting the "private club" requirement for entering a bar. Or, at least, the governor may ask the Legislature to consider doing this.

(For those who aren't Utahns, the way bars work here is this: Bars in Utah are "private clubs" that are not open to non-members. You can enter a bar if and only if you are a member or you're there with a member. You can easily become a member for a small fee --- you just sign a card at the door.)

Some of the talk you hear in support of private club requirement is bad economics, and I thought I'd point that out.

Alcohol consumption certainly has negative externalities associated with it. Go read Paying the Tab by economist Philip J. Cook for more on this. One solution to the problem of externalities is to tax the activity that leads to the externality. Higher alcohol taxes would probably mean lower consumption, less drunk driving, and lower incidence of alcohol-related illness.

Without doing a lot of statistical analysis, it's hard to say exactly what the "right" alcohol tax should be, but I think it's safe to say that economists would agree that alcohol should be subject to taxes that are higher than other goods.

The private-club requirement isn't a tax on alcohol, it's a tax on entering a bar. So maybe that's the same as a tax on alcohol?

Some in the legislature seem to think so --- Senate President-elect Michael Waddoups said in the SLTrib "Unless we find something better that protects our children and protects us from drunken drivers, we want no change in private club memberships. Someday we may find a better solution, but it hasn't even been suggested at this point."

What the Legislature is doing here is taxing a complement to alcohol consumption, rather than taxing consumption itself. "Entering a bar" and "drinking" usually go together, so economists call them complements.

The problem with this approach is that it's the "drinking" part, not the "entering a bar" part, that causes harm. So a more effective way to reduce drunk driving would be to tax alcohol consumption directly.

This approach of "tax a complement to the harmful activity" is a pretty common thing for governments to do when the tax itself is unpopular. This is what the Congress did when it implemented the CAFE (Corporate Average Fuel Economy) standards on the auto industry in the 1970s. Taxing gasoline would be unpopular, so instead the Congress imposed a hidden tax on a complement. By forcing auto firms to raise average fuel economy of the vehicles they sell, Congress made it more expensive to own a gas guzzler --- this is a tax on a complement to fuel consumption.

There are (at least) two problems with the CAFE standards: (1) Driving a Hummer 1 mile to work uses less gasoline than driving a Honda Civic 50 miles to work. That is, the harmful activity is "burning gasoline" not "owning a Hummer". The CAFE standards didn't do anything to deter driving your Civic a lot, but that's an activity that causes a lot of harm. (2) The CAFE rules had an exception for light trucks, and so this meant SUVs were subject to different rules.

My suggestion for the Legislature on reforming alcohol laws? Raise the tax on alcohol itself, and get rid of the other rules.

Information on Utah's comparative alcohol taxes is available from The Tax Foundation. Turns out that Washington, Oregon, Alabama, Virginia, Alaska, Michigan, Iowa, North Carolina all have higher "Spirit Taxes" than Utah. Utah's "Beer Tax" trails Alaska, Alabama, Hawaii, South Carolina, North Carolina, Florida, Georgia, and Mississippi.

Wednesday, January 14, 2009

Education Reform

A Fin 6250 Student sends this link about Washington DC education reform:

D.C. Schools Chief's Plan Faces Opposition

DC School Reform will be a really interesting laboratory over the next few years, in part because the federal government is so heavily involved in administering the District.

One interesting feature of the DC plan is to get rid of tenure for public school teachers. This raises the question: Why does tenure exist in the first place?

The main reason for tenure cited by teachers' unions is to protect teachers from arbitrary firings. This raises an important issue, namely that it can be hard to provide incentives for supervisors to do subjective evaluations of performance. Getting rid of tenure would require principals to make hard decisions about who stays and who goes, and DC will have to grapple with this issue.

One reason for tenure in at the university level comes from the work of Lorne Carmichael, an economist who was thinking about hiring decisions. University professors have very specialized expertise --- my dean just isn't trained as an economist, and so it's very hard for him to figure out who we should hire if we're going to hire more business economists. In fact, I'm really the only person at the whole university with the expertise and connections to know which business economists are really good and which are just OK.

This means the "hiring" problem is really an "information extraction" problem. The university has to provide incentives for me to tell them the truth about who we should hire.

What role does tenure play?

Well, suppose I didn't have tenure. Then I might have an incentive to hire an economist who's good, but not quite as good as me.

