Tuesday, May 26, 2015

Why Raise Rates?

The Federal Reserve keeps reminding us that it must eventually raise interest rates after six years of keeping short term rates near zero.

Why must it raise rates?  Low rates are usually thought to encourage economic growth, which is supposed to be a good thing.  There appear to be two reasons: "side effects" and "holding fire."

The side effects story is that keeping rates low is desirable, but doing so will eventually cause bad side effects, like inflation and asset price bubbles.

The holding fire story is that rate reductions are needed to fight recessions.  If rates are already at zero, then when the next recession comes the Fed will not be able to fight it by lowering rates.

Do these stories make sense?  The main side effect of low rates, inflation, is certainly not showing up in the consumer or producer price indices.  Asset prices are high, but it isn't obvious that they are at crazy, tulip-bulb levels.  Capitalization rates have come down to maintain historic spreads over treasury rates, but I can't see why they should continue to fall just because rates are low - if rates stay low, spreads will find their natural level, and asset prices will increase or decrease with expectations of future earnings, not the level of interest rates.  Why couldn't we maintain current rates and multiples forever?

One reason low rates might not be sustainable is if excess bank reserves are building up that will eventually cause inflation.  But excess reserves are at least in part due to the fact that the Federal Reserve is paying interest on reserves.  If banks reduce their reserves too quickly, the Fed could always raise that rate or raise reserve requirements to slow the process down and prevent inflation.

The holding fire story seems to suffer from similar logical fallacies.  If interest rates are an accelerator, then keeping them at zero provides maximum stimulus for the economy.  Why would it be a good thing to raise rates, discouraging growth, just so that we can lower them again in case of weakness?  Why not keep the accelerator down all of the time? 

Suppose we reasoned that hunger results from cutting back on eating, and is relieved by increasing our food consumption.   We might conclude that when we have been well-fed for a while, we should stop eating three meals per day, so that if we get hungry we will be able to add a meal.  So we might start skipping lunch, and sure enough, we would get hungry, and when we added lunch the hunger would go away.  The Food Fed might think it had done a great job of hunger-cycle management, but the entire exercise would have been unnecessary - we could have just stuck with three meals per day all along.

Of course, four meals per day might make us fat - a bad side effect.  But if we are maintaining a healthy weight on three meals, why not stick with the plan?  In other words, if inflation shows up we could raise rates, but why do so before there is even a hint of inflation?

Similarly, suppose a race car driver let up on the gas so that he could floor it at the end of the race.  That might make sense if he was tricking another driver into holding back so that he could pass him at the last minute, but if he is only racing the clock (as we are in managing an economy) then the best strategy is to floor the accelerator as much as possible - as long as there are no bad side effects, such as crashing the car, or inflating prices.

I admit that these are very simplistic arguments, but where is the economic theory showing that we need to raise rates in order to be able to lower them in bad times?  If Larry Summers is correct, the equilibrium real rate of interest is well below current levels, and raising rates will only make things worse.

Friday, May 8, 2015

La Guardia

Why do people hate La Guardia airport so much?  I was in New York City two weeks ago, and had to leave a conference and get on a plane in a hurry.  It was mid-afternoon and it took 20 minutes to get there from midtown.  That is amazing, and one of the great things about New York.

The airport doesn't have inspiring architecture or big glass domes enclosing wasted space like this:




La Guardia is mostly a plain Jane airport:

















But it gets the job done.  That isn't enough for Vice President Joe Biden, who called it a "third world" airport.  Reporting this comment, the New York Times ridiculed LaGuardia and praised other airports:

In Hong Kong, travelers can play a round of golf while waiting for a connecting flight. In Seoul, South Korea, people can strap on skates and kill time on the ice at the Sky Rink. And in Paris, parents are provided with free strollers as they cruise newly upgraded terminals


In today's New York Times, a civil engineer and transportation planner called for closing La Guardia, so that travelers can enjoy the amenities of JFK and Newark instead of enduring the "grubbiness" of the "dilapidated" Queens airport.

When I fly I want to get into and out of the airport.  I don't want to spend an extra hour getting to the airport so that I can walk through a longer duty-free shop zone or stop at a spa on my way to the gate.  I would put up with a lot of grubbiness to shave another 30 minutes off of my travel time and lower my taxes at the same time.

