"Fed Up" With Paul Ryan

The following is an essay that I wrote for NextGen Journal, an intercollegiate journal focusing on the issues facing America's youth.
............................................................................................................................

I'm “Fed” up with Paul Ryan. No, I don't mean his regressive tax proposals, unrealistic budget projections, or his peculiar approach to health care. I am frustrated by something more serious: his call for monetary policy that would only put our economy in greater danger.

The Federal Reserve controls U.S. monetary policy by controlling the growth of the money supply in accordance with its dual mandates of “maximum employment” and “price stability.” Traditionally this has meant unemployment below 5% and inflation around 2%. According to Paul Ryan, recent Fed actions have caused massive inflation and have debased the dollar. Therefore, the Fed should abandon the “maximum employment” part of its mandate and focus on targeting headline inflation. However, a look at the data suggests this would be the wrong way to go.

Ever since July of 2008, annual inflation has been at about 1.1%, far below the 2% target traditionally set by the Fed. Long-run expectations of future inflation are also below target, at 1.26%. Moreover, the dollar is actually stronger now than it was at the beginning of 2008. The value of the dollar last peaked at the height of the financial crisis, showing that, in recent times, a strong dollar is not a sign that markets are doing well, but rather that they are breaking down.

So if inflation is low, and the dollar is strong, why has the Great Recession been so great? The biggest reason is that nominal gross domestic product (NGDP), a measure of the total dollar value of goods produced by the U.S. Economy, has collapsed. During the worst quarter of the crisis, NGDP fell at an annualized rate of 8.4% and has not returned to the previous trend, something unheard of for any other recession in the past century. The fall in NGDP has harmed the economy in many ways, most importantly by making it harder for families to meet mortgage payments and for business to justify further expansion.

To change this, the Federal Reserve should announce a policy of restoring NGDP to its pre-crisis 5% trend, and take all necessary steps to reach the goal. This policy enjoys a wide range of support from conservative and liberal economists alike. Changing the Fed's mandate to targeting the level of NGDP would change market expectations and work to stimulate growth now.

Representative Ryan might argue that this shift would be just another form of discretionary monetary policy that jeopardizes the stability of the economy. In his 2010 op-ed with John Taylor, Ryan demanded that the new monetary regime have “greater simplicity; a description of interest-rate responses to economic developments including how the Fed will achieve those responses through money growth; and greater attention to commodity prices, including food and energy, as opposed to a myopic overemphasis on core inflation.” Fortunately, NGDP targeting addresses the root of those concerns just as well, if not better, than Ryan's version of inflation targeting.

First, unlike the current vague balance between “maximum employment” and “price stability”, an NGDP mandate only targets NGDP, making it a simple rule based regime. On the other hand, inflation is not so simple. To properly measure it, you need to pick which prices to check, and then adjust for quality increases. If a car's quality and price both improve, how do you know the real value? Moreover, which inflation – CPI, PCE, or PPI – should the Fed target? This confusion only complicates matters and increases policy uncertainty.

Second, a simple mandate translates into self-evident money supply responses to economic conditions. If NGDP growth is above the 5% trend, the money supply should contract. If NGDP growth is, as in the current situation, below the 5% trend, the money supply should expand. I choose not to talk about interest rates because, as Milton Friedman once wrote, they are misleading guides to monetary policy. Low interest rates can be the result of easy money that pushes down the market interest rate, or the result of tight money that leaves nobody wanting to borrow.

Third, an NGDP target shifts away from a myopic focus on core inflation and can effectively deal with asset bubbles. Even if focusing on core inflation led the Fed to hold interest rates “too low for too long” during the housing bubble, the inclusion of commodity prices in inflation indicies doesn't solve the issue. During the 90's and the early 2000's, the computer revolution increased productivity growth, driving down all measures of inflation, including those that took commodity prices into account. In response, the Fed eased monetary conditions and lowered interest rates to hit its inflation target. On the other hand, NGDP targeting would have actually tightened money in response to the 7% NGDP growth, thus popping the bubble before it grew large enough to hurt the broader economy.

Yet in spite of all this, Ryan wants to limit the mandate of the Fed to only inflation. On one issue, I agree with Paul Ryan, “we are on an unsustainable path” and “it doesn't have to be this way.” But to embark on a new path, we need an NGDP target, not inflation mongering. The first is an enlightened path forward, the second is but a dangerous step back that compounds our economic stagnation.

Eyes on the Prize

A look to the final frontier and then back to energy

Picture from The Guardian

While evidence in favor of global warming piles up by the day, there seems little political motivation to try to address the issue. Both cap and trade and a carbon tax would be politically impossible in such a low growth environment, and subsidies for more clean energy research would be shouted down with cries of "Solyndra!" Moreover, these policies are likely to have little substantive effect either. Without an international governing body, trade would nullify cap and trade or carbon taxes, and research subsidies tend to increase the price of research without raising the quantity. With this in mind, Evan proposes an alternative funding mechanism: research prizes. Instead of having the government subsidize firms or directly do basic research, the Federal government could sponsor research competitions, allowing innovators from all over the world to pool their collective wisdom to solve the energy crisis. To get a better idea of this would mean, we should look at a current example in which prizes played a large role: NASA's development of private, low earth orbit vehicles.