Why would I do this?

Sometimes universities face budget cuts (like this year). And without tenure, the university might think about laying off some business economists. And if I hire someone better than me, I might be the one laid off. But if I hire someone worse than me, then maybe I'd be the one to keep my job.

To summarize, my incentives to build an excellent department would be weaker if I feared that doing so might cause me to be the first to go in a downturn.

We're trying to hire another business economist right now, and frankly the guy we're talking to is a lot better than me! Good thing I have tenure....

Tuesday, January 13, 2009

Tax Cuts

Some details of the Obama stimulus package have been announced, and I was surprised to see $300 bn in tax cuts. It's looking like the plan is to send stimulus checks to individuals, sort of like what happened last year.

The problem with tax cuts like this is that people don't tend to spend that money, and as a result there is little impact on overall demand for goods and services. And this means that these tax cuts probably won't do much to stop the bad employment news that we've been seeing.

This point was illustrated for me today --- I was giving another talk on the economy downtown (similar to my last one, so no new slides to post here), and I made the point about people not spending their stimulus checks.

One audience member works for a local firm that is a collection agency; the kind that calls you when you get behind on your mortgage or your credit card payments. She said that when last year's stimulus checks went out, her firm saw a big increase in payments, and asked how this fit in with my claim that people didn't spend their stimulus checks.

This fits my point exactly. Rather than using the stimulus checks to buy NEW goods and services (which would then increase demand and help keep people employed), lots of individuals used their checks to pay off the debt they owed on goods and services they had already bought. So.... not the sort of thing that's going to lift us out of our 500,000 per month job loss cycle.

Wednesday, January 7, 2009

Why (Some) Behavioral Economics Bothers Me

Before I launch into my tirade, I want to emphasize that Behavioral Economics is useful and economics has learned a lot from it. Understanding how people actually make decisions and what drives their preferences is important. Economics should be studying preference and choice, and we are.

But....

Some of it is just bluster. I've already criticized behavioral economist Dan Ariely's NYT piece on incentives in this space. He's just choosing to ignore reams of research on how financial incentives work in real firms. I don't know why he is doing this.

Other times, regular old economics gets labeled as behavioral, because it's easier to attract attention that way.

I've seen this a number of times in seminars and journals and the like, but here's an example from the New York Times recently. (And I should note that this article was submitted by a Fin 6250 group for their article discussion assignment --- way to go, group!)

Budgets Behaving Badly

This article is about how behavioral economics can help make better policy. I don't disagree with this claim; everyone should read Nudge for more on this.

But here's the thing: Read the section on how "Medicare separates hospital insurance and drug insurance into different programs." Any time an organization separates responsibility like this, you run the risk of across-department externalities. Here the drug insurance side of things is limiting access to drugs, thinking they're saving money. This, of course, makes people sicker, which means that the costs for the hospital insurance side of things goes up. It would probably be better if the people making the drugs decisions were motivated to care about the hospital side of things, but as of now they're not. Intra-organizational externalities like this are common (see Fin 6250). And they're understandable with regular old, garden-variety, neoclassical economics --- nothing obviously "behavioral" about it.

(Sorry for all the caveats in this post --- but here's another. I doubt it's Dana Goldman at RAND who's marketing this idea as behavioral. I'm guessing here, but probably this is the reporter's doing. I haven't been able to find Goldman's work on this, though.)

I'm of course showing my biases, but I think one reason that behavioral economics is popular is the notion that it's showing that regular old economics stuff is all wrong. "Oh, that economics you took and hated in college? The reason you hated it was because that stuff was wrong anyway because of strict rationality assumptions."

The big problem with college economics isn't that it's wrong; the big problem is that it's taught badly. Teaching regular old economics well is teaching exactly "the study of everyday life as it actually happens," just at Leonhardt wishes for.

On top of that, the big problem with our application of economics to policy isn't that economics is wrong (even though it surely is, at least on occasion); the big problem is that economists (and other social scientists) get ignored too often.

Tuesday, January 6, 2009

CEO Pay

Hope everyone had a great holiday! I just got back from the American Economics Association meetings. You have not lived until you've gone to some econ conferences, that's for sure.

Two notes about CEO Pay for today:

(1) A former student mails this link from Sunday's NYT:

Should Congress Put a Cap on Executive Pay?