So called "infrastructure investment" isn't always worthwhile.

.

Sunday, May 3, 2015

Baltimore

Today the Wall Street Journal is reporting results from a poll asking whether riots in Baltimore and other cities were the result of "Longstanding frustration about police mistreatment of African-Americans" or "People seeking an excuse to engage in looting and violence." 

In order to maximize the apparent difference in response between whites and blacks, no mention is made of whether anyone answered "both" or "neither."  Looting and mayhem are nothing new, however, and any reflection at all on the history of these kinds of events makes it clear that both explanations are important.

Looting often occurs when there is an opportunity to do so.  I looked up news reports after several U.S. hurricanes and earthquakes and found that looting was a major problem after all of them.  Looters were found carrying human fingers after the Galveston hurricane of 1900, cut from the dead to retrieve expensive rings. There is often a racial element to these reports, as described in a recent Master's thesis, and in contemporaneous reports.  But looters are diverse - there were white looters during the 2008 Iowa floods, and even Japanese looters during the tsunami.  Looting was widespread in the U.S. during the Great Depression, and large numbers of whites took part in riots and looting during the 2011 English riots.

Rioting and looting is a problem like forest fires - good management can keep them to a minimum, but they are bound to happen from time to time, which is why police departments stock riot gear

Riots work something like this:
The pattern appears to proceed roughly through three stages: (1) A primarily symbolic looting stage, where destruction rather than plunder appears to be the intent. It often seems initiated by alienated adolescents or ideologically motivated agitators in an area. (2) A stage of conscious and deliberate looting, in which the taking of goods is organized and systematic. It frequently appears spurred by the involvement of omnipresent delinquent gangs and theft groups operating on pragmatic rather than ideological considerations. (3) A stage of widespread and nonsystematic seizing and taking of goods. At this point, plundering becomes the normative, the socially supported thing to do. 
 In a hurricane, earthquake or blackout, stage one is given, and stages two and three follow if there is opportunity.   But in Baltimore or Ferguson, the riots would not have happened without the spark of the deaths of local residents while interacting with the police.  The spark would not have gotten the fire going without the underbrush of widespread resentment.

In an excellent book about the 1977 blackout in New York City, Robert Curvin and Bruce Porter outline the theory of and historical data on looting.  Evidence suggests that absolute deprivation, unemployment, etc. are not factors causing riots and looting; the important factor is change in status.  High expectations that are not fulfilled cause discontent.  The civil rights movement raised expectations among blacks during the 1950s and early 1960s, and when their economic conditions did not rapidly improve the potential for unrest increased.

From the early 1980s until 2008, economic conditions for blacks in the U.S. improved, and the election of Barack Obama as president raised hopes even more than economic history suggested.  The crash of 2008, the slow recovery, and the moderation of the Obama administration dashed these expectations, creating the conditions for sparks to grow into fires.  The sudden explosion of social media allowed previously unremarkable events, such as deaths in police custody, to be those sparks. 

As I mentioned earlier, non-black Americans also have the potential to riot and loot, but their outbreaks of disorder are triggered by random events that create the opportunity to loot and destroy.  The U.S. black population is different in that it has continuing resentment against whites from slavery and discriminatory policies through the mid-20th century, and it has had significant swings in hope and disappointment, partly because lower levels of financial wealth and human capital cause greater exposure to business cycles.  Blacks also compete with immigrants, resulting in lower wages and higher unemployment for blacks.

Vigorous policing and mass incarceration achieved dramatic reductions in crime over much of the same period, resulting in significant social and economic gains.  There is a tradeoff between police brutality and crime, just as there is a tradeoff between cargo weight and fuel consumption in an airplane.  Technological improvements can increase the amount of weight carried for a given amount of fuel, but the tradeoff still exists.  We might invent better police methods that require less brutality, but it is unlikely that we will quickly improve upon methods that have been developed over the past few hundred years.  Perhaps body cameras will change policing, but it is far too early to tell.

