Thursday, November 29, 2012

The Fiscal Cliff: a non-Keynesian Analysis

Much of the discussion and analysis of the fiscal cliff is purely Keynesian: incentives are ignored and it is assumed that moving money around has a multiplier effect (negative in this case, because it involves moving money "from the consumer" to the Treasury).

Here's a back of the envelope on incentives:

  • Emergency unemployment benefits are ended "prematurely".  Keynesians say that end depresses the economy, but it really expands it because it reduces the pay people can receive by not working.  In terms of marginal tax rates for middle income people, this provision will lower them about 2.2 percentage points (see Chapter 5 of The Redistibution Recession, this is expressed as a percentage of total  compensation).
  • "Bush rates" for individual income tax expire:  rates rise about 3 percentage points (using the same compensation units as above -- not the usually units you see cited).
  • The payroll tax cut expires: raises rates 1.8 percent points (same units as above).
  • AMT hits a lot more people.  If the Bush tax cuts had not expired, that would probably rate rates about one point.  But the cliff  expires the Bush rates, so it is possible that adding on the AMT lowers rates.  So let's call it zero.
  • TOTAL: marginal tax rates for middle income people go up 2.7 points.  Their after tax share falls by 0.049 log points (that is, about 5 percent).
IF all of this happened (it won't), my model predicts that the labor market shrinks 1.8-3.7 percent relative to the constant marginal tax rate baseline. At the baseline, the labor market grows at something like population growth (+1 percent over one year).  So maybe the labor market shrinks 0.8 - 2.7 percentage points in absolute terms.  That's a noticeable recession, but much smaller than what happened 2008-9.

I doubt that the Bush rates expire on the middle class (the rich don't count much in the employment statistics), so that entry can be forecast as zero.  So then marginal tax rates actually go down (the emergency UI expiration more than offsets the payroll tax cut expiration -- see my book on this point), and the labor market grows a bit faster than one percent in absolute terms.

If the Bush tax rates were maintained (at least on the non-rich), and just part of the unemployment benefits were to expire, then marginal tax rates might edge up a bit.

In my view, the tax rates on the rich matter little for what happens to employment in the short run, because the rich don't count much in the employment statistics.  They count much more in the spending and productivity statistics, which would be depressed by returning to the Bush rates for them. 

What could be interesting is if we go over the fiscal cliff, have a mild recession, but the entry into that mild recession opens the spigots for still more help for the poor and unemployed (think ARRA), and turns a mild recession into a deep one (sound familiar?).

The bigger issue for the labor market in the medium term is what happens with the ACA.  If that goes in as planned, it should be a lot more depressing than the fiscal cliff (although I am still preparing my estimates here -- not a simple law).

Wednesday, November 28, 2012

A Time for More Nations

Copyright, The New York Times Company

Catalonia, which includes Barcelona, has long been a part of Spain, but its peaceful residents increasingly talk about being an independent country again. In elections over the weekend, where independence was one of the most discussed campaign issues, a majority of offices were won by parties that support more Catalan independence, in one form or another.

An independent Catalonia would reinforce a worldwide trend. The world’s economics and demography are changing, and economic theory predicts that national borders will change with them (see “The Size of Nations,” by Alberto Alesina and Enrico Spolaore or “A Theory of the Size and Shape of Nations,” by David Friedman in The Journal of Political Economy).

The number of countries has grown since World War II, especially since 1990. The Soviet Union broke into multiple nations. Czechoslovakia split into the Czech Republic and Slovakia; Yugoslavia dissolved.

In most cases, many citizens of the parts wanted independence from the larger whole. The large countries were often divided by language or ethnicity and were often held together by nondemocratic leadership. The new independent countries emerged as democracy took hold, or shortly after.

In a few instances, countries combined. East Germany and West Germany unified. North and South Vietnam became one when North Vietnam won the Vietnam War. North Yemen and South Yemen were unified. As their names suggest, they have some common language, culture and history, more so than many former Soviet republics did.

Catalonia has its own language, Catalan, and a long history. Under Franco, Spain suppressed many Catalan institutions. And labor was mobile in Spain during the Franco regime, with many Spanish-speakers moving to Catalonia, Spain’s most prosperous region. The prevalence of Spanish in Catalonia, as well as the heavy hand of Franco, may have undercut an independence movement. But Franco’s death in 1975 and the emergence of democracy in Spain did not foster an independent Catalonia.

New generations have been learning Catalan, however, and that may be tipping the balance toward independence.

Catalonia objects to the amount of taxes it pays Spain’s central government, compared with the benefits it receives. One potential step would be to address that situation without full independence, by having Spain’s central government “charge” Catalonia less for being part of Spain by providing tax breaks or more public services.

As governments and redistribution grow, richer regions find taxes to be increasingly burdensome. With the cold war over, ethnically unique regions no longer perceive the same national security benefits of being part of a larger nation.

Catalonia’s situation is worth watching, as it may hold lessons for Libya, Iraq and even the United States, where regions sometimes diverge in terms of culture, language and preferences for governing. A small nation can be established peacefully and may prove to have long-term benefits.

