Thursday, October 29, 2009
Cool Graph, II
The estimates are based on state legislative voting, which might make you wonder how you could possibly compare legislators in one state with those in another. The trick is that some state representatives (for example, Barack Obama) also end up in Congress. There are enough of these overlap cases that you can put legislators from all 50 states on a common scale.
Here’s Boris’s graph showing the estimated positions of Democratic and Republican legislators in all 50 states in the past decade:
The chart below gives a nice summary of state attitudes on gay rights issues, based on estimates from national polls. It’s from a new paper, by Jeffrey R. Lax and Justin H. Phillips, both of Columbia University, that was recently published in the American Political Science Review. (Methodology for the survey estimates is on page 32 here.)
Bubbles are placed to represent public opinion on a gay rights issue, with bubbles farther to the right indicating greater public support. For example, the red bubble on the line for California shows that slightly less than half of Californians say same-sex couples should have the right to marry.
Filled-in bubbles signify that the policy has been adopted in that state (either by legislative or judicial action). The red bubble for California, for example, is not filled in, indicating that gays in the state are not currently allowed to marry.Jeffrey R. Lax and Justin H. Phillips
So who is more liberal on gay issues: the public or the public servants who write and interpret law?
Let’s take a look at the graphic above, keeping an eye on the dashed vertical line in the middle that designates 50 percent support for any given policy. Bubbles to the right of that line have support from the majority of a state’s population; bubbles to the left of the line have support from less than half the state’s population.
Notice that there are many more unfilled bubbles on the right side of the line (representing policies that the majority of people support but that have not been put into effect) than there are filled bubbles on the left side of the line (representing policies that the majority of people do not support, but that have been implemented anyway).
Something for the new economics think tank to consider
Because of course it wasn't always so. E. P. Thompson's work on early modern Britain reminds us that there was a "moral economy of the crowd" that profoundly challenged the legitimacy of the market; that these popular moral ideas specifically and deeply challenged the idea of market-defined prices for life's necessities; and that the crowd demanded "fair prices" for food and housing (Customs in Common: Studies in Traditional Popular Culture). The moral economy of the crowd focused on the poor -- it assumed a minimum standard of living and demanded that the millers, merchants, and officials respect this standard by charging prices the poor could afford. And the rioting that took place in Poland in 1988 over meat prices or rice riots in Indonesia in 2008 are reminders that this kind of moral reasoning isn't merely part of a pre-modern sensibility. (For some quotes collected by E. P. Thompson from "moral economy" participants on the subject of fair prices see an earlier posting on anonymity.)
So where do contemporary Americans show a degree of moral discomfort with prices and the market? Where does the moral appeal of the principles of market justice begin to break down -- principles such as "things are worth exactly what people are willing to pay for them" and "to each what his/her market-determined purchasing power permit him to buy"?
There are a couple of obvious exceptions in contemporary acceptance of the market. One is the public outrage about executive compensation in banking and other corporations that we've seen in the past year. People seem to be morally offended at the idea that CEOs are taking tens or hundreds of millions of dollars in compensation -- even in companies approaching bankruptcy. Part of the outrage stems from the perception that the CEO can't have brought a commensurate gain to the company or its stockholders, witness the failing condition of many of these banks and companies. Part is a suspicion that there must be some kind of corrupt collusion going on in the background between corporate boards and CEOs. But the bottom line moral intuition seems to be something like this: nothing could justify a salary of $100 million, and executive compensation in that range is inherently unfair. And no argument proceeding simply along the lines of fair market competition -- "these are competitive rational firms that are offering these salaries, and therefore whatever they arrive at is fair" -- cuts much ice with the public.
Here is another example of public divergence from acceptance of pure market outcomes: recent public outcries about college tuition. There is the common complaint that tuition is too high and students can't afford to attend. (This overlooks the important fact that public and private tuitions are almost an order of magnitude apart -- $6,000-12,000 versus $35,00-42,000!) But notice that this is a "fair price" argument that would be nonsensical when applied to the price of an iPod or a Lexus. People don't generally feel aggrieved because a luxury car or a consumer device is too expensive; they just don't buy it. It makes sense to express this complaint in application to college tuition because many of us think of college as a necessity of life that cannot fairly be allocated on the basis of ability to pay. (This explains why colleges offer need-based financial aid.) And this is a moral-economy argument.
