An interesting read from Noahpinion: What I learned in econ grad school

I found the following blog entry to be very interesting. In some sense, it touched on my dissatisfaction with macro, and why I quit it.
Noahpinion: What I learned in econ grad school: I always find it interesting that criticisms of economics education focus more on the gra...

Names, names...

I remember one argument my friends and I got into. Qiao and I believed in the non-neutrality of names. Naming a concept frames peoples' thoughts, and it is the gateway to introduce bias. We both felt the name of "China Proper" to denote "Inner China" is inappropriate since it seem to imply the frontier region during Qing dynasty is not properly China's land. Another friend of us disagree.

Why I suddenly think of this? I recently realized that what we economists call "price discrimination" is elegantly called "price customization" for marketing people. The very act of charging different prices to extract more surplus is depicted as a special service---customization by the marketing people. I found that to be really amusing. We actively use language to frame, mislead, and remove stigma.

I am aware of some research in political science and economics trying to study the neutrality of of media (Jesse Shapiro from Chicago Booth had a paper on newspaper ideological leaning, Ali from Stanford GSB had one one cable media, Feng from HBS looked at bias in wikipedia). Most of the work I have seen involves searching for "key words". This has lots of problems. The baseline might employ keywords very differently from media (for example, congressman with known political leaning might have different frequencies of keywords as a result of differential participation in certain topics, while media might use the same keywords to ridicule opponents). Empirically, the estimates have been noisy. Maybe one could get more mileage by looking at the different naming of the same concept.

As for me, this is a very effective screening tool for me. I like to read news, but not biased news. For example, I can pick up an article from NYT, if I found manipulation via naming, I can quickly threw it to trash can, ignoring the prose carefully crafted to convey a sense of balance and neutrality.

Smart as you are, you must be aware that I chose to name NYT in the previous example to "manipulate" you into associating NYT with that tactics. Absolutely, an eye for an eye.

What is Barbaric?

Let me start with a quite unrelated note. There is a custom of foot binding in Chinese culture. In short, the culture is about applying painfully tight binding to the feet of young girls to prevent further growth. In fact, the previous statement understates the pain associated with it. The following are two pictures. The first is a picture of a bound foot and the second is a Schema of an x-ray comparison between an unbound and bound foot.










It is a part of Chinese culture, but that does not prevent me from labeling this particular element as barbaric and backward. Despite several champaigns against it led by Woman's Christian Temperance Movement and other feminists, it did not end until Mao came to power:

When the Communists took power in 1949, they were able to enforce a strict prohibition on foot-binding, including in isolated areas deep in the countryside where the Nationalist prohibition had been ignored. The ban remains in effect today.

There are two takeaways. 1)The fact that certain custom/practice/etc is part of culture does not mean we cannot pass value judgement one it. Sometimes, it is outright barbaric. 2) To end certain barbaric behavior, reason might not be as helpful as a strict crack down.

Now let us imagine a church. The church pours resources from everyone and spend them on a very few selected people. Those people get educated at the expense of others being deprived of education. And the culture of that society is that the uneducated mass placing much faith on the few elites. That does not sound that inviting. Does it?  Let me go to what I intend to discuss in the beginning. To know more about Tibet, I consulted a close friend on what it is like in Tibet, and here is his/her response:

I have seen pious Tibetans, their religious devotion is incredible, but they are not perfectly kind, compassionate, and peaceful people (in fact ... someone's recent argument that Tibetan Buddhism is a culture infused with both extreme violence and extreme compassion). some are aggressive, and many are incredibly erotic and sexual. one can even argue that the complete devotion and trust of guru required in tibetan buddhism can be used to cultivate terrorism. but rather than concluding that religious devotion is the problem, I'd rather see the problem as the extreme lack of and inequality in education (religious and secular, in general). the way tibetans train lamas are perhaps elite intellectual education at its finest, and they do produce most incredible human beings, but that's the resources of the entire society poured into a few individuals. the vast majority is so un-educated that it is difficult to think about rationality. an uneducated population entrenched in a single belief system, that's like material waiting to be mobilized for collective aggression and chaos.