Tuesday, November 27, 2012

Myth: Constant Marginal Tax Rates in Europe

One of the myths (perpetuated here and here) about the labor market since 2007 is that marginal tax rates have been constant in Europe. Somebody needs to carefully measure marginal tax rates around the world (my book provides a framework for doing so -- see chapter 3), but with news stories like this (Portugal), this (UK), this (France), this (Japan, admittedly not Europe), and this (Ireland), I don't see how one could assume that European countries have had constant or declining rates. "Austerity" is by no means the same as cutting marginal tax rates.

International readers, if you have information or anecdotes about marginal tax rate changes, please leave a comment. Don't forget that so-called "tax cuts" do not cut marginal tax rates if only-low income people are eligible for the cut. For the same reason, government benefit cuts do not cut marginal tax rates if the cuts occur primarily by withholding benefits from people with high and middle incomes.

Sunday, November 25, 2012

Recession by Redistribution

Doubt of the benefit
Why increasing unemployment aid is prolonging the recession


More families used food stamps this past Thanksgiving than ever in history, while Congress is pushing to extend benefits — again — for the longterm unemployed.

But what if such aid isn’t helping us weather the recession, but instead prolonging it?

The White House, and other believers in Keynesian policy refer to subsidies to the unemployed, poor and financially distressed as “automatic stabilizers” and insist that subsidies have a large positive effect on national income.

Yet when the subsidy spigots were opened wide in 2008 and 2009, labor market activity contracted sharply, and stubbornly refuses to rebound.

Getty Images
It is time to reconsider the old-school economic idea that paying people to be unemployed reduces employment. The more we pay poor people, the more poor people we will have. The more we help people and institutions in financial distress, the more financial distress there will be.

It is easy to look at a particular instance of redistribution — say, unemployment benefits — and conclude that its aggregate effects are minimal, or approximately zero. But policymakers did not expand just one provision of one program.

Food-stamp recipients were given a big raise in October 2008, and then another raise six months later. Thanks to the elimination of asset-testing by the majority of states, just about anyone who is the sole earner in their household now find themselves eligible for food stamps during periods of unemployment.

The American Recovery and Reinvestment Act, popularly known as the stimulus, gave unemployment insurance recipients a weekly bonus, and offered to pay for the majority of their health insurance expenses. FDIC and Treasury reduced some “unaffordable” mortgage payments, which means that successful people need not apply. The list goes on and on.

The essential consequence for all of these is the same: a reduction in the reward to activities and efforts that raise incomes.

I’ve studied how redistribution affects the “reward” for working.

We start with a monthly index of government benefits. Before the recession began, an unemployed person typically received about $10,000 a year in government benefits. By the end of 2009, program rule changes alone had increased the typical benefit to almost $16,000.

Plot that against the hours an average American adult spends away from work in a year (the difference between total hours in a year and hours at work). Largely because of the increase in the number of people without jobs, the average work hours were about 120 fewer at the end of 2009 than they were at the end of 2007 — a 10% decline.

But things start to change at the beginning of 2010. Slowly, the number of hours of work by the average American begins to climb. Not coincidentally, the average annual government benefit for the unemployed dropped to $14,000.

The increase in benefits provides a disincentive to work. From 2007 until 2009, I found a startling 13% decline in the “reward” for working — that is, how much better the average job would be over collecting benefits.

Considering that, why is the labor market still so far from a full recovery? Because government benefits are still far from returning to pre-recession levels.

But wait, a Keynesian would say. Unemployment benefits are good for the economy overall, since it is money spent and not saved.

It’s true that the poor and unemployed tend to quickly spend what they have on basic needs. Yet Keynesians have gone further to claim that spending patterns of the poor are why redistribution raises total spending and thereby employment. Redistribution changes the composition of spending and employment in the direction of industries like discount groceries and low-cost retail that disproportionately serve poor customers and away from industries like, say, airlines. The stimulating effects of benefit spending for the overall economy is limited.

Redistribution is not free. Redistribution depresses employment, aggregate spending and GDP, by implicitly punishing the successful and implicitly rewarding the unsuccessful.

We don’t like to see people suffer, and it’s a natural instinct to want to increase redistribution in a time of recession. But the better economic solution reduces the implicit penalties of government aid, and gets more people working, so they don’t need the help.

Casey B. Mulligan is a professor of economics at the University of Chicago and author of the new book “The Redistribution Recession” (Oxford University Press);

Friday, November 23, 2012

USDA edits food stamp report, but embarrassing stats remain

For at least 7 consecutive years (going back at least to FY 2004), the USDA had explained in its annual food stamp report how the program was growing because of new eligibility rules and USDA outreach efforts. They would also comment on how the state of the economy was an additional factor expanding program spending and participation or, in some of the economic-expansion years, how the economy by itself was a factor that would have reduced participation and spending.

Here is a typical (FY2009) example from the executive summary:

"The continued growth in SNAP participation from 2008 to 2009 is likely attributable to the deterioration of the economy, expansions in SNAP eligibility, and continued outreach efforts." [emphasis added] (FY2009, p. xiii)

See also FY 2010 (p. xv), FY 2008 (p. xiii), FY 2007 (p. 11), FY 2006 (p. 10), FY 2005 (p. 10), and FY 2004 (p. 9).