First, some history. What is often left out of the public memory of the push into space is what happened after we made it to the moon. Optimism about U.S. space policy led people to make "conservative" predictions of Mars landings by 1988, with a Mars base by the end of the century. Pan-Am even started taking reservations for flights to the moon, as it was forecasted that trips to the moon would be a quickly realized affair. So what happened? As the geopolitical impetus for the space program started to fade, so did the funding. Instead of going back to the Moon and beyond, we settled with the Space Shuttle program, which functioned as a space taxi that would ferry astronauts from Earth to Low Earth Orbit (LEO) and the International Space Station.

In theory, the Space Shuttle was supposed to be a temporary affair, replaced by a vehicle capable of going to the moon and beyond. However, cost overruns and budget problems eventually ended the program, resulting in a problem: after the retirement of the Space Shuttle in 2011, there was no way for U.S. astronauts to get back to space. Instead, we had to buy seats on Russian Soyuz shuttles at a price of $51 million, round-trip. 

Enter Commercial Orbital Transportation Services (COTS), a program designed to spur private sector solutions to the space transport program. It was implemented through special arrangements known as Space Act Agreements (SAA) coordinating development between NASA and private sector firms. Firms submitted proposals, NASA gave initial approval, and as the firms met certain milestones NASA gave them more funding. The funding was capped though, so no matter how much the companies spent, NASA would not pay them extra. This forced the competing firms to cut costs and streamline projects. NASA promised to purchase resupply and crew transport services from the final successful companies.

The program has been a resounding success. About 26 companies submitted proposals for the first stage, CCdev-1. Another 11 propsals were submitted for the second stage, CCdev-2. Space-X, the private space company that docked with the ISS just a few months ago, was actually denied funding for CCdev-1, but was later granted funded for CCdev-2. This showcases the resiliency of the SAA funding structure. Even though SpaceX was passed over, it still pulled through and is now the front runner in the COTS race. There is even discussion on how SpaceX's Dragon and Dragon Heavy rockets may make a trip to Mars by 2017, nearly 15 years before the same projection for the NASA planned Space Launch Vehicle at less than one hundredth of the cost. Of course some of it is hype and overoptimism, but it nonetheless stands testament to how far SpaceX and commercial space ventures have come since the era of Apollo.

The development of commercial crew highlights a few lessons about prizes and innovation. First, prizes often save money. There are stories of Elon Musk, the founder of SpaceX, being so unsatisfied with the market price for a certain injector that goes into the Dragon rocket that he decides to build it in-house at less than half the price. This is a particular issues as politicians like to see results for their billions of dollars spent. Lower costs equal happier politicians, which is a plus for a research program.

Well, except if the politicians are hungry for the pork offered by the traditional Space program. In the commercial transition, Senator Hutchinson of Texas has been a notable offender, fighting hard for the traditional government programs such as the Space Launch System as their budgets are cut in favor of commercial crew. This is the second lesson on why prizes are useful; they guard against rent-seeking, as if a firm meets the requirements, they are eligible for funds. No backdoor deals are needed.

A third lesson is that when you give the private sector the chance to directly work with the technology in the hope of generating disruptive innovations, they are in a better position to develop the technology into further disruptive innovations. In the vocabulary of growth economics, prizes promote "learning-by-doing", and have domino effects in promoting further development down the line. The fact that many firms are all trying their own ideas means that the market learns at a much faster rate once the knowledge gained from the innovations diffuses outwards. You have decentralized tinkering, instead of a top-down solution, vastly increasing the probability that someone thinks of the golden idea.

So does this mean this kind of fixed-cost research prize system is always superior to government sponsored direct subsidies and research funds? If anything, just the opposite. SpaceX, United Launch Alliance, Sierra Nevada, and all the other commercial space companies would not have gotten off the ground without the initial investment from NASA. Imagine Kennedy delcaring "We choose to pay the private sector to bring us to the moon in this decade and do the other things, not because they are easy, but because they are hard" There would have been no gravitas, no national pride, and importantly, no technology. At that point, the technology had not been developed. Somebody needed to go do the basic science to get us up there. 

To give an example of the wide range of science involved, think about the "simple" task of linking two orbiting objects together. It requires a firm understanding of the science of orbital rendezvous to build the correct equipment and to pilot correctly. Buzz Aldrin actually wrote a dissertation on this issue, and although it seemed useless at the time it was critical in the development of the space program. Buzz also spent time solving other physics problems such as the differential effects of gravity on large objects in space and the implications for navigation. If that science seems like something the private sector would be willing to fund, recall that the initial analyses of rocketry were conducted by a German Nazi* that was working on military rockets. None of this was easy; this was part of the reason why there were so many failures. Without the military impetus and government support, there would have been no space program. If the SpaceX Falcon rockets are able to fly so far now, it is because they launch off of the shoulders of giants. This is true figuratively as well as literally. The Falcon rockets launch from Cape Canaveral, the former launch site of the rocket that took us to the moon: the Saturn V.