And what about that other necessity of life -- gasoline? Public complaints about $4/gallon gas were certainly loud a year ago. But they seem to have been grounded in something different -- the suspicion that the oil companies were manipulating prices and taking predatory profits -- rather than an assumption of a fair price determined by the needs of the poor.
Finally, what about salaries and wages? How do we feel about the inequalities of compensation that exist within the American economy and our own places of work? Americans seem to accept a fairly wide range of salaries and wages when they believe that the differences correspond ultimately to the need for firms to recruit the most effective personnel possible -- a market justification for high salaries. But they seem to begin to feel morally aggrieved when the inequalities that emerge seem to exceed any possible correspondence to contribution, impact, or productivity. So -- we as Americans seem to have a guarded level of acceptance of the emergence of market-driven inequalities when it comes to compensation.
One wonders whether deeper resentment about the workings of market forces will begin to surface in our society, as unemployment and economic recession settle upon us.
This could be interesting
[F]inancier George Soros is announcing a $50 million effort to speed things along. This week Soros is gathering some of the leading practitioners of the market-skeptic school, who were marginalized during the era of "free-market fundamentalism," among them Nobelists Joseph Stiglitz, George Akerlof, Michael Spence, and Sir James Mirrlees. He's also creating an "Institute for New Economic Thinking" to make research grants, convene symposiums, and establish a journal, all in an effort to take back the economics profession from the champions of free-market zealotry who have dominated it for decades, and to correct the failures of decades of market deregulation. Soros hopes matching funds will bring the total endowment up to $200 million.
Tuesday, October 27, 2009
Putting your econometics skills to use
Choose randomly in the age group that you are interested in, and date every
single one of the girls in the sample for the same amount of time (so, you’d
have to dump them when the time comes, or you need to keep them with you as
long as the time takes, if they want to dump you first). Then you rate them,
look at the distribution, calculate the mean, and the standard deviation.
2. Now for every girl that you date afterwards, we need to do a hypothesis test:
H0: this girl’s rating is rating is the same as the observed mean.
Suppose the previous sampling follow a normal distribution, then you can do a
t-stats on the current girl friend’s rating is significantly larger than the
sample mean, then congratulations, you’ve found your “significant” other.
3. But in order to save time so that you won’t end up wasting your time on the
girls who are not statistically “significant”, you can run a multi-regression
on the sample, with repressors such as height, weight, Body Mass Index, color
of hair etc. Find out which one plays a major role in your rating, then with
these rules, you can narrow your girlfriend pooling to look for the significant
Wednesday, October 21, 2009
Health Care Costs and Wages
"There are a few things economists believe in our souls so strongly that we have a hard time actually explaining them," he said. "One is that free trade is good and another is that health-care costs come out of wages." To put it another way: Economists are pretty united on this point. A firm's compensation for its workers is pretty static, and if relatively more goes to health-care costs, relatively less will go to wages, and vice-versa. But this isn't just a matter of theory. The following graph charts the percent growth in the median household income versus the percent growth in health-care costs since 1990. The correlation is striking:
The correlation between the two data sets is -0.8, which is incredibly high for this sort of thing. To give you an idea, I also ran the correlation between GDP growth and median wages over the period: It was 0.7, which is to say that there was a weaker connection between economic growth and median wages than between health-care costs and median wages.
There is, in other words, very good evidence that employers pass health-care savings onto employees. A Rand study by Dana Goldman, Neeraj Sood and Arleen Leibowitz examined a particular firm's response to a period of premium increases and found that "about two-thirds of the premium increase is financed out of cash wages and the remaining one-thirds is financed by a reduction in benefits." Another study by Katherine Baicker and Amitabh Chandra found that a 10 percent increase in premiums "results in an offsetting decrease in wages of 2.3 percent," which is fairly impressive given that income is much higher than health-care premiums.There's good reason to think that if health-care costs can be tamed, wages will rise. But one of the big problems in health-care reform is that workers don't understand this connection. They think of health-care coverage as a "benefit," rather than a form of compensation engaged in a fairly zero-sum competition against their wages.
This created an internet firestorm -- particularly because the main climate scientist they rely on feels his views on the importance of CO2 emissions and geoengineering were misrepresented. Here's a very large collection of links on the issue. It can be a bit overwhelming, but I encourage you to read some of them. The authors "responses" can be found here (note -- there's a lot of attacking the messengers without responding to the message).
Tuesday, October 20, 2009
Education and Economic Growth
... an extra year of school is associated with more than a 30 percent increase in per capita income.