Now, for the first time in years, such discussion is absent from the SNAP report. The words "expansion" and "outreach" are now absent from the report. But look at the very first chart in the report (click to enlarge).

Our labor market has not been doing well lately, but it did not contract in FY 2010 as it did in, say, FY 1991, FY 2001, or FY 2002. Nevertheless, SNAP participation grew more in FY 2010 than it did in those years. I suggest that USDA add this sentence to its executive summary:

"The continued growth in SNAP participation from 2010 to 2011 is likely attributable to expansions in SNAP eligibility and continued outreach efforts."

Although the USDA does not offer an opinion as to the reasons for SNAP participation growth, they do cite some facts that scholars might find useful:

In fiscal year 2011, Colorado, Hawaii, and Iowa adopted BBCE policies for the first time. California, Maryland, and Minnesota expanded existing BBCE policies, increasing SNAP eligibility in those States, while Idaho and North Dakota restricted existing BBCE policies, decreasing SNAP eligibility in those States. In particular, California and Maryland expanded their policies to include households without children, and Minnesota expanded their policy by dropping their asset test and raising their gross income limit to 165 percent of the poverty guideline for all households. Conversely, Idaho and North Dakota restricted their policies by adding an asset test and gross income test, respectively.

For more on the mutual feedbacks between the economy and the food stamp program, see my new book.

Wednesday, November 21, 2012

The Future of Employer-provided HI and the Market

Copyright, The New York Times Company

The future of employer-provided health insurance is better considered together with the future of total employee compensation, both cash and fringe benefits like health insurance. From that perspective, the likelihood that most employers will continue to offer health insurance is not necessarily good news for employees.

The Patient Protection and Affordable Care Act, President Obama’s initiative, offers large health-insurance subsidies to the majority of the population beginning in 2014, but only if their employer does not offer affordable insurance. The subsidies are frequently much larger than the subsidies coming through the tax exclusion of employer-provided health insurance.

Some economists are predicting that eligible employees, especially those in line for the largest subsidies, will prefer employers who do not offer affordable insurance. As a result, they say, many more employers will not offer insurance.

Others have different expectations, pointing out that employers dropping insurance will pay penalties and throw away the tax exclusion for their employees who are not subsidy-eligible (typically the ones who earn more). Moreover, perhaps because people are comfortable with their existing coverage even if it is not subsidized, employer coverage did not decline in Massachusetts when it began a similar plan (by my estimate, only 5 percent of the people in Massachusetts who could get subsidized individual-market insurance actually receive it, largely because they have coverage through the employer of the head of the household or that person’s spouse). Note that Massachusetts has lower subsidies and a narrower eligible population than the Affordable Care Act and lower employer penalties for dropping coverage.

How many employers will drop their coverage when the new health care law gets under way? The answer makes for a nice headline, but that’s the wrong question. Would it be so bad if many employers dropped their coverage but replaced it with huge cash raises? Or would it be so good if every employer continued to offer coverage but required employees to take big pay cuts?

All sides agree that some otherwise subsidy-eligible employees will work for employers that keep their coverage, and other subsidy-eligible employees will work for employers that drop it. Market forces must be considered, because some employees will be moving between these two types of employers.

Low-income employees will ultimately cost less to employers without coverage (or without “affordable” coverage; the important issue is that their low-income employees are subsidy-eligible) than they cost to employers with coverage. If they didn’t, low-income employees would be better off at employers without coverage and would line up to work there. Meanwhile, the employers with coverage would find it more difficult to retain and attract low-income employees. That situation defies supply and demand.

Another way to see the same result: by getting low-income employees at lesser cost, employers without coverage can, without going out of business, compete aggressively for the high-income employees who are considering positions that offer coverage.

By the same logic, high-income employees will cost more to employers without coverage than they do to employers with coverage. Thus, high-income employees will lose one way or another — either they will lose their tax exclusion because their employer eliminates coverage or they will see their cash compensation fall below what it would have been without the Affordable Care Act.

At the same time, the low-income employees will enjoy the subsidy either way: either their employer drops coverage, in which case they receive the subsidy directly, or their employer increases their compensation above what it would be without the Affordable Care Act to attract them from the employers without coverage. (Tax economists will recognize this as the Harberger model applied to the Affordable Care Act; international economists will recognize it as the Heckscher-Ohlin model.)

The same sorts of market competition will ultimately prevent most employers from dropping their coverage and thereby incurring the penalties. Employers keeping coverage will raise the pay of subsidy-eligible employees and get by with fewer of them. Those who remain will typically not want to leave for no-coverage employers because doing so would cut their pay. The same employers will hire a few more high-income employees at lesser pay, because for those employees, the alternative is a no-coverage employer.

Sunday, November 18, 2012

Obamacare and the Quantity of Labor

The Patient Protection and Affordable Care Act (ACA) subsidizes health insurance for many, but not all, persons.  It also levies a fine on employers who do not offer health insurance.  The net result of all of this will be to reduce employment, especially among less skilled people.