Rapid urbanization, labor unrest, and prohibition all caused police activity, and police brutality, to increase during the late 19th and early 20th centuries.  A reaction against harsh policing began in the 1930s and culminated in supreme court decisions that limited police conduct.  Popular reaction to the resulting increase in crime, illustrated in movies like Dirty Harry and Death Wish, eventually resulted in a crackdown on crime that continues to this day. 

Perhaps the pendulum will swing again, and police brutality will decline, reducing the number of events that spark riots.  The price of this swing will be an increase in crime.  Perhaps the economy will improve for blacks, reducing the potential energy of riots.  Or perhaps riot control will attract more resources.  Using the forest fire analogy, we might see fewer sparks, less underbrush, or more fire suppression efforts.  Or perhaps we will just see more fires.

Saturday, April 25, 2015

The Mediterranean Exodus and the Trolley Problem

How should we think about the migrants who are trying to cross the Mediterranean from Africa to Europe?  Thousands of them are paying smugglers to take them in unsafe boats to countries where they have no legal right to live, work or visit.  Many are dying in the attempted crossings.

Until very recently, Europe had more people than Africa, and so immigration pressure was limited.  Over the next few decades, however, Europe's population will be steady, while Africa's will explode.  The divergence between perceived opportunities in Europe and Africa will also increase.

If the possibility of a European population that is majority African and Arab is of no concern, then of course Europe should end the humanitarian crisis, run a free ferry service across the Mediterranean, and allow anyone who wants to settle in Europe.  But if Europe wants to remain majority European, it must do something to stop people from crossing over. 

There are three options:  Working to improve conditions in Africa so that people there will be less interested in migrating; paying people to stay in Africa; and using force.  The first two options are unrealistic.  The more Europe spends to improve conditions in Africa, the more apparent it will be to Africans that things are even better in Europe.  It is not the desperately poor who emigrate, it is the ones who have begun to glimpse the advantages of middle-class life and want more.  Foreign aid programs are unlikely to succeed in improving conditions in Africa - as P.T. Bauer said, the money spent will end up like water thrown on sand.  But even if it succeeds, it will increase immigration pressure, not reduce it.  And given population projections, Europe will not have enough money to bribe everyone to stay home.

The final option is force.  Europe could pay African dictators to patrol the beaches and prevent boats from launching, as they did before the Arab Spring.  Europe can also intercept boats in the water, arrest the smugglers and return the passengers to Africa.  Fatalities at sea will be an inevitable consequence, at least until Africans are convinced that Europe is serious about stopping immigration.

Many other countries, such as Australia, Israel, Caribbean nations, and the United States take a hard line on illegal immigrants coming by sea, but Europe seems to be having a crisis of conscience over the issue.

There are two questions that must be answered:  First, is keeping Europe majority European a legitimate objective, and second, is this goal worth the cost of thousands of drowned migrants?

My view is that Europe has a unique and valuable culture that is not shared by most Africans, Arabs and Asians, even those who have lived in Europe for generations.  If Europe becomes majority non-European, that culture might die.  Majorities always oppress minorities, and an African or Arab majority in Europe would be no exception.  Whenever Africans, Arabs and Asians have conquered European territory, European civilization has suffered.  When Europeans have conquered territory, they have attempted to impose their own values on local civilizations.

European civilization is something that European people possess, and might lose if unlimited immigration is allowed.  Do people have a right to defend their possessions?  Most individuals seem to think so.  Most nations also seem to think so.  Deadly force is what ultimately backs this right.  We catch burglars and put them in prisons patrolled by guards who shoot escapees.   European civilization is certainly more important than many of the possessions we defend in this way.

Perhaps the real moral dilemma comes from the fact that while the mass of millions of migrants might threaten European civilization, individual migrants in isolation do not, and it is these individual migrants who are dying at sea.  Our moral intuitions, having evolved while humans lived in small hunter-gatherer groups, are confused by questions like this.

Some might say that the value of European civilization is reflected in Christian values, and that if we must allow migrants to drown to protect it, we have already lost it.  Others believe that European civilization means more than turning the other cheek.  Jesus himself suggested that there were values higher than charity to individual poor people.

It seems to me that Europe is past the point of diminishing returns to immigration, and that it is absurd to think that morality demands that Europeans extinguish their own civilization by allowing hundreds of millions of foreigners to move in.  Half-hearted efforts only encourage more hazardous crossings.  The only way to defend Europe and stop migrants from trying is to convince them with force that they cannot succeed in reaching Europe.