One explanation for the extraordinarily strong relationship between national earnings and education is that the correlation is largely spurious.
Perhaps, richer countries choose to become more educated, so that higher income causes higher levels of education rather than the reverse. Perhaps countries with other positive attributes, like better governments, are both richer and better schooled. The individual level research has labored hard to find so-called “natural experiments” — like mandatory schooling laws that start abruptly and relatively arbitrarily in a particular year in a particular state — that enable researchers to estimate the returns to education holding individual aptitude constant. Cross-country work is not nearly as well-identified and it never will be.
Yet there are reasons to think that the education-income relationship is not mere happenstance.
Historic educational enrollments predict subsequent income growth quite well, even when holding past income constant, which seems to reject the view that education is just following income. Moreover, the link between initial education and subsequent income growth has proven to be remarkably robust, surviving any number of country-level controls, including controls for governmental quality. I’m not confident that all of the measured cross-national relationship between schooling and income is real, but more evidence supports that interpretation than any obvious alternative.
If the link between country level income and education is real, then we need to understand why the link between schooling and education gets magnified at the country level. Why should there be a social multiplier that causes the schooling-income link to increase at higher levels of aggregation?
One hypothesis is that there are spillovers from education, and that human capital enables places to gain access to better technologies. Human capital externalities occur when one person’s education makes his or her neighbors more productive. Comparing people within a country only picks up the direct effect of more education on the person being educated, and won’t reflect any of these externalities. Comparing across countries picks up all the externalities that would occur if one’s persons schooling provided benefits for everyone around.
Within the United States, researchers have found that when holding an individual’s own years of schooling constant, that individual’s earnings increase by around 9 percent as the share of college graduates in that individual’s metropolitan area increases by 10 percent. If you work around skilled people, you earn more, either because you have learned from those people or because more skilled entrepreneurs make production more efficient. After all, Buenos Aires in 1900 had significantly fewer innovative technologies than did Chicago in the same year, perhaps because of lower levels of schooling.
But country-level income and education may also be closely linked because of politics. There is a very strong correlation between quality of government and education. My work with Felipe Campante suggests that the link between Argentina’s low level of education in 1900 and its poor economic performance over the next century reflects, in part, the fact that Argentina’s lower levels of education led to worse political outcomes.
Monday, October 19, 2009
Online STATA tutorial
Friday, October 16, 2009
Thursday, October 15, 2009
The value of school
Experiences and skills acquired in school reverberate throughout life, not just through higher earnings. Schooling also affects the degree one enjoys work and the likelihood of being unemployed. It leads individuals to make better decisions about health, marriage, and parenting. It also improves patience, making individuals more goal-oriented and less likely to engage in risky behavior. Schooling improves trust and social interaction, and may offer substantial consumption value to some students. We discuss various mechanisms to explain how these relationships may occur independent of wealth effects, and present evidence that non-pecuniary returns to schooling are at least as large as pecuniary ones
Oreopoulos and Salvanes go on to make a back-of-the-envelope calculation of the size of the nonpecuniary effects. They suggest that
about three quarters of the schooling effect on selfreported life satisfaction is due to non‐pecuniary factors. A 12 percent increase in annual earnings would then imply that the total non‐pecuniary gains are equivalently worth another 16 percent increase in earnings (for a total of 28 percent).
Book of Odds
1 in 10 the odds a man in the US does not own a pair of blue jeans
1 in 100,000 odd a man will be diagnosed with breast cancer in a year
1 in 4.76 the odds that an ever-married or cohabiting man has cheated during the relationship
1 in 9.09 the odds that an ever married or cohabiting man has cheated during the relationship
And here's an article for shark week fans -- you are twice as likely to be killed by a vending machine than a shark (do note this follow-up though which provides some important context).
Perhaps these will help people better maximize their expected utility.
Wednesday, October 14, 2009
For Ostrom, you can kinda boil it down to this Curb Your Enthusiasm clip.
From the archive: Women and negotiation
I attended a conference yesterday on women professionals. Some interesting stuff was presented. I particularly enjoyed Linda Babcock's talk on women and negotiation. Babcock has written extensively on this topic. The results she presented yesterday were from this paper:
Two experiments show that sex differences in the propensity to initiate negotiations may be explained by differential treatment of men and women when they attempt to negotiate. In Experiment 1, participants evaluated candidates who either accepted compensation offers without comment or attempted to negotiate higher compensation