First Approximation: ACA = $2000/Employee Levy on ALL Employers
To analyze this, I break the ACA subsidies (as of 2014) and taxes into conceptual pieces that have the same sum total as the ACA itself
  1. A levy on ALL employers in the amount of $2000 for each employee
  2. A subsidy to employers offering health insurance in the amount of $2000 per employee
  3. A subsidy to insurance-purchasing employees at employers not offering health insurance
  4. A subsidy to insurance-purchasing people not working
You won't find either item (1) by itself or item (2) by itself discussed in connection with the ACA.  Rather the combination of (1) and (2) are usually described as a bundle: as a $2000/employee levy only on employers not offering health insurance.  But we'll see that it's easier to analyze them separately.

Suppose for the moment that (3) and (4) were each about $2000 per employee or potential employee.  In this case, the combination of (2)-(4) has no effect on employment because non-employment and both types of employment relationships all receive the same subsidy.  That leaves us with (1), which obviously reduces employment, especially low-skill employment.

Take, for example, someone who would be working full-time and full-year at minimum wage: the $2,000 levy would increase his employer cost by 13 percent.  It would increase the employer cost of the median full-time employee by 5.5 percent.

Because labor demand is more wage elastic than labor supply is, workers would bear the majority of a $2,000 levy on employers in the form of lower wages.  Each one percentage point increase in employer cost would reduce employment by 0.36 to 0.83 percentage points: at the median that's an employment per capita decline of 2.0 to 4.5 percent: a recession by any definition.

In fact, the average subsidy per employee or potential employee is greater than $2,000, so the net employment effect of (2)-(4) is not zero.  (2) is smaller in magnitude than (3) and (4), which means that people are pulled from employers offering HI to non-employment or employment without employer-provided HI.  In other words, the combination (2)-(4) depresses employment, and raises wages, in addition to the effects of (1).  Perhaps the overall employment effect of (1)-(4) is in the range -2.5 to -5.9 percent.

The direction of the overall wage effect of (1)-(4) is ambiguous, but I suspect that it is slightly negative. Employer cost will go up, but by less than $2,000 per employee.  This means that the ACA will drive a wedge between labor productivity and wages in the amount of about 5 percent of productivity at the median (perhaps 4 percent on average).

ACA Elements Not Yet Considered Here
I haven't thought much about the dynamics of these effects yet, but I doubt that it will all happen on January 1, 2014.  Presumably some will happen before and some after.  Interestingly, we began to see wages fall about 2-4 percent behind productivity beginning almost exactly when the ACA was debated and passed.

This does not consider the subsidies to small employers providing health insurance, and the levy exemptions for small employers not providing health insurance.  The exact size of the wedge between wages and productivity depends on the degree to which employer subsidies offset the revenue from the employer levy.  Nor does the above consider the expansion of Medicaid coverage (although that expansion is considered in my book).

Between 2014 and 2018, the levy and subsidies grow, which means that the employment impacts should grow too.

**The above does not consider that the subsidies (3) and (4) are on a sliding scale function of household income.  As a result of those provisions alone, some households will see their marginal tax rates increase by 40 percentage points.  I have not yet quantified the effect of the sliding scale on nationwide average marginal tax rates, but it is significant and in the direction of depressing employment.

Previous Studies have Underestimated the ACA's Employment Impact
Some economists say that Massachusetts' healthcare law, which shared many features with ACA, did not reduce employment in the state.  However, the MA levy was only $295 rather than $2000, and people in MA tend to make more than they do in the rest of the U.S.  Thus, I expect the ACA to depress employment per capita at least 6-7 times more than MA's law did.  In other word's, my approach says that the MA law would reduce employment per capita in MA by 0.3 to 0.8 percent, which is small enough that an econometric study might not detect it and significantly less than population growth over the time frame that the MA law matured.

(Hardly anyone was on the MA version of the sliding scale I mentioned above).

I also understand that the CBO says that the ACA would depress employment by only 0.5 percent.  Scaling by 10, I guess that means they think that a $20,000 levy would reduce employment by only 5 percent?!  When it comes to estimating employment impacts, the CBO keeps repeating the same flaw: they say that transfers and related government transactions expand employment through a "multiplier" that (at best) applies to government purchases of items (like military spending) that individuals would not purchase for themselves.  Their approach is radically different from a labor supply and labor demand analysis, which recognizes that transfers and other subsidies reduce the incentive to work and raise employer costs.

Wednesday, November 14, 2012

Job Openings: What do They Mean?

Copyright, The New York Times Company

A high ratio of unemployed to job openings means that the unemployed are competing a lot for jobs, many news reports say, when in fact it could indicate the opposite.

It’s true that a reduction in labor demand — from, say, a new tax on employers — would motivate employers to get by with fewer employees. As they do, employers would reduce job openings and lay off workers. One result would be fewer job openings and more unemployed people, and thereby more unemployed people per job opening.

But a reduction in labor supply in the form of additional subsidies for unemployed people would have similar effects. Unemployed people would be choosier about the jobs they accept, especially the low-wage ones. With more help for people after layoffs, employers and employees in struggling industries would do less to avoid layoffs, especially layoffs from low-paying positions. Either way the result would be more unemployed people.