In short, government directed research and prizes are complements. First, government R+D and subsidies are best for discovering fundamental disruptive technologies and sciences, whereas prizes are an effective way to further organize and commercialize that existing knowledge. Without DARPA, there would have been no Internet. But without subsequent innovations from companies such as Google or Facebook, the Internet would not be as vibrant as it is today. Second, it is important the government is to be a consumer, or an anchor tenant, of the innovation. For commercial space, NASA has committed to buy launches from whichever company that ends up developing the rocket. For energy, this would mean a guarantee from the government to purchase the electricity or fuel cells produced by a revolutionary firm. Third, private sector innovations from prizes can be used by the government. Government investments that fed into the private sector may feed back to the government again. NASA may end up using the improved rockets from COTS instead of the current United Launch Alliance Atlas V rocket for future Mars missions. The military may extensively deploy improved private sector solar panels to increase readiness in times of energy price volatility. So by all means, use prizes to spur innovation. But don't neglect the foundational role of the government in other dimensions.

*The rocket scientist in question was Wernher von Braun, and Tom Leher once wrote the following poem making fun of his political history:

Once the rockets are up,
Who cares where they come down?
'That's not my department',
Says Wernher von Braun.

Food (Price Shocks) For Thought

Global food prices and inflatable BRICs

While Shanghai recovers from the aftershocks of the Haikui typhoon, many other areas in the world are dealing with record droughts and rising food prices. The bad weather has hit U.S. farmers hard, with corn futures last week surging 59% from mid-June and soybeans jumping 21%. This may soon have international spillovers as U.S. crops count for more than half of the export market in corn and soybeans, both important inputs for the food industry, especially meat markets.

Although the recent food price spike has been significant, global food reserves and good harvests in other crops will likely prevent it from causing mass starvation. Nonetheless, food inflation is now putting the heat on global central banks as they consider whether they need to tighten monetary policy to maintain inflation credibility, or whether they should stick to maintaining short term growth instead. The BRIC countries stand on a dangerous precipice, as their real growth in recent months has slowed dramatically. The Brazilian Central bank is dealing with inflation on its doorsteps as its own employees are striking and demanding a 23% wage hike to compensate for higher cost of living. India's growth engine is also losing its spark and global food prices coupled with an already poor monsoon season could push inflation further beyond the central bank's comfort levels.  China is barely holding on, and monetary tightening at this juncture would have serious implications for both broad growth and the stability of the shadow banking sector. Russia is dealing with its own drought and its central bank is also under pressure from IMF officials calling for a monetary tightening. These food inflation problems are compounded by a rise in the value of the dollar, making purchases of U.S. corn, whose futures are 6% more expensive than their 2008 peak, even more costly.

No doubt, the current situation is quite severe, but what can history tell us about how food price affect inflation in the BRIC countries? Econometric evidence suggests that world food prices are a key driver, more so than oil, of global inflation, but can we generalize to the BRIC countries in the current situation? The first thing to note is that global food prices, as measured by the IMF food price index, have been on a secular rise since 2000, but that in June, the last measured month, food prices were still below where they were during the 2008 or 2011 food price crises.

Given that average food expenditure as a percentage of income for Brazil, Russia, India, and China all hover around 25 to 35%, we should expect that increases in food price growth should quickly show up in each country's inflation rates . However, by looking at the time series for each CPI and the IMF food price index, we see that the time series do not match up well and that the real story is a bit more complex. In each graph, CPI year over year growth rates are plotted on the left axis, while food index year over year growth rates are plotted on the right axis.

I split the countries in the above two groups for more than aesthetic reasons. If you look carefully at the time series, you can see that in the first group, China and Russia, food price growth and inflation rates seem to move together at all levels of food price growth. Over the entire period, China's inflation rate and food price growth had a correlation value of 0.7, which is enough at the 99% confidence level. Russia's correlation is more limited, as inflation only starts to move in sync with food prices after 2007. But in the period of time since 2007, the correlation value is 0.18, which is enough at about the 90% confidence level. I call this the unconditional inflation group, as the correlation between food prices and inflation is not conditional on the rate of food price growth.

On the other hand, if you look at the second graph, there's less of a discernible pattern for India or Brazil. Food prices spike in 2004 and 2008, but neither of the magnitudes of the two countries' change in inflation match the large swing in food price. However, the time series do start to line up in times of crisis, such as in 2009. This is especially evident for India, as from 2009 on, its inflation rate seemed to move in tandem with the food price growth rate. I call this group the conditional inflation group, as the correlation between food prices and inflation seems to be conditional on whether food price growth is sufficiently high.

