Saturday, January 17, 2015

Madness of Economists

According to the secular stagnation hypothesis, equilibrium real interest rates are negative.  Since anyone with cash can hold it in a bank account or as currency at a 0% return, the savings and investment market cannot clear, and many investment projects that savers would otherwise fund and investors would undertake will not happen.

I have written before about why I think this is wrong, but what struck me at the ASSA meetings in Boston earlier this month was the amazing things some economists are willing to consider to deal with this supposed problem.

 To believers in the stagnation hypothesis, the problem is that cash earns at least a 0% return.  They would like to fix this by eliminating currency and forcing everyone to keep all of their money in bank accounts, where it could be taxed (by the bank or the government), creating a negative interest rate for savers.

I heard Kenneth Rogoff make this suggestion, and in another session I heard Greg Mankiw suggest requiring that money be stamped every year to remain valid.  A fee could be charged for the stamp, creating in effect a negative interest rate.  Both Rogoff and Mankiw are considered conservatives.

Another way to put this is that losses on cash holdings would make savers more willing to invest their money instead of holding it in bank accounts or currency.

When famous economists say things in public, there is a chance that someone will take them seriously and actually implement their ideas, so I think it is worth thinking through the consequences.

The most obvious effect of a currency crackdown would be capital flight out of the U.S.  Savers would move money out of the country and convert it into foreign currency that had a high likelihood of at least maintaining its nominal value. The value of the dollar would fall.

The second effect would be a reduction in the size of the underground economy.  I have written about the underground economy before - probably somewhere between 10% and 20% of the U.S. economy is underground, and conducted using paper currency. 

My guess is that currency elimination advocates would see a reduction in the size of the underground economy as a feature, not a bug. Politicians and government bureaucrats see little value in economic activity that they cannot tax.  But the underground economy keeps a lot of people fed, and I believe that it has large multiplier effects.  Eliminating cash would eliminate a lot of shadow jobs, and many newly unemployed people would end up on public assistance.

A third effect, which might not be so bad, would be the emergence of private alternatives to government-issued cash.  Bitcoin and gold would be big winners.  Merchants would expand their gift card offerings, which could function as zero interest cash.  Eventually these alternatives might replace dollars for everything except payment of taxes.

Actually, I might have just talked myself into supporting the elimination of paper dollars - the private alternatives that would spring up might result in huge net benefits!  Maybe that reasoning makes me one of the mad economists.

Saturday, January 10, 2015

ASSA Meetings - Second Best for Europe?

Happy New Year, and thank you for visiting my website!

I spent time in New Orleans, Boston, and Williamstown over the last few weeks, and I hope to write about all of them over the next several days.  I'll start with some stories about the ASSA meetings - an annual three day gathering of thousands of economists, this year in Boston.

I seem to recall sessions from the late 1990s in which speaker after speaker praised the coming of the euro and predicted strong economic growth in Europe following currency integration.  Now, of course, everyone claims that they said all along that the euro was a big mistake.  But in fact, many economists did warn of possible problems well before the euro was launched - perhaps they weren't speaking as loudly at the time, or perhaps I only heard what I wanted to hear.  Paul Krugman, for example, was clear at the time that the euro was a bad idea.  Here is an article by Rudy Dornbusch from 1996 warning that a euro would be a mistake.  Martin Feldstein wrote in 1997 that
"the adverse economic effects of a single currency on unemployment and inflation would outweigh any gains from facilitating trade and capital flows among the EMU members."
 The thinking of American economists who favored the euro is described here.  Robert Mundell continues to defend the euro, saying that Europe needs to fix its real structural problems instead of papering over them with monetary policy.  The answer to Mundell appears to be that Europe cannot fix its structural problems, and so it needs to keep a bag of tricks available.  This is a classic "theory of the second best" problem, which I wrote about here.

I attended a session in Boston on the euro in which Kenneth Rogoff, Martin Feldstein, and Jeffrey Frankel all agreed that the euro had been a "big mistake."  I was surprised to hear Feldstein advocate what he called "revenue neutral tax incentives" for Europe.  In other words, they should load up their tax codes with loopholes and raise tax rates.  This is exactly the opposite of what Feldstein has advocated for the U.S.