As an economist, I see gigantic problems right here.  Educating the rest of the population gonna be hard, because the lamas understand that once the people get educated and enlightened, they will not be as susceptible and worship lamas as much they did. The lamas stood to lose, just as the church in the past viewed science and enlightenment as its enemy. What the church did, is to turn the people against scientists and leaders of enlightenment. I suspect that is what happened in Tibet.

I called this practice "backward at best, barbaric at worst", but my friend protested against my use of the language. I stood by what I wrote however. Yes, every culture has good sides and bad sides, but certain elements are still barbaric.

As my friend later noted, most of the lamas are good-natured. Sure, just as I believe that most people in the church are wonderful people throughout history, but that does not mean, we should give them unlimited power.  For power abuse, it is not the majority that matters, it is the very few evil outliers that matters.

What should we do? There are only two ways. 1) leave it as it is 2) eliminate that element. The latter definitely is going to be painful. So I am not sure if we should go with the latter. But what I do wish to say here is that there is no middleground. It is not possible to campaign and advocate, hopefully people will ditch their old ways, as illustrated in the footbinding history. It is like a tumor that just does not heal by itself. Of course it is debatable if we need to remove the tumor or the pain of removing the tumor does not justify the removal, but I think it is overly romantic to wish for a middle ground.

How Seriously should We Take Economic Policy Analysis 1?

The short answer is: not much. We can take the qualitative analysis quite seriously, but in many areas, the quantitative results are far from as trust-worthy as most would like you to believe.

There are many reasons, and I have intended to blog about this for a long time, but I will only discuss one issue briefly.

In economic analysis, we do something called welfare analysis---that is we find the utility function for individuals, and then we find their utility under different policies.

The problem lies in finding the utility function. We know there is no way utility function is parametrically identified, thus, we need to assume a functional form for the utility and for that functional form, we fit the data. Almost surely (that is with probability 1), we will be using the wrong functional form and the utility function we claim to identify is a very crude approximation within that functional form. As it turns out the welfare calculation could be very sensitive to that. There is one paper that finds that the welfare improvements is infinity!! The authors admirably kept the results and explained that it is due to their functional form assumption and continued with other functional forms. I could imagine in other studies authors would just delete crazy results from the paper and pretend nothing happened.

The problem continues. For welfare calculation, we want an "experience utility", that is how people actually feel. However, we can only back out what I call "decision utility", that is agents behave as if they are maximizing their "decision utility". Unless we assume that agents do maximize their experience utility, there is no reason to believe that these two would coincide. In fact, the study of behavioral economics point to many problems. Any inconsistency problems or cognitive bias would break the link, and they exist.

Let me give another example. We all know about business cycle--that is the boom and bust, recessions. We kind of feel there is a big cost to it, and it would be quite desirable if we can eliminate it. In fact, it is so desirable that we will be willing to pay an insurance fee to eliminate it or protect ourselves completely from it. But if we apply a standard welfare calculation, an upper estimate of the welfare cost of business cycle is 1/5 of 1 percent. An individual with average consumption of \$50,000 would be willing to pay \$100 to eliminate fluctuations. This is still a very small amount compared to the implications of long run growth on income. Most economists feel this is weird, but that is what is spit out. I think this does point to the limitation of our standard welfare analysis. Perhaps, some habbit formation adjustments would give more plausible answers.

Thus, the utility function we get needs to be taken with a grain of salt and so does the policy recommendation.

What does this imply?

It does NOT imply we should stop doing welfare analysis in economics. We need to and we need to expand the scope. But we need to be careful in interpreting our results and realize the error margin is larger than it looks. I always think of the wisdom of Ariel Pakes:"A decision needs to be made in real time. This is not perfect, but this is the best we can do." On the other hand, I would not shoot for any drastic policy changes that promises a slight/moderate welfare improvements on paper.