To test this hypothesis, we can generate 2-year backwards looking rolling correlations and see how they evolve through time. These price correlations are plotted below, with the value of the 2-year rolling price correlation plotted on the left axis and year over year change in the IMF food index on the right.

China and Russia:

India:

Brazil:

In these graphs, we see the difference between the groups in a different light. The value of the food correlation for China and Russia seem quite uncorrelated with food prices, whereas for India and Brazil the correlation between food prices and inflation is higher when food prices are higher. With further analysis, it can be shown that we can reject the null hypothesis that food prices don't affect the value of the food correlation for Brazil and India, but we fail to reject the same null hypothesis for China and Russia. This is the reason why Brazil and India are grouped together as conditional inflation countries. Food price changes affect their inflation rate only if food prices are growing quickly enough. On the other hand, Russia and China are unconditional inflation countries, as food prices strongly affect their inflation rates at all levels of food price growth. The scatter plots of correlation versus food price growth for India and Russia are particularly illustrative of this difference. First, India:

Second, Russia:

While India's food correlation values look to be affected by food price growth, Russia's food correlations seem to just cluster horizontally around values of y=-0.75 and y=0.5. With more detailed regression analysis of India's results, we obtain a 95% confidence interval of (-0.25, -0.05) for the intercept and a 95% confidence interval of (0.0053, 0.0156) for the slope. A similar regression for Brazil returns a 95% confidence interval of (-0.35, -0.18) for the intercept and a 95% confidence interval of (0.0022, 0.0141) for the slope. Both these numbers suggest that the effect is real: higher food price inflation is associated with a tighter positive relationship between food prices and inflation. Food prices are a convex predictor: little effect when prices are low, much stronger effect when they are high.

What implications does this have for food inflation and the BRIC countries? First, we should expect China's and Russia's inflation rates to be hit the hardest by any food price growth. They unconditionally inflate, which means that the historical relationships suggest that a rise in food prices will directly translate into higher inflation rates for those two countries. On the other hand, India and Brazil only conditionally inflate. Statistically significant relationships are unlikely to form at current food price growth levels, and we need to be looking at at least 10% year over year growth in food prices before we should expect each country's inflation to becomes statistically linked to global food prices. Therefore, their inflation rates will likely only rise after China and Russia's inflation rates rise. However, this analysis does not say anything about welfare costs to these countries. Given that India and Brazil have higher inflation rates than China or Russia, convex costs to inflation may end up leading to more damage in the conditional inflators than in the conditional inflators. Nonetheless, it shows that the relationship between food prices and broad inflation is not so clear cut, and that some statistical manipulation can be invaluable in teasing out the connection.

NGDP Autoregressions and the Lucas Critique

Is NGDP growth sticky? In other words, does above NGDP growth in one period affect GDP growth in the next? Evan Soltas has previously shown that RGDP appears to be sticky. He constructed some impulse response functions and found that there is no instantaneous self-correction mechanism. On the nominal side, there is substantial evidence indicating that inflation is sticky. The correlation coefficient for the relationship between the current inflation rate and the inflation rate measured one quarter ago is 0.75. Another way of saying this is that about 50% of the variability in current inflation can be predicted by inflation one quarter ago.

Does NGDP, the sum of inflation and RGDP, suffer from the same stickiness? If it does, this could have serious implications for NGDP level targeting. If it takes a long time for past NGDP surges to slow down, this could affect the speed at which central banks can change expectations of NGDP. central banks may need to take even more drastic action to adjust NGDP at the necessary speed, causing monetary policy to be blunt and not credible.

To answer this, I looked at the NGDP time series from 1947 to today and constructed a multiple regression model to explain the current NGDP continuously compounded annual rate of growth as a function of the NGDP growth rate for the past six quarters. As only the coefficients for the past two quarters were statistically significant at the 95% confidence level, I settled with testing NGDP as an AR(2) model. The results of my regression are listed below:

Call:

lm(formula = n[, 1] ~ n[, 2] + n[, 3])

Residuals:

     Min       1Q   Median       3Q      Max 

-14.5968  -2.1681  -0.1866   2.0848  13.5262 

Coefficients:

                 Estimate  Std. Error t value   Pr(>|t|)    

(Intercept)  2.82257    0.47431   5.951     8.93e-09 ***

n[, 2]         0.43320    0.06257   6.923     3.67e-11 ***

n[, 3]         0.13255    0.06250   2.121     0.0349 *  

---

Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 

Residual standard error: 3.943 on 251 degrees of freedom

Multiple R-squared: 0.2631,     Adjusted R-squared: 0.2572 

F-statistic:  44.8 on 2 and 251 DF,  p-value: < 2.2e-16 

From this, we can say that about 26% of the variability in current NGDP growth is explained by past NGDP growth. To get a better idea of what the relationship looked like, I also plotted the predicted values of NGDP versus the actual values:

While the intercept is not statistically significant different from zero, the slope is 1, with standard error of 0.10, suggesting that the model does do a reasonable job of estimating actual NGDP. So does this data prove NGDP is sticky, making instantaneous NGDP expectation adjustment impossible?