Subsidies for unemployed people also make labor more expensive as low-wage jobs are more likely to end by layoff and unemployed people can be choosier about the jobs they take. When labor is more expensive, employers have an incentive to get by with fewer employees and for that reason may well reduce the number of job openings they have.

In this way a reduction in labor supply by itself, a reduction in labor demand by itself or both together can increase the ratio of unemployed to job openings. It makes little sense to point to a high ratio as proof that labor demand is low, because it could just as easily tell us that labor supply is low. All a high ratio tells us is that the labor market has contracted, and that we could readily and more reliably detect without any data on job openings by just looking at the unemployment rate itself, or the ratio of employed to population.

My conclusion is not new to labor economists, who have long understood that supply factors could increase the ratio of unemployed to job openings. Christopher A. Pissarides, a professor at the London School of Economics, literally wrote the book on job openings and unemployment, and his book explains how more generous unemployment compensation would have these effects (see Figure 9.2 from his latest edition; I thank my colleague Robert Shimer for this reference).

The black series in the chart below shows the ratio of unemployed to job openings. The chart also shows in red the marginal tax rate on labor income (the extra taxes paid, and subsidies forgone, as a result of working, expressed as a ratio to the income from working) for a typical head of household or spouse based on the ever-changing eligibility and benefit rules for safety-net programs. The ratio increases fastest between the first half of 2008 and the first half of 2009, just when the marginal tax rate series increases the most. Both series peak in late 2010 and decline thereafter. Neither series has returned to its prerecession level.

Ratio of unemployed per job opening is calculated from Bureau of Labor Statistics seasonally adjusted monthly figures for number of unemployed and total nonfarm job openings, as provided by the St. Louis Fed. Marginal tax rates are as calculated by Casey B. Mulligan in Ratio of unemployed per job opening is calculated from Bureau of Labor Statistics seasonally adjusted monthly figures for number of unemployed and total nonfarm job openings, as provided by the St. Louis Fed. Marginal tax rates are as calculated by Casey B. Mulligan in “The Redistribution Recession” (Oxford University Press, 2012).

For the reasons mentioned above, the chart is by no means proof that supply was a major factor during the recession. That proof requires other sorts of analyses, which are shown in my book.

Nevertheless Paul Krugman continues to cite the high ratio of unemployed to job openings as evidence that demand, rather than supply, contracted the labor market: “There are now four job seekers for every job opening, which means that workers who lose one job find it very hard to get another” (see Page 9 of “End This Depression Now!”). He and other economics commentators citing this fact never explain why the very same ratio should not be interpreted as a drop in supply, or as a combination of reduced supply and reduced demand. Instead they contend that the labor market would rebound with still more help for the unemployed.

Believe it or not, Keynesian economics is not the only way to interpret the job openings data.

Tuesday, November 13, 2012

An Old School Keynesian on Marginal Tax Rates

"Our present system of welfare payments does just that [with 100 percent taxes], causing needless waste and demoralization. This application of the means test is bad economics as well as bad sociology. It is almost as if our present programs of public assistance had been consciously contrived to perpetuate the conditions they are supposed to alleviate."
Tobin, James. “On Improving the Economic Status of the Negro,” Daedalus (Fall 1965), 94(4).

Another interesting quote on the marriage tax implicit in welfare programs:

All too often it is necessary for the father to leave his children so they can eat. It is bad enough to provide incentives for idleness but even worse to legislate incentives for desertion.
Undergrad readers of this blog might benefit from this:

Introducing the Hayek Fund for Future Scholars

Do you know great students or recent graduates applying to graduate school this year? IHS wants to help! We can increase their chances of being accepted and funded, by making it more affordable to apply to more schools.

The Hayek Fund for Future Scholars awards up to $300 for grad school application fees.
Please pass this along to the promising, liberty-oriented people you might know applying for full-time doctorate or master's degree programs in the US for the 2013-14 academic year. They can find out more at


Keri Anderson
Student Coordinator
Institute for Humane Studies

Saturday, November 10, 2012

My Neighbors Know

In my precinct, 79 percent of registered voters voted (I was one of the voters).  Of the voters in that precinct, only 31 percent voted for Congressman Jesse L. Jackson, Jr (I was not one of those). Nevertheless, thanks to the many other precincts, the Congressman got 70 percent of the total votes in his Congressional District shortly before heading for prison.

In case you were wondering, that same precinct cast 59 percent of its votes for Obama-Biden.

Friday, November 9, 2012

Blind Squirrel Finds No Food

It's a good sign that there are critics of The Redistribution Recession, and that so far every one of them admits that he hasn't read any of it. I'm trying to think about how I could encourage such behavior, because any citation is a good citation.

With enough attempted criticisms and a little luck, the law of large numbers predicts that non-reading critics will eventually put forward something that might be valid. That's what's impressive about this admitted non-reader's piece: it makes quite a few claims about the book and every single one of them is false! I'm sure because I HAVE read the book ... many times.

(please click through to take a look: it has a nice picture and page views are the reward I mentioned above)

Example: The book allegedly does not address specific criticisms (without reading a book, how do you know what's missing from it?), when in fact I anticipated these criticisms years ago and devote entire chapters of the book to them.  (Oxford owns the copyright, and I will not steal from them in order to further help non-readers by reprinting or paraphrasing those chapters here or elsewhere on the internet.)