People are Over-excited

This is my reaction to facebook buying WhatsApp. It is a stretch for me to believe that it is rational.

First, I would like to put 19b in perspective. It is a large number and we have very bad intuition for it. Some benchmark helps. See this excellent comparison.
I am kind of old school and I always use a cash flow model as a benchmark (Though I think it gives very very crude estimates, since the result is purely assumption-driven, but it does give good sense of what assumptions are needed to justify a result. Also, it gives an anchor).

So We need to consider two scenerios: 1)what will be the annual income of facebook if it buys whatsApp $Y_t$ 2)what will be the income of facebook if it buys whatsApp $Y_t'$ Then we need to consider the differential of $\Delta Y_t=Y'_t-Y_t$. As a very crude benchmark calculation, let us hold $Y_t$ and $Y'_t$ fixed and thus $\Delta Y_t=\Delta Y$. When we apply the annuity principle to find what is the value of the discounted present value of all those differential income $\Delta Y$.
\[
NPV=\Delta Y/r
\]
where $r$ is the interest rate.

If we consider an interest rate of 5 percent (risk -adjusted), then it means that the annual income differential needed to justify this purchase is 1 billion usd. That seems awefully large, as the net income of facebook in 2013 is only 1.5 billion.

When it come to tech stuff, we often believe the old rules do not apply. It is the ideas! This time is different! I regard this as nothing but bullshit. The earliest innovation that drives much excitement is the South Sea Bubble in Britain, and the Mississippi bubble in France. They were wonderful financial innovation that really could be very beneficial. The railroad bubbles seem to be got forgotten as well. We are not talking about worthless craps--like many in the tech bubbles. We are talking about things of value, and people just get over-excited.  Nothing more, nothing less.

Financial Times has a pretty interesting analysis.

If I have to speculate, the tech industry contains lots of hype. There was a time it created lots of value, and truly made things better. Skype was a wonderful start. But later things just get boring. It is just some marginal incremental improvements whose value seldome justifies the extra cost. (like iphone's first appearance was great and a couple of improvements were good, but the later innovations have run out of steam). I think the demand is now sustained more by commercial manipulation that exaggerates the differences, and prey on people's catching up with the latest mentality. This is not sustainable. There is much to catch up, and smartphones will not always be on the list.

Greatest unfairness in China?

This blog is about the registry system in China. It is one of the most controversial issues in China, often making headline, and debates often getting too heated for rationality to remain.

Background

In China, people gets assigned to a registry (a city or a town), which determine what benefits they are eligible to. For example, your registry in Shanghai will entitle you to the public education system in Shanghai. If your registry happen to be not in Shanghai, but you happen to work in Shanghai, will unless you find some way to change it or work out some other way, it is going to be a lot of trouble for you. Your choice will be limited, you might need to pay for some of the choices others can enjoy for free, and there might be a longer line for some services. The gist of the problem as many see it is that there is underprovision of public service to migrant worker. They ask, why is there a differential treatment to people? Why are people not born equal?

"Walls" of Social benefits 

The short answer is, people are never born equal.

I am talking about eligibility for social benefits here. If you look around the world, there is huge discrepency in social welfare. Some countries are doing great--Denmark is an example, but no one in his right mind would suggest that Denmark should extend its benefits to people from all over the world. In fact, Denmark would not allow people from all over the world to migrate there freely.  Many people migrate to work in other countries, and it is very common to find themselves excluded from the social benefits in the other country (they pay the taxes, by the way), in addition to lots of obstacles in finding a job. I do not wish to dwell on this too much, but it suffices to say that countries do set up walls for social benefits.

Beyond Country Level?