I'd be surprised if Congress did any sort of hard cap on pay --- it's just too obviously a bad idea, for the reasons Frank identifies.

(2) My paper (co-authored with Rachel Hayes) on CEO Pay and the Lake Wobegon Effect is now forthcoming the Journal of Financial Economics. Here's a non-technical press-release type summary that the marketing folks here at the DESB put together:

In humorist Garrison Keillor's fictional hometown of Lake Wobegon, all children are above average. Now, University of Utah finance professor Scott Schaefer says that corporations might want investors to believe the same thing about highly paid CEOs. "CEO Pay and the Lake Wobegon Effect," a study by Schaefer and Utah accounting professor Rachel Hayes, argues that this effect may help explain high CEO pay. The paper is forthcoming in The Journal of Financial Economics.

The Lake Wobegon Effect in CEO Pay is the idea that a firm might want to convince investors that the firm's CEO is above average. An above-average CEO might, after all, be able to deliver above-average performance, which is what investors are after.

But how can a firm convince investors of a CEO's quality? "Everyone knows that in well functioning labor markets, better performers earn higher salaries," Schaefer points out. "The reason the Yankees are paying star pitcher C.C. Sabathia $161 million is that some other team was willing to pay him $160 million."

Schaefer applies this reasoning to CEOs like this: Think about a firm that hires a bargain-basement CEO. The CEO's labor-market options probably aren't that good, so suppose the firm can pay the CEO a low salary; say, $1 million a year. Investors, seeing that salary, might conclude the CEO isn't so great, and will downgrade the firm's stock.

The firm might wish it could hire a better CEO, but good help is hard to find. What if, instead, it simply decided to pay its bargain-basement CEO as if he were a superstar? If the firm bumps the salary to $3 million, investors might conclude the CEO is a superstar ---- and the firm's share price might jump. If the increase in the stock price is bigger than the increase in the CEO's salary, then this could be a good move for the firm.

Schaefer notes that this idea --- which has been around in the business press since at least the late 1990s --- has not received much careful scrutiny by economists. "The big problem with the Lake Wobegon idea as applied to CEOs," he says, "is that it seems to presume investors aren't very smart. You can imagine investors getting fooled by this trick once or twice, but over time they'd probably catch on as highly touted and well paid CEOs consistently fail to deliver." And if investors understand a firm's incentives to goose CEO pay just to pump up stock prices, then wouldn't firms give up trying?

The answer, Schaefer argues, is no. "What we show in our research is that the Lake Wobegon Effect can drive up pay even if investors are super smart about it. The key," he notes," is the role of investors' expectations."

His reasoning goes like this: Suppose investors expect the firm to play this game. Then, when investors observe a CEO being paid $3 million, they think "Oh, that's a bargain-basement CEO who should really only be paid $1 million. So we're not impressed." But should the firm then pay its CEO only $1 million? Perhaps not; if investors are expecting CEO pay to be inflated, then they'll figure any CEO who is paid "only" $1 million must be truly unqualified for the job. And the firm's share price will drop like a rock. The firm might therefore be better off overpaying its manager, even if investors are not fooled.

Although Schaefer believes he's demonstrated that the effect can occur, and identified conditions under which it can occur, one question remains: Is it actually occurring? Finding the answer, Schaefer says, might prove to be difficult with the data that's currently available.

"It's hard to calculate the exact financial impact a CEO has on a company," he says. "We would like to know what the firm's profits would've been if the CEO had not been there. It's hard to tell whether firms are profitable because of their CEO or in spite of their CEO."

With CEOs' pay packages in the headlines following the recent wave of bankruptcies, Schaefer says the practice of rewarding managers regardless of companies' success deserves to be re-examined.

"Pay packages for CEOs have gotten so high that the repercussions of getting fired are minimal because these guys are so wealthy," he says. "It can useful if CEOs are afraid of getting fired because this makes them cautious. Exorbitant pay packages can make it so they're not afraid of taking risk."

Despite this, he's not a proponent of limitations on pay levels. "If we limit CEO pay amounts, then it will be harder for firms to hire top talent. If private equity guys or investment bankers are earning tens of millions per year but CEOs are capped at a million, well, talent tends to follow the money." Instead, he thinks corporate America need to find better ways of rewarding managers for the long-term success of the firm.