It was puzzling at first to hear free market economists advocating interventionist policies, but I think the reason must be that they believe that real reform in Europe is impossible, so we should give up on the idea of market solutions and think about how the state can best manage the economy.  Only Mundell clings to the hope that by taking away gimmicks like currency devaluation, countries will be forced to adopt true free market policies.

I wonder how long it will be before most mainstream economists in the U.S. give up on the possibility of structural reform here, and focus instead on optimal government direction of the economy.

Saturday, December 20, 2014


I will take a break from blogging until sometime in January, but in the spirit of the holidays I thought I should link to some articles I agree with in publications I sometimes criticize. 

I agree completely with Paul Krugman about Russia, except the gratuitous shot at Giuliani.  There is more in his blog.  (Giuliani was not taking Putin's side, he was saying that Obama should be tougher and more decisive in order to counter him.)

There was also very good piece today on New York's rejection of fracking while it encourages casino development - a very poor economic strategy.  Nuclear power is the best option - much better than oil and gas from fracking, but since the environmental movement has crippled the nuclear industry, things like fracking are necessary.

I also think it is worth saying that the United States is ending 2014 with strong economic growth and political stability.  The Republican comeback demonstrates once again that the two party system is remarkably resilient.  Whenever one party looks like it has been defeated forever it comes back in less than a decade.  By absorbing fringe movements, the parties defang revolutionary movements while preserving the illusion of choice.  The U.S. is also slowly but surely making progress against its adversaries around the world without major military activity.  All in all, we seem to be entering a time of peace and prosperity unmatched since the late 1990s - also a time with a weakened Democratic president and a Republican congress.

Merry Christmas and Happy New Year!

Sunday, November 30, 2014

Homeownership and Newspapers

Today's lead editorial in the New York Times extolls homeownership, in spite of the disasters caused in recent years by government promotion of it.  Is it possible that the Grey Lady is biased by the fact that homeowners are more likely than renters to subscribe to print newspapers?  Renters are able to move easily to take advantage of job opportunities, better schools or changing neighborhoods, but frequent movers often neglect subscriptions to dead tree publications.  Secure apartment complexes can also be more difficult for newspaper delivery than single family homes.

Of course this couldn't be true, any more than it could be true that the Times is biased in favor of immigration because a major shareholder made his fortune on cross-border phone calls.

Thursday, November 20, 2014


The first major confrontation between President Obama and the new Republican congress will apparently be over immigration.  I hesitate to write about immigration because it is such a touchy topic, but I suppose the time has come to stop ducking it.

The first question to ask about immigration is whether we favor completely open borders.  Before thinking that this is an absurd straw man, recall this quotation from a Wall Street Journal editorial in 1984:

               ...we propose a five-word constitutional amendment: There shall be open borders.

Economists tend to support higher levels of immigration, and some support completely open borders.  What would be the effect of completely open borders on the U.S.?  According to a Gallup survey, 150 million adults around the world would like to move to the U.S.  If other destination countries kept their borders closed, a large number of the total of 640 million adults who would like to move somewhere would come here.  Open borders would trigger a rush of immigration, since many immigrants would worry that the border might later close and would want to get in while it was still possible.

Each adult who immigrated would probably bring an average of one to two family members, for a total that might exceed the current population of the U.S.  Over time, the number of immigrants would probably grow, because families would attract extended family members, and immigrant enclaves would ease the transition of additional immigrants less eager to assimilate into native American society.

My guess is that the vast majority of my readers would oppose a rapid doubling of the population of the U.S. from immigration.  Why?  Probably because they like things as they are and are reluctant to gamble on the political, cultural, and economic changes that might result.

On the other hand, most of my readers probably believe that immigration has benefitted the U.S. in the past, and that some level of immigration is desirable.  So what would be the best amount of immigration?  And who would make the best immigrants?  The answer depends on the economic, political, and social effects of immigration.

My view is that the U.S. government should act to maximize the well-being of U.S. citizens.  If immigration benefits current citizens, we should have more of it, but if it doesn't we should have less.  Of course, if we can benefit non-citizens at a trivial cost to citizens we should consider doing so.  But the New York Times editorial welcoming Obama's speech made no mention at all of possible benefits of Obama's actions to current citizens, only benefits to immigrants themselves.  The Times made its position clear on Sunday, saying the following while arguing against admitting Cuban doctors to the U.S:
American immigration policy should give priority to the world’s neediest refugees and persecuted people.
By legalizing current residents who entered the country illegally, Obama will encourage many more people to come.  They will not be deterred by his talk of border security - they will rightly reason that those being legalized now entered illegally, and so if they enter illegally they will probably be legalized sometime in the future.  Will that help current U.S. citizens?