I understand that reading takes time, so in order to help the blind squirrels find an acorn every once in a while, I have prepared a www page (with Oxford's permission) with a brief summary of the book, a brief Q & A about the book, and a short video presentation.

Another alternative, or prelude, to reading: take a look at the opinions of some people who have read the book, or are currently reading it:

"Rethinking one's views" is even more costly than regular reading, which is one more reason to talk, blog, or curse about the book without actually reading any of it.  Even opening to page one could prove to be expensive. 

Thursday, November 8, 2012

You too Suffolk?!

Suffolk County NY begins gas rationing.  Suffolk County has a lot to brag about, but their regulations are not  among them.

I have an idea: while Suffolk County official require boaters to take a boat-operator's course, the county should require their officials to take an economics course -- at the University of Chicago!

Wednesday, November 7, 2012

Gas Lines are Unnecessary

Copyright, The New York Times Company

When it comes to making the last week unpleasant, Hurricane Sandy got some help from government officials.

As the water from the storm began to recede, people in the New York metropolitan area wanted to repair, rebuild and get back to normal, But one of the most visible obstacles has been 1970s-style lines for gasoline. Many customers waited in line for hours only to learn that fuel had run out. Gasoline was rationed in New Jersey, where license plate numbers determined which days drivers were permitted to purchase fuel.

Waiting in line is a waste of time. The people there were certainly not helping bring more gasoline to the region and could instead be helping rebuild or could be productive in other ways.

Economists on the right and on the left agree that market prices – prices that reflect both supply and demand location by location – are much better at allocating scarce resources in extreme situations like the storm’s aftermath. But state and local government regulations, in the form of antigouging laws, effectively outlawed market pricing.

Early on, Steve Bellone, the executive of Suffolk County on Long Island, warned suppliers that he would punish anyone charging prices that were too high. Gov. Chris Christie of New Jersey sent similar messages. In New York City, federal officials interfered with the market by giving gasoline away; those lines were so long and contentious that New Jersey decided not to use that strategy.

If officials had allowed the price system to work, it would have alleviated lines in a number of ways. As suppliers seek the maximum profit, temporary and extraordinary prices encourage them (and make it affordable for them) to go to extraordinary lengths to get the electricity and fuel needed to have gasoline available to customers where it is needed the most.

Were they permitted, high prices would also have encouraged customers to economize creatively on their usage and acquisition of gasoline. If it had cost $10 or $15 a gallon, some people on those lines might have been willing to delay vehicle usage, leaving more for people who were willing to pay that price or who had no other choice.

Of course, many suppliers and customers take prudent steps because they want to be helpful during a time of emergency. But why not let the market bring forth more supply and more customer conservation by adding a financial reward?

Instead, officials resorted to begging customers to conserve – Gov. Andrew Cuomo of New York said, “Now is not the time to be using the car if you don’t need to” – and spending law enforcement resources dealing with hoarding, gouging and other crimes that would not exist if the price system had been allowed to work.

Mistakes of economics will happen sometimes, but it is too bad that government officials in the New York area are making so many when residents can least tolerate them.

Flashback: Don't Read This Until Nov 7

from last week:

If your candidate lost yesterday, I'm sorry. If you lament because you assumed that your candidate would have implemented superior public policies, then you can feel better already because your sorrow is based on a false assumption.

Democrats and Republicans clearly have different rhetoric. But rhetoric is not policy. Republicans talk a great game when it comes to cutting government spending, but President Clinton's administration had one of the lowest ratios of government spending to GDP. President Bush added immensely to Medicare spending with the Prescription Drug Act. Democrats talk a great game about helping the poor, but they pushed through a bill to tax America in order to bail out Wall Street. FDR started Social Security, but Nixon did the most to increase its spending. Democrats talk about limiting the power of the state when in comes to the death penalty, but a Republican Governor (Ryan in IL) put a moratorium on the death penalty.

Do you remember when Democrats were devasted because Roe-v-Wade would be overturned once President Reagan made his Supreme Court appointments? Well, those appointments happened and Roe-v-Wade still stands. I could go on and on with examples.

Economic theory suggests that political party might not affect policy, but instead merely reflect public policy preferences of the citizens. With some exceptions (see below), political parties compete with each other. Obama was one of the most liberal U.S. Senators because he faced little contest in Illinois, but became quite middle-of-the-road when it came to the Presidential race. Politicians are politicians first and (at best) ideologues second. A public opinion shift may give one party or another a small advantage and thus create a correlation between public policy and party-in-power, but this does not mean that political party itself has a significant impact on policy. Indeed, it would be inefficient if it did.

A number of economic studies have failed to find a correlation between party-in-power and public policy. Others have found a correlation (Professors Besley and Case have a nice survey in the JEL), but even there the implied impact is quite small. For example, Besley and Case look at state governments (where spending is about 1000 1982-dollars per capita per year) and find that governor's party is not correlated with spending and that a 10 percentage point increase in the Democratic party's share of the state legislature is associated with additional state government spending in the amount of $10 per capita per year. $10 per capita per year could be less than the cost of voting itself! Furthermore, effects at the state level may be larger than they would be at the national level because state-legislature elections are often uncontested and the whole economic logic cited above presumes competition.