As with most economic questions, the answer is "not necessarily". Perhaps NGDP is sticky because of certain nominal frictions in the economy, such as staggered wage contracts or menu costs. These traits, while they endow nominal shocks with real effects, also mean that nominal levels in the economy have momentum. If wages are suppressed for extended periods of time, the economy would be able to stay at higher levels of aggregate activity for longer. Alternatively, information itself can be sticky. So even if economic agents stand at the ready to change prices and to be flexible, they don't observe economic conditions quickly enough, acting in a way that creates nominal momentum.

A first pass analysis would suggest that these frictions would fundamentally limit the ability of NGDPLT to achieve stability. If elevated current NGDP always predicts elevated future NGDP, then there would be no way for a central bank to credibly commit to quick NGDP corrections.

However, given some recent discussion of Milton Friedman's thermostat, we have to ask if this result is regime dependent. Take inflation for a concrete example. The Federal Reserve has, for most of its recent history, targeted the rate, not the level of inflation. What this means is that if the Fed overshoots one year, there's no expectation for the Fed to compensate with lower than normal inflation for the next year. There's no expectation for monetary policy to correct the elevated inflation, which allows all the frictions mentioned above to keep inflation persistent. But if there's an expectation that the Fed will engage in corrective policy as in level targeting, then inflation may not be as persistent. Higher inflation today would actually predict lower inflation tomorrow as the central bank quickly acts to restore the original price level trend. Applied to NGDP targeting, if people perceived that current NGDP growth was higher, they would have an expectation that future NGDP growth would be slower. They could then act on that expectation and quickly restore trend NGDP. We could also test this hypothesis by testing if inflation or NGDP autocorrelations are higher in rate targeting countries than level targeting countries. However, I do not know of any central banks that have a formal committment to any kind of level targeting, so I'm not sure how robust my results would be.

This discussion of autoregressions and NGDP stickiness is also an instance where the Lucas critique defends the effectiveness of monetary policy. While I can run a multiple regression and find "evidence" that NGDP growth is sticky, because my arguments are not microfounded I have no theoretical reason for why NGDP growth would stay sticky in a level targeting regime. So to fully understand how nominal persistence can affect NGDP targeting, we need a microfounded model that can analyze the intertrelated process of nominal frictions, policy and expectation formation: a hole that market monetarists must be able to fill.

Labor Shortage, Misallocation, and Abuse: Big Mac Edition

Let's start with an interesting photo from a McDonalds restaurant:

Similar to my past observations on KFC, this Shanghai McDonalds seems really eager to hire. I'm always amazed by how happy and cute the pictured workers always are. Happiness is also the theme of the banner message: Join McDonalds and take the first step towards happiness. The message is particularly clever because it capitalizes on how the character that makes the "Mc" sound has the same pronunciation as the character 迈, which means to step.

McDonalds is also a great example of Balassa-Samuelson effect. Even if aggregate productivity grows much faster in China  than in the United States, it's unlikely that a Chinese McDonalds franchise can raise its productivity any faster than an American McDonalds franchise. Just look at the registers; how do you expect the cashiers to ring people up any faster?

The low productivity growth can ultimately be traced to one factor: there's not much space for innovation within McDonalds. No doubt, running a McDonalds franchise is no easy task, but for the individual cashiers, cooks, and janitors, there's only so much you can "learn-by-doing". Compare this with other sectors such as, solar panel production or apparel, and you can see why there's a large productivity differential between fast food and manufacturing.

These two factors, aggressive hiring and low productivity growth, are connected; the second can lead to the first. This is not necessarily only because low productivity growth requires a higher amount of labor to produce the same amount of product, but rather also because low productivity enables employers to abuse an initial probationary period to hire more aggressively with minimal downside. During probation, employees can be fired without cause and are paid lower wages. Limited room for innovation means there's little need to train employees to create new techniques, thereby shortening training times. Since  training time for an average employee is relatively low, McDonalds can afford more worker turnover without significantly impacting productivity. But to fill this demand for workers, they need to aggressively hire. This is arguably a violation of Chinese labor laws, but given McDonald's other violations, it should not a surpise. Chinese labor laws such as minimum wage and worker's insurance have also been frequenly abused by Yum Brands, the parent company of KFC, Pizza Hut, and Taco bell.

This story is problematic because it contradicts the idea that there's a labor shortage. If workers are hard to find, how can McDonalds and other fast food restaurants afford such high turnover? A possible explanation is that they pay high enough of a wage to entice workers from other restaurants, and therefore exacerbate the worker shortage for other businesses. Yet because they do not advertise as aggresively, I do not notice.  