Bringing in foreign doctors would lower medical costs and improve access to medical care.  Bringing in accountants would lower business costs.  Bringing in low-skill workers lowers wages of native unskilled workers, hurting them and benefiting their employers.  Most immigrants into the U.S. are unskilled, probably because doctors, accountants and employers have better political representation.

To determine the overall effect of low-skill immigration we need rough estimates of the government services consumed by immigrants, and the economic benefit of their employment in the U.S.

How much does government spend on new immigrants?  A commenter at marginalrevolution last year pointed out that total US healthcare spending per year divided by total hours worked in the US would tell us something about which workers earn enough to cover average healthcare costs.  Assuming that lifetime expected healthcare costs are the same for everyone, if immigrant wages are less than average healthcare costs per hour, then it might be true that the native population is subsidizing immigrants. (the answer also depends on benefits to employers, which I will discuss)

Total healthcare spending in the US is estimated to be $3.8 trillion in 2014.  Subtracting out discretionary healthcare costs and all out-of-pocket expenditures, the amount is $2.8 trillion.  I subtract these amounts because poor immigrants might do without these services.  The idea of immigration reform and Obamacare is to provide medical insurance to everyone, so the remaining $2.8 trillion is an appropriate amount to use.

According to the Conference Board, the total number of hours worked in 2013 was 248.61 billion.  Division produces a result of $11.26 per hour.  The same calculation for education adds another $1.80 per hour.  Assuming that immigrant households receive 46% more welfare than native households, welfare spending adds another $2.91, for a total of $15.97.  The overall effects of immigration on the federal budget are discussed here.  As George Borjas put it:
There’s also been a lot of fake fog thrown into the the question of whether immigrants pay their way in the welfare state. It’s time for some sanity in this matter as well. The welfare state is specifically designed to transfer resources from higher-income to lower-income persons. Immigrants fall disproportionately into the bottom part of the income distribution. It is downright ridiculous to claim that low-skill immigrants somehow end up being net contributors into the public treasury.
On the other hand, immigrants lower wages, benefiting employers and in turn, consumers of products made by these employers.  I did a very simple calculation of the "immigration surplus" by specifying demand and supply curves for labor. This is a very, very simple model and I don't claim much for it, but I think it might give some idea of how to think about the costs and benefits of immigration, and why it is probably the case that it costs more than it comes to.

I assumed that no one in the U.S. would work for less than $0 per hour, and that no one would hire anyone for low-skill jobs for more than $30 per hour.  Denmark pays fast food workers $20 per hour, but how high could low-skill wages go before workers are replaced by automation?  Assuming that a single robotic arm costs $100,000 and fast food work consists of at least 6 tasks, I assume a cost per employee replaced of $600,000.  Assuming a 12 year machine life and a long-term corporate borrowing rate of 3.5%, and assuming that maintenance costs are similar to costs of maintaining human employees (training, uniforms, HR departments, etc), I think that low-skill workers would be replaced if their wages rose to $30 per hour. Since approximately 3.3 million people are employed at or below the minimum wage of $7.25 per hour, I have the other point I need to determine the demand and supply curves. 

I model immigration as a pivot of the labor supply curve.  If a single low-skill immigrant comes to the U.S., the wage floor is still zero, but at equilibrium one more worker is employed.  This additional worker drives the wage down slightly below the minimum wage, which might happen with illegally low payments, unpaid overtime, worse working conditions, etc.

The overall addition to worker and employer welfare from adding an additional worker is shown by the lavender area. Since everything is linear, it is simple to calculate the areas of the triangles formed by the supply and demand curves.

I did the same calculation for non-linear supply and demand curves.  I just assumed that the wage was proportional to the log of employment plus a constant and fit the points required by the assumptions. Finally, I took the integrals of the curves to calculate the areas under them. The curves look something like this (not to scale):

I calculate the total economic surplus added by an additional immigrant to be $15 per hour using the linear model, and $1.87 using the log model.  Either way, the surplus is less than the $15.97 estimated cost per hour.  If the true economic surplus is somewhere in between, the net cost of low skill, illegal immigration in the U.S. might be between 0.14% and 2% of GDP.  These are not numbers that will bankrupt the country, but they are not trivial either.  Either way, it is difficult to believe that low-skill immigration provides net economic benefits.