Professors Snowberg, Wolfers, and Zitzewitz tried to look at situations in which party-in-power was significantly different even when citizen preferences were not. They found some effects, but they were also quite small. Eg., a Bush administration (rather than Kerry or Gore) was expected to increase stock prices by 2-3%. That is pretty trivial, given that the stock market fluctuated that much in the 20 minutes it took me to type this entry (back in October 2008).

Tuesday, November 6, 2012

Professor Quiggin Could Learn a Lot from My Book

I recently criticized Paul Krugman's recent book for ignoring marginal tax rates, which were hiked by the stimulus law and would be further hiked by the bigger stimulus that he proposes in his book. Moreover, I asserted that high marginal tax rates are responsible for a lot of the U.S. labor market's recent depression, and that Krugman's plan would have depressed it further.

Although not the author I was criticizing, Professor Quiggin recently wrote that

  1. I was obviously wrong because of what happened in other countries,
  2. The very recent (last 6 months or so), withdraw of some of the 99 weeks of UI benefits proves that marginal tax rates don't matter, and
  3. "As for food stamps, the expansion in the number of recipients is not due to changes in policy."

All of these points are addressed  in my book, before Professor Quiggin even wrote them.  Even if he had not read my book, a little investigation would have quickly shown him that his claims are incorrect. I take the points in reverse order.

3.  SNAP (aka, food stamps). Professor Quiggin has been repeatedly refuted by the US Department of Agriculture (it administers SNAP), most recently in its Sept 2011 report where it says "The continued growth in SNAP participation from 2009 to 2010 is likely attributable to the slow recovery from the recent economic recession, expansions in SNAP eligibility, and continued outreach efforts."  [emphasis added]  My book agrees that all three were a factor, provides estimates of their separate quantitative importance (Table 3.4), and discusses the academic literature on the subject.

One way to quickly see how Professor Quiggin is wrong about food stamps is to look at SNAP participation as a ratio to either (a) persons in poverty, (b) persons on Medicaid, (c) persons on SSI, or (d) persons on SSDI.  Of course a "bad economy" expands participation all of these things, but why would SNAP grow so much more than the others?  The answer is simple: SNAP policy changed, while the definition of poverty was constant and the policy rules for Medicaid, SSI, and SSDI were relatively constant.

2.  99 weeks no more.  As of October 2012, unemployed people could not collect 99 weeks UI, thanks to UI rule changes going into effect this spring.  But they still could collect 60+ weeks, not to mention remain on Medicaid, SNAP, and other programs indefinitely.  My book quantifies all of these factors, and finds that the difference between 60+ weeks and 99 weeks (actually, 96 weeks was the national average) is real but fairly small (extending UI from 26 to 52 weeks is a big deal).  So my model predicts that, adjusted for age and other factors, the labor market would rebound slightly during 2012, which is exactly what happened.  (There are also issues of timing here, which are discussed in my book).

3.  Austerity depresses the economy.  I agree (see also here) that European governments have typically failed to revive their economies, and probably further depressed them.  But austerity is not opposite of redistribution (ie, hiking marginal tax rates).  Think of how austerity might be implemented in the U.S.: we might cut Medicare and Social Security, but only for the more successful beneficiaries.  Regardless of whether redistribution is achieved by withholding benefits from families with high incomes, providing more subsidies to families with low incomes, or both, an essential consequence is the same: a reduction in the reward to activities and efforts that raise incomes.  Many kinds of austerity enhance redistribution, and that’s an important reason why austerity depresses the labor market.

With that said, I am very much in favor of cross-country comparisons.  I would love it if Professor Quiggin or anyone else measured marginal tax rate time series for any country that we could compare to my series for the U.S.  But Professor Quiggin has failed to do that, and instead  claims without evidence that marginal tax rates were constant (or falling) everywhere outside the U.S.

Finally, if marginal tax rates were found to be constant in Estonia (the only specific country that Professor Quiggin points to), does that mean that marginal tax rates do not matter in the U.S.?  Please let me know so I can notify American economists that Estonia is our ideal laboratory, and notify policymakers that they can safety hike marginal tax rates to 100 percent without noticeable consequences.

Added: I have heard the claim that the ratio of SNAP to Medicaid increased so sharply because Medicaid was cut sharply, rather that SNAP changing its rules.  The claim ignores the help that Medicaid got from ARRA and, more important, that Medicaid spending and participation per person in poverty was pretty flat.  Why don't the allegedly sharp Medicaid cuts result in sharply less Medicaid per person in poverty?  The answer is simple: Medicaid cuts, if any, where nowhere near the magnitude of SNAP expansions.  For more on SNAP expansions, see The Redistribution Recession.

Monday, November 5, 2012

WSJ reviews The Redistribution Recession!!

What if Keynesian economics is a bankrupt theory and the massive "stimulus" bill in 2009 made the economy worse, not better? Those are among the questions that Casey Mulligan asks in "The Redistribution Recession," a biting analysis of our current economic malaise.