It makes people less comfortable when it goes beyond country level. Now let us look at the data. The following graph plots different countries and different provinces in China, on a two dimensional graph, showing their GDP per capita and life expectancy. It shows the amount of heterogeneity among different provinces in China.
www.bit.ly/aV32Mn
As you can see, the country China is like a miniature world (excluding the poorest niche). If it is reasonable for different countries to set up walls of their social benefits, then the same goes for a country like China. In fact, this graph underestimate the amount of heterogeneity, because the biggest difference lies in countryside vs. cities, and thus by mixing them together, some heterogeneity is lost.

Disruptive Flows

Of course, all these are not satisfying. We still want to ask why we need walls? The answer is simple, because flows are disruptive. Flows of migrant worker, while being instrumental for the cities' development, could be highly disruptive when it is extended social benefits too quickly. Consider a minuscule case: flows of mainland Chinese to Hong Kong to shop. These shoppers like certain things a lot, and the flow is disruptive enough for Hong Kong government to impose a ban on sales of certain items to mainlanders. The truth is, supply of public benefits, adjust even more slowly, and when more people demand a slice of the pie, the slice becomes thinner and thinner.

Finally, A Model and some General Equilibrium

We now consider the issue using the following model. There are two types of labor: skilled and unskilled labor. In the city, both types of labor exist and for simplicity, the migrant workers to the city are unskilled labor. What happens when we allow the flow of migrant worker to increase? The supply of unskilled labor goes up, and hence the relative wage of unskilled labor goes down. Thus, the flow of migrant worker hurts the welfare of unskilled labor while improving the welfare of the skilled labor (informally by decreasing the price of goods produced by unskilled labor, or more formally it is isomorphic to the Stolper-Samuelson effect in trade theory).

In reality, the effects on welfare of unskilled labor will be cushioned by public education, health care, and other benefits provided by the local government.  We consider two alternatives--1)extend those benefits to the migrant worker. 2)deny those migrant workers the benefits. If we extend the benefit, since the number of recipients increased the quality will decrease. Thus from the perspective of unskilled labor in the city, it is a double hit--their wage goes down and the public service becomes crappy. From the perspective of migrant worker, they have chosen to come to the city---via revealed preference, must have improved their life from that in their hometown. (On a side note, four years ago, I did use a Computable General Equilibrium model calibrated to model this, and it does suggest that unskilled labor in city will lose, and migrant worker will gain a little from such a flow) Empirically, the unskilled labor is much more dependent on social programs as the rich can afford private schools and now increasingly private health care. Thus, unskilled worker is the only loser in this game, and if we take the status quo as a reference point, there is a reason for denying social welfare to newcomers. What is important here is to realize that we are faced with the tradeoff between unskilled labor native to the city and the migrant worker, it is not the trade-off between rich and poor.

The reason why migrant worker will only gain a little in the model when we improve the access to social benefits is due to its general equilibrium nature. As I commented in the previous blog, we need to think general equilibrium when we consider a policy. Now consider the following shock: we give migrant workers more benefits. What will happen? In the short run, migrating to the cities to work becomes more attractive, and thus, more people will flow into the city, bidding down the price, and adding to the burden of social programs--ultimately these two forces will make working in the city less attractive to induce more flows. These two forces work on the unskilled worker in the city as well, and there is no initial positive shock to counter it--in fact, they only have an initial negative shock. I suspect, this general equilibrium consequences are never in people's mind when they discuss this issue.

Afterthoughts

I have written enough, but this is by no means a complete discussion of the issue. I only wrote out some things that I see blatantly missing in public discussion. This entry is in response to a friend's blog about the system, and I was asked to give some feedback. Thus it is for that purpose, to balance out a more proactive voice. I often feel I am a terrible person to ask for feedback, as I disagree too often. But I think I will stick with it, because not speaking what I see makes me feel worse than disagreeing. Finally, I think economics, especially (good) economic models have a role to play in public discussions. They help us organize thoughts and see things more clearly.