It is important to note that the benefits go to employers, while native workers lose.  Conservatives often justify income inequality by arguing that it is needed for overall economic growth, but with low-skill immigration we get more income inequality and lower overall economic well-being.


From a political point of view, immigration is always bad for Republicans.  As I argued a year ago, immigrants are most likely to vote for Democrats.  So Republicans understandably begin the debate biased against immigration, while Democrats are biased in favor.  If immigrants happened to vote Republican, these biases would be reversed.  Given current realities, anyone favoring lower taxes and smaller government has a good reason for opposing any additional immigration. 

Republicans were happy that they almost received 40% of the Hispanic vote in 2014, but they will do worse than that in 2016, a presidential election year. 

There is also a case to be made that most countries from which the US attracts immigrants are far less democratic than the US.  Immigrants bring their own political ideas with them, and might water down support for some traditional American ideas and freedoms, as Pim Fortuyn argued in the Netherlands before he was murdered for saying so.

Setting partisanship aside, importing new voters as a political tool just doesn't seem fair.  As the playwright Bertolt Brecht put it (although he was not talking about immigration):
Would it not be easier
In that case for the government
To dissolve the people
And elect another?


The social and cultural effects of immigration are complicated.  A diverse population brings an interesting variety of food, music, ideas and lifestyles.  On the other hand, traditional American food, music, ideas and lifestyles are comfortable and desirable for many Americans.  Too much diversity might crowd out traditional American culture, whatever that is, just as 19th century immigration crowded out traditional English colonial culture in America.  Just as many people bemoan the growth of chain businesses at the expense of mom-and-pop restaurants, stores and other businesses, too many immigrants around the world might homogenize world culture, dissolving native cultures.  

Of course, this has happened many times in American history.  Waves of immigrants swept away traditional lifestyles and institutions in many cities and rural areas.  Immigration strengthened and invigorated the country in many ways, but old ways of thinking and doing things, some good, some bad, were lost due to massive immigration. There were important positive effects of immigration, but there were also negative effects, such as city government corruption, labor violence, organized crime, and disease in overcrowded cities.

Immigration was severely curtailed from 1921 until 1965, with negative and positive social and cultural effects.  Cultural diversity declined, and perhaps the perceived blandness of American culture contributed to the youth rebellion of the 1960s.  On the other hand, national cohesiveness and trust in institutions were at high levels.  A recent academic paper summarized the results of Robert Putnam as follows:
for the United States, that at least in an unspecified short run ethnic diversity is likely to reduce social solidarity, social capital and particularly social trust between citizens.
Robert Putnam is not anti-immigration, but his work demonstrates that there are some social costs of large-scale immigration. The current academic consensus appears to be that there is evidence suggesting that diversity in North America does have some cost in terms of social capital, but the situation may be different in Europe.

Disruptive effects of immigration are not always by accident.  In the UK, for example, the Labour Party allowed mass immigration in order to "radically change the country and 'rub the Right's nose in diversity.'"


Massive immigration of low skill workers costs more in added government expenditures than the economic benefits it provides, and at the same time it exacerbates income inequality within the U.S.  It also has potentially disruptive social and political effects.  The effects of high skill, high income immigrants would be different, and possibly more beneficial to Americans.

By legalizing immigrants who entered the U.S. illegally, President Obama will encourage massive new illegal immigration of people hoping for future amnesties.  The vast majority of these immigrants will be low-skill, low-wage workers and their families, and they will cost U.S. taxpayers much more than their low wages will benefit U.S. employers.  They will also have disruptive political and social effects.  Even the Democratic Party might lose if more native voters turn against them than new immigrants vote for them. 

But employers of low-wage labor will win, and unions believe that they will organize newly legalized immigrants.  Bleeding hearts favor immigration regardless of the costs, as do libertarians who oppose border restrictions on principal, regardless of the additional taxation it might cause.  This is apparently a very strong coalition, stronger than the majority of the population that will pay the price.

Friday, November 14, 2014

Another Global Warming Update

President Obama has announced major initiatives in two policy areas for his last two years in office:  global warming and immigration.  I will discuss immigration next week.  For this week, I noticed that it has been a year since my last global warming update.