Click here for Stephen Moore's favorable review

Saturday, November 3, 2012

Opportunity for Election Forecasting

I have absolutely no expertise in election forecasting. But I have enough faith in the stability of human behavior, properly understood, to believe that election forecasting experts (not me!) should be able to accurately and confidently predict who will be president next year.

This not to say that such a forecast would merely consist of taking the leader in a large national poll, or the leader of a large poll conducted in Ohio. You have economic data, rallies, campaign contributions, early voting patterns, billions of tweets and facebook chatter, and many other glimpses at what people might do on Tuesday. I would think an expert could process all this and tell us who will win.

Theory 1: The Experts do Know
One view of the state of forecasting is that the experts have done exactly that, and President Obama will win. One reason I hesitate to embrace this view is that the experts I have seen (not a good sample, because I am not an expert on experts, either!) have not considered all that much data (I don't consider repeats of flawed data to be much data, regardless of how many times it is repeated).

Nate Silver, for instance, gives the President an 84% change primarily on the basis of state-level polling data, and gives essentially zero weight to the national polling data (see the middle of this long post), let alone any of the other sorts of data I mentioned. He may have the right weights -- he's the expert -- but an explanation of his dramatic departure from Bayesian good practice (ie, throwing out important data sources rather than down-weighting them on the basis of assessed flaws) is conspicuously absent.

Justin Wolfers looks at who people expect to win (which also points to victory for the President). That's an interesting data source, and he explains why he puts less weight on the usual polls. But what about all of that other data?

Intrade, meanwhile, puts the President's chances at 2/3 (ie, 2-to-1 odds of losing). That's pretty close, and not the confidence I am expecting from a good forecaster. (for confidence, look at silver's odds, which exceed 5-to-1). Either intrade refutes the theory that the experts know who will win, or intrade has done a poor job of aggregating information.

If the President wins by a non-trivial margin (ie, by more than Ohio, or by a 5+ percentage point margin in Ohio), it will tough to reject Theory 1. But that outcome would also support the theory that the experts got lucky.

Theory 2: It Really is Too Close to Call

Some elections, such as Bush v Gore, are genuinely close and I don't expect the experts to know those outcomes ahead of time (or even the day after the election). But they should know that they don't know. Silver, and maybe not even Wolfers, are saying that they don't know. Other experts are calling it a tossup, but I am worried that they work for TV networks who want viewers.

If either candidate wins by a non-trivial margin, that casts a lot of doubt on Theory 2 unless we have good reason to believe that the world changed significantly in the last hours before the election.

Theory 3: The Experts aren't Doing a Very Good Job

If the President loses by a significant margin, Theory 3 has to be the favored theory, which is an opportunity for an ambitious election forecaster to get it right next time.

We are Supposed to Trust Them with Trillions

DeepenEnd This Depression Now! insists that we taxpayers should trust the author and other Keynesians with trillions of our dollars, and in return they supposedly would end this depression fully and quickly. Meanwhile, the best the author can do to earn our trust is to take a vacation from both logic and economic understanding:

  1. He and I agree that soup kitchens did not cause the Great Depression.  But only he can conclude from that fact that incentives should be ignored when studying the labor market, and that the best way out of this recession is to further broaden the population of people with no incentive to work.

  2. He and I agree that a massive safety net expansion all by itself would increase wages (he never says by how much, though ... my book looks extensively at that: it's just a couple of percent in the short run, and then falling back to trend).  Then, with the intention of showing that safety net expansions were trivial, he puts up a graph showing that wages continued to increase (sic) after the recession began!  Notice in particular the vertical axis in his chart (and ignore that he cherry-picks a disportionately manufacturing sample ... my book shows how manufacturing did experience a sharp demand reduction but that manufacturing is not the entire economy): the axis is measuring wage CHANGES and all of its numbers are positive.

  3. If he had wanted to test the theory a little more carefully, he might have looked at wage levels, and acknowledge safety net contractions as well as expansions.  That's what I did in my book, and in my "Why did wages rise and then fall?"  Here's what you get: real wages rose above trend when the safety net expanded, and did not start to fall until part of the "stimulus" started to expire.
  4. We can quibble about whether wages went up a couple of percentage points or went down a percentage point or two, but the real issue with wages is what happened to after-tax wages: they fell about 12 percent below trend because of the massive hikes in marginal tax rates. So even if you really did have a demand drop that by itself depressed pre- and after-tax wages by 3 percent, and a safety net expansion that by itself increased pre-tax wages by 2 percent and depressed after-tax wages by 9 percent, then the net result would be pre-tax wages falling by one percent -- Professor Krugman and friends could have their "gotcha" -- yet still after-tax wages fall by 12 percent, three-quarters of which is due to the safety net expansion. If your choice was to ignore the safety net and focus on demand or ignore demand and focus on the safety net, you would get a lot closer to the truth with the latter approach.
  5. when it comes to real wages, my model is Keynesian in the sense that it says that wages are counter cyclical (high when employment is low). It's kind of funny that, regardless of what the data show, Krugman would attempt to discredit the Keynesian part of my model by insisting that wages are procyclical. I guess he sides with Kydland and Prescott on wages and indeed he does appreciate nonKeynesian approaches :)