Another possibility is that the target demographic, college students, is not the same demographic of workers that coastal plants are lacking. College students often won't or can't hold low-skill manual manufacturing jobs, but they may be willing to take up a job at a McDonalds while they go to school. College workers also reside in a grey area of labor law. Guangdong marketing director of Yum Brands has been on record saying "part-time workers are neither full-time workers or non-full-time workers (既不属于全日制用工也不属于非全日制用工)" to justify the low pay of part-time college student employees. Ignoring whether this is ethical, if college students are advantageous from a labor/wage perspective, this would match both the type of advertising we see on both the KFC and McDonalds hiring notices with the concept of a skill mismatch. The firms would have an incentive to hire college students to save on wages and insurance fees, not to capitalize on their unique skills.

The story from here would be that China does not suffer from just a lack of workers, but rather that there's a severe skill mismatch. College students, who should be building human capital, are being cycled through low-skill, low productivity jobs. This pushes the unskilled to even lower productivity jobs such as street sweeping or selling road-side trinkets. Yet while all this happens, coastal firms struggle to find enough talented workers to stay open. This, of course, does describe the Chinese labor market in much depth, but it does point to a key structural problem holding back further Chinese growth.

How to Unscrew a Generation

A Newsweek article by Joel Kotkin highlights why monetary policy is so important

Newsweek asks: are the Millennials really "screwed"? While I can't speak for fellow undergraduate blogger Evan, I would like to declare that I would strongly prefer that the economy not turn me into an inclined plane wrapped helically around an axis. And for this reason, as a member of the Millennials, my strongest demand is that monetary policy starts to effectively target nominal GDP.

This may surprise some readers who believe that structural factors are the reason holding back my generation. That would be an understandable first impression from the Newsweek article. But remember that the housing bubble popped in 2006 and the proximate cause of the recent recession was actually monetary policy failure. So instead of seeing the article as a list of damning reasons against taking action on monetary policy, I saw it instead as a list of damning reasons for taking further action, and why a failure to do so will have serious long run implications for growth.

Let us start with labor markets. Kotkin claims that younger workers are being hurt by the longevity of their parents who aren't leaving the workplace:

One key reason: their indebted parents are not leaving their jobs, forcing younger people to put careers on hold. Since 2008 the percentage of the workforce under 25 has dropped 13.2 percent, according to the Bureau of Labor Statistics, while that of people over 55 has risen by 7.6 percent.

"Employers are often replacing entry-level positions meant for graduates with people who have more experience because the pool of applicants is so much larger. Basically when unemployment goes up, it disenfranchises the younger generation because they are the least qualified," observes Kyle Storms, a recent graduate from Chapman University in California.

But this argument is missing something. If medical advances raised the number of years people could work, most economists would call this an aggregate supply increase, not a reason why younger workers should suffer. If anything, it should at least increase aggregate growth. Yet the article suggests just the opposite. This is because there is a bigger problem: a lack of aggregate demand. There's plenty of slack in labor markets with unemployed graduates, we just need some more nominal GDP to take advantage of it.

The above passage also highlights another perspective on hysteresis and why aggregate demand shortfalls have serious long-run implications. Besides causing the formerly employed to lose their skills, unemployment also causes the not-yet-employed to miss the chance to gain new skills. Those outsiders become "disenfranchised", and never have the chance to learn. This is especially problematic if human capital, like regular capital, exhibits diminishing marginal returns. What this means is that the additional human capital gained by current workers is less important for long-run growth than the initial capital gained by newer workers. The first bit that new workers learn has a higher impact than just marginally increasing the human capital of existing workers. So when people talk about "labor-mismatch" and a shortfall in human capital, remind them that aggregate demand is critical in developing and organizing the skills of new generations of workers, thereby fostering long run growth.

Another reason unemployment devastates human capital is because it forces highly educated individuals into jobs typically reserved for lower-skilled workers:

More maddening still, the payback for this expensive education appears to be a chimera. Over 43 percent of recent graduates now working, according to a recent report by the Heldrich Center for Workforce Development, are at jobs that don't require a college education. Some 16 percent of bartenders and almost the same percentage of parking attendants, notes Ohio State economics professor Richard Vedder, earned a bachelor's degree or higher.

"I work at the Gap and Pacific Pak Ice, two jobs that I don't see myself working long term nor jobs that are specific to my major," notes recent University of Washington graduate Marshel L. Renz. "I've been applying to five jobs a week and have gotten nothing but rejections."

This then spills over to actual low-skilled workers, who find themselves competing against overeducated peers in a ever-worsening trend. College graduates like Renz work in jobs that neither let them invest in further human capital or take advantage of their current human capital. While they get trapped, high-school graduates and dropouts fall even farther behind in the human capital race, resulting in more social problems on their way down.

The worst effects of the "new normal" can be seen among noncollege graduates. Conservative analysts such as Charles Murray point out the deterioration of family life—as measured by illegitimacy and low marriage rates—among working-class whites; among white American women with only a high-school education, 44 percent of births are out of wedlock, up from 6 percent in 1970. With incomes dropping and higher unemployment, Murray predicts the emergence of a growing "white underclass" in the coming decade.