Last year I argued that the most reliable data show no statistically significant warming since 1997, and no statistically significant relationship between temperatures and CO2 levels over the same period.  My main caveat was that new data was expected from Argo, a set of deep ocean probes that might show that the missing heat is being stored in the world's oceans instead of in the atmosphere.

Data are starting to come in from Argo, and they show negligible ocean heating.  A longer time series is needed to have much confidence in the results, but for now there are no data to contradict the lack of observed atmospheric warming over the past 17 years.  Of course, if these small amounts of data showed significant ocean warming there would have been national headlines.

I have updated the regressions I ran last year, and the results are the same - a Chow Test shows a structural break in the temperature time series in 1997, with a statistically significant increase before that time, but no statistically significant rise afterward.  Multiple regression of the 12 month change in temperature on the change in CO2, sunspots, and a measure of the intensity of the El Nino effect in the Pacific Ocean shows no statistically significant effect of CO2 on temperature from 1998 to 2014.

The earth has been warming off and on since the last ice age, with a relatively rapid increase from 1975 to 1997, a time when global CO2 levels also increased.  Given theoretical reasons to expect CO2 to increase temperatures, the idea that human activity was causing warming was plausible.  It is certainly a long way from being ruled out, but 17 years of flat temperatures while CO2 levels continued to increase is good reason to doubt the theory.  I think these doubts, plus uncertainty about the true costs of warming should be enough to convince us to put the anti-warming program on hold.

Wednesday, November 5, 2014

Democratic Dilemmas; Republican Resurgence

Last night's election was about age and race.

There have been two sets of offsetting trends in US elections over the past 20 years:
  • Young voters increasingly prefer Democrats, but a decreasing percentage of voters are young.
  • White voters increasingly prefer Republicans, but a decreasing percentage of voters are white.

Since the early 1970s, I think that the attention young people pay to conventional economic and foreign policy questions has declined, but the attention they pay to issues involving group identity has increased.  Why?  Perhaps college students and recent graduates prefer not to dwell on issues on which they might feel pressure to take a stand against their own economic interests.  At the same time, they find refuge in support for policies that they believe will cost them nothing but will still convince others of their political correctness.

As college attendance rates have increased, the average academic ability of those not attending or graduating from college has declined, and so has their understanding of complex political issues, so they also focus on simpler issues involving group identity.

A combination of increased cynicism and ignorance has caused the decrease in young voter participation.


Democrats have capitalized on the opportunity presented by the changing preferences of young voters by increasingly emphasizing group identity issues.  The dilemma for Democrats is that this strategy tends to alienate white voters.

The critical task for Democrats is to appeal to blacks and young whites who care about racial issues without alienating whites.  In 2008 they found the perfect tactic - nominating a black man.  White voters are happy to vote for black candidates, but they mostly disagree with nearly all actual policies that blacks believe would benefit black people.  Black voters like voting for black candidates so much that they are willing to overlook the fact that once elected, they do not deliver on issues they care about.  Obama won because he appealed to all voters, and was able to placate whites without alienating black voters.

White Democrats will have a much harder time running for office.  They won't appeal to black voters unless they deliver something substantial to blacks, but if they do they will alienate whites.  Eventually black politicians might also need to deliver results to maintain black support, but then they will also alienate whites.


Suppose that minority voters all vote Democratic, and that the percentage of voters who are minority increases linearly.  At the same time, the percentage of white voters who vote Republican increases exponentially.  This occurs as whites increasingly fear the power of growing minorities.

I start in the year 2000 with even strength for both parties and 25% minority voters.  Minority voters increase their share, reaching a majority in 40 years.  Republican support among whites increases proportionately with the square of the number of years since 2000, calibrated so that 100% Republican support among whites is reached in 20 years.

The percentage voting Democratic evolves as follows:

There is randomness in election results, so not all elections would exactly follow this pattern, but results would roughly move in the indicated directions. 

The 2000 election is 50/50.  Democratic strength increases, reaching a peak around 2008.  By 2014 Republicans have the advantage, and their advantage continues to increase until they capture the support of all white voters.  As minority voters increase in number, however, Republicans eventually start to lose again.

Of course, Republicans could use their electoral dominance from 2014 to 2039 to change the demographics of the electorate, perhaps with immigration policy, or perhaps with policies designed to selectively discourage or encourage voter turnout.  Democrats attempted to do so during their period of dominance with some limited success.