It is a trolley problem.  Achieving the best end result for the most people requires tough action against current migrants for a short period of time.  Europe will be better off with a European majority, Africa will be better off by retaining its most ambitious people, and lives will be saved if fewer people attempt to cross the sea illegally.  To achieve this end, the migrants who might have snuck into Europe will be worse off, and some will drown while the message is still getting through that crossing is hopeless.

Hand-wringing over the problem helps to milk votes and sell newspapers, but most rational people agree that sometimes you have to throw the switch on the trolley track.

Saturday, April 18, 2015

Global Warming Update

I updated my data on global temperatures, CO2 levels, the index of the Southern Oscillation Index, and sunspot activity to see if there has been any change in the relationship between temperature and CO2. 

A Chow Test still shows a structural break in the temperature trend in 1998, and there is no statistically significant temperature increase since then.  After 1998 there is also no statistically significant relationship between the 12 month change in temperature and the change in CO2, either in a simple regression, or in one where changes in the SOI and sunspots are included.  I am not cherry-picking time periods in which there is no relationship - regardless of the date endpoints, there is no consistent relationship between CO2 and temperature after 1997.

Since the 1880s there has been about a 2.2 degree Celsius increase in the global temperature, for an average of  about 0.16 degrees per decade.  ARGO data on ocean temperatures were supposed to demonstrate that all of the heat trapped by CO2 was going into the oceans, which would explain why we were not measuring significant atmospheric heating over the last 17 years.  The latest ARGO data, however, show ocean heating of about 0.02 degrees per decade since the ARGO system was activated in 2004.  This is consistent with the atmospheric readings showing that warming has essentially stopped in recent years.

Of course, the warming that occurred between late 1970s and the late 1990s was unexpected, and something similar could happen again.  But if it doesn't, how many years will it take before the elite consensus on global warming begins to crumble?

Saturday, April 11, 2015

Inequality and Age

David Steele recently gave an excellent talk defending the top one percent of the income distribution. I agreed with almost everything he said, but there was one point he made that I had to think about.  He suggested that some of the income inequality we see is due to the fact that older people earn more money than young people.  As the population ages, we should see an increase in inequality.  

Steele also made the point that as the need for skilled labor has increased, this effect should have become more important.  Long ago, when a strong body was the most important requirement for most jobs, young people could earn more than older people, but today the young lack skills and experience that gives older people an advantage.

Can the age distribution account for the magnitude of measured income inequality?  Here is a simple model that I found helpful to begin to think about the question.  Suppose that people work from age 18 to 67, a total of 50 years.  Suppose also that people receive two raises per year, so that there are a total of 100 distinct pay rates a person has during his career.  Assuming a uniform age distribution, the top 1% will therefore be exactly those people in the second half of their 67th year of age.

Suppose incomes increase by x% every 6 months, and that income at age 18 is normalized to equal one.  Then the total income of 100 people evenly distributed across all ages from 18 to 67 will be:

This is a sum of a geometric series, which is equal to: (see http://en.wikipedia.org/wiki/Geometric_series)


The income of the person in the top 1% will be:


Dividing (3) by (2) gives the fraction of income that goes to the top 1%.  Doing the algebra gives the following formula for this share:


We can compute (4) for different values of x to see if we can generate observed income inequality levels.  Here is a graph:



At a realistic growth rate of 2% every six months (4.04% per year), the top 1% only receives 2.3% of all income.  It would take growth of 30% every six months (69% per year) to produce an income share for the top one percent of 23%, which is the currently reported level.  Actual observed lifetime income increases on average by approximately 3.3% per year, and growth rates tend to fall with age. (see http://www.mrrc.isr.umich.edu/publications/papers/pdf/wp302.pdf)

So it seems unlikely to me that age explains much of the income inequality that has been reported.  I do believe that idiosyncrasies of the U.S. tax code explain a great deal of inequality, but the fact that other developed countries report similar trends in income inequality suggests that something else is also going on.  

Just because there is inequality doesn't mean we should try to reduce it, however, as Steele argues very well.