In many ways, this white underclass has already developed. As Evan noted a few months ago:

To summarize the findings, Ip's missing 5 million are disproportionately young (ages 16-19), male, nonwhite, and of low educational attainment as compared to their respective fractions in the American labor force.
...By sex, 3.6 million men and 2.3 million women are missing from the labor force. Given that men compose 53 percent of the labor force, men are 15 percent overrepresented relative to women. Interestingly, the ratios of under- and over-representation between men and women has remained roughly constant as the number missing has grown -- and this is true for most of the other demographic breakdowns.By age, 1.3 million Americans between the age of 16 and 19 inclusive and 5.0 million above the age of 20 are missing from the labor force. Given that those between 16 and 19 compose 3.7 percent of the labor force, the young are 7 times overrepresented relative to the working-age. By ethnicity, 4.6 million white Americans, 0.8 million black Americans, 0.8 Hispanic Americans, and 0.6 million Asian Americans are missing from the labor force. Given that these ethnicities compose 71, 10, 14, and 5 percent of the labor force respectively, white Americans are 5 percent underrepresented, black Americans are 14 percent overrepresented, Hispanic Americans are 18 percent underrepresented, and Asian Americans are 90 percent overrepresented.FRED doesn't seem to have population data on the number of college and high-school degree-holders, or those with some college, but just by comparing the declines in their labor force participation ratios, it is obvious that high-school graduates are significantly overrepresented in the missing millions relative to college graduates and those with some college.

Recent micro-level analyses strongly suggest that contractionary monetary policy systematically increases inequality, adding to the underclass. This further destroys any possibility of fostering human capital. Given that the first few years of human development are enormously important to a child's later success, how does one expect healthy future generations if they don't have a proper environment in which to grow? Yet, this lack of human capital might not be quite as important as the lack in, well, humans:

Inevitably, young people are delaying their leap into adulthood. Nearly a third of people between 18 and 34 have put off marriage or having a baby due to the recession, and a quarter have moved back to their parents' homes, according to a Pew study. These decisions have helped cut the birthrate by 11 percent by 2011, while the marriage rate slumped 6.8 percent. The baby-boom echo generation could propel historically fecund America toward the kind of demographic disaster already evident in parts of Europe and Japan.

Instead of leveraging higher population growth to support the large stock of pension liabilities, we're getting slower population growth because of poor monetary policy. At least we can rely on more immigrants, right? Not if the current republicans have their way:

Right now, politics is just another place where American millennials are getting screwed. Republicans want to deport young Latinos while cutting investments, such as roads and skills education, that would benefit younger voters. Democrats, meanwhile, seem determined to mortgage the future with high spending on pensions, predominantly for aging boomers; cascading indebtedness; and economic policies unfriendly to the rapid growth necessary to assure upward mobility for the new generation.

If you look carefully at that paragraph, you can see a whole host of reasons why stable aggregate demand would improve aggregate supply policy.  First, good monetary policy can help with the investment problem, as it can increase R+D spending by helping risk-averse firms plan for the future. This would partially counteract the drop in public investment that has been the result of entitlement spending.

Second, and potentially more importantly*, stable aggregate demand could change the perception of immigrants from "job stealers" to productive workers. The problem with immigration is that if the monetary authority fails to stabilize aggregate demand, an increase in the labor force may crowd out existing workers from their jobs. But because immigration expands aggregate supply and increases total factor productivity growth, it actually raises the nation-wide standard of living.  As a result, stable aggregate demand would greatly alleviate the political opposition to low-skilled immigration and raise the possibility of reform. So libertarians who are in favor of liberalized immigration should also be scrambling for stable aggregate demand. Otherwise, the politics of immigration will never allow for real action.

Immigration is especially important because it's one of the low-hanging fruits that Lady Liberty still has within reach. The Newsweek article tries to use Tyler Cowen's "The Great Stagnation" to provide evidence that the problems we face are all structural, but immigration is a fruit that still hangs low. Yet as a society, we refuse to eat.

Allowing these increased flows of legal immigration may also help solve the political problem highlighted in the Newsweek paragraph. The problem with entitlement spending is that there is an increasing number of older voters, causing entitlement reform to become "a thoroughly rigged boomer game, providing guaranteed generous benefits to older public workers while handing the financial upper echelon a 'Wall Street boondoggle'" More immigrants would lower the average age of the population, forcing the parties to cater more to the needs of young people. Also, better monetary policy, by reducing inequality, would empower immigrants and other working class young individuals to have a stronger voice in politics, giving us a better shot at addressing the serious issues of public finance that we face today.

So while, yes, we do face serious structural problems, we can't hope to truly address them without an adequate monetary policy response. Stabilize aggregate demand, and the dust will settle. Then we can know what we truly need to do. Otherwise we may have to settle for a much more stagnant future:

Once known for their optimism, many millennials are turning sour about the future. According to a Rutgers study, 56 percent of recent high-school graduates feel they would not be financially more successful than their parents; only 14 percent thought they'd do better. College education doesn't seem to make a difference: 58 percent of recent graduates feel they won't do as well as the previous generation. Only 16 percent thought they'd do better.

This perception builds on the growing notion among economists that the new generation must lower its expectations. Since the financial panic of 2008, "the new normal" has become conventional wisdom. Coined by Mohamed El-Erian at Pimco, it's been used to describe our world as one "of muted Western growth, high unemployment and relatively orderly delevering."

We cannot afford to lower our expectations and ignore how monetary policy has failed an entire generation. So Ben Bernanke, if you dare not act for the sake of your mandate or for the sake of your reputation, act for the sake of a generation. For the high school graduates worrying about the value of a college degree, for the college graduates struggling to make ends meet, for those everyday workers struggling to buy groceries, and for your own two children and all of their peers, for their sake, I beg that you act.

*Full Disclosure: I am a 1.5 generation immigrant. My parents are 1st generation immigrants.

Inaccurate Estimation in a Gaussian World

How accurate is estimation in a Gaussian world?

Critics of finance lavish lots of attention on "fat-tail distributions" and how they call into question the way finance deals with low probability events. Nicholas Nassim Taleb is particularly angered by the way financiers use the Gaussian bell curve, affectionately known as the Great Intellectual Fraud, to "predict" and optimally hedge bets. While sympathetic to this argument, I still find the Gaussian bell curve to be a useful tool to help demonstrate how fragile and unpredictable low probability events are even in a normal world. Working on the problem also proves to be a convenient time to orient myself in R, a statistical package that I will likely be using in my undergraduate research at the University of Michigan this fall.

So here's the problem I want to look at:

Given a sample from a normally distributed population
1) How accurate of an estimate of the population standard deviation is the sample standard deviation?
2) As a result of (1), by how much do you over or under-estimate the probability of tail events?
3) How does the over or underestimation change as the event you're trying to estimate becomes more extreme (ie higher sigma event?)

I've previously written about (2), but the point of this post is to run the full simulation and what are the results.

To start, I generate a 500 element population, which has both a standard deviation and mean of 1.

From here, I take 1000 samples of 50 each, and from each sample I calculate the standard deviation. Below is the distribution of the percent errors of those standard deviations estimates.

As you can see, in spite of the fact that the standard deviation was measured 1000 times, there's still a substantial amount of spread in the distribution of standard deviations. While they average a 3% underestimate, they range from an underestimation of around 40% to an overestimation of over 20%.

This spread in the standard deviation estimate is particularly worrisome when one starts estimating the probability of low probability tail events. Below is the ratio of the actual left tail probability of a 3 sigma event divided by the estimated probability based on the standard deviation estimates. A big number implies an underestimation of the tail risk, while a small number implies an overestimation of the tail risk. Note the skew.

So when we're talking about a 3 sigma event, there's a very sizable risk of a dramatic underestimate. While most of the underestimation is concentrated around 1.14, which corresponds to a 13% underestimation, the ratio can go up to around 15, suggesting that risks can be massively underestimated even in a Gaussian world. Astute readers may note that the estimation error distribution looks to be log-normal, which at least has finite variance. But if a normal population can lead to a log-normal distribution of estimation error, it means that they end up with even more potential for underestimation. As pictured below, a log-normal population means your average underestimate is even higher at around 5 times, while the tail is even fatter.

Now that we've explored part 2, we can look at part 3; how does this change at different levels of standard deviation? I generate a population with a standard deviation , and collect data on it as per the sampling procedure above. I do this over and over again at different populations at different standard deviation levels. Surprisingly, I don't get a robust result for the estimation of error, as it's highly dependent on the population generated. However, most of the time, I do get the general trend for the mean of the log of the error ratio, pictured below. Note that it becomes more and more negative, suggesting that the mean underestimation goes down as the severeity of the tail event increases. This is likely because of the way the skewness works out with the over and under estimation of the standard deviation.

But is this necessarily good news? Not really. While the mean might be tending towards less underestimation, the proportion of underestimated risk stays relatively constant.

While the mean value of the underestimation given the risk is underestimated shoots up.

From these graphs, we can see that the average magnitude of the underestimation is increasing very quickly, while the proportion of underestimation stays relatively constant. It is actually increasing faster than it looks because the y-axis is the log of the ratio. Combined with the fact that the proportion of underestimation is staying constant, this implies the distribution is getting flatter and more dispersed, opening the possibility of catastrophic loss. The mean underestimation likely understates the actual damage that would result, as only one extremely bad underestimation can send the firm bankrupt, along with possibly the rest of the industry.

These graphs show with great detail at how insufficient risk measurement techniques like VaR or even ES are. Small probabilities are hard to estimate, even in a normal, Black-Swan free world. But add some grey swans, fragile balance sheets, and large banks, it's a time bomb that's just waiting to explode.