STEM controversy

So I was trying to do research about stem cell controversy when I ran into another STEM controversy--the debate about the policy of opening America up to people pursuing a career in science and technology. Leading the opposition is Norm Matloff, who happened to be a veteran in computer science, actually a professor in UC Davis.

When it seems that Matloff's experience would give him more credibility at the issue, I would argue it is exactly his experience that brings suspicion of conflict of interest. This policy, after all, pits the technology industry against veteran IT workers. At the risk of caricature, here is how: veterans cannot compete against the new international comers, in terms of cost efficiency, so they lose from the policy. The tech companies, now having a larger pool of labor to draw from, benefit. It is very much the same idea in medicine, when cardiologists set up associations, and refuse to recognize degrees from other nations (like EU or India, which is doing an amazing job in cardiology), they are setting up a monopoly, so as to shield them from competition. Since I mentioned conflict of interest, I should disclose mine: I am an international in US, but I am not in STEM. Thus even if I chose to stay in US after graduation, I do not benefit from STEM, and in fact, since the policy is leaning towards STEM and away from others, I might lose from it.

The main opposition is that STEM brings in cheap labor that replaces American labor, instead of "best and brightest" as in the rhetoric. Of course, reality and rhetoric is different. So I ask what is the alternative? no STEM law. Unlike medicine, which is non-traded goods/services. Technology is traded. Now we are looking at factor intensity differences, and I tend to think of the classic HO model, which tells you that trade in goods/services makes trade in factors unnecessary. What does that mean in plain English? So if we have different economies, India and US. Suppose India has more skilled labor and US has less. Then one would expect that in India, the relative wage for skilled labor will be lower than in US. HO's insights are that if you allow the country to trade, these two countries will specialize in such a way that equalizes the relative wage. Of course in reality, trade frictions prevents perfect equalization, but the force more or less works. The wage differential we observe today between rich and poor countries are mainly due to the premium to skills (that is they are different labor, rich countries have more skilled labor so they get paid more). According to HO, US will begin to outsource many of its IT job (though it gonna be a slow process, during which veterans like Matloff will still enjoy the benefit), and the demand will not grow as fast. More specifically, when an entrepreneur have an idea, he will find it more expensive to find technical people to implement it, and this deters potential innovations. It is not a new story--after all, US manufacturing lost its dominance due to outsourcing, which takes place in the absence of cheap labor inflow.

What is the bottom line? Matloff is right in that STEM will not matter for the "best and brightest", but for an industry it is not the "best and brightest" that matters. What matters for the thriving of an industry is a ready pool of competent, and cheap labor. Above all, the dynamics of being able to replace old, impotent, expensive labor with new , competent, and cheap labor is the core of growth.

What about welfare. First of all, IT people are not underpaid. They earn way more than the average American. When I say cheap, I am still talking about 100k. Second, it is THE misconception that to be labor friendly, we have to kill mobility. We have to guarantee that a laborer can always hold on to his job at the price he used to get. Some of us know it is bad for the economy in the long run, but they feel it is a fair price to pay. This needs not be true. Denmark has one of the lowest inequalities, and yet it has no minimum wage law, does not restrict companies to fire its workers. Indeed, if you look at it, their labor has high turnover rate. But they retrain themselves and find a job in a different industry. This ability to cope with change is the essence of growth, it is the best security a laborer can get.

Finally, I wish to talk about general equilibrium. Basically, I told a general equilibrium story about how trade balance things out. I never conceal my conviction that economists have a fetish for general equilibrium. We do, every models we write now must be a general equilibrium one. This has its problems because as I said general equilibrium is not immediate and the economy could be out of equilibrium for a while. Thus it will be impossible to explain some of the stylized facts we observe with a general equilibrium model. Nevertheless, I believe that any story that is not general equilibrium is not complete--it is temporary. It is fine to explain things with it, but to debate a policy with it is unacceptable. You are leaving things out with an incomplete story. You are not telling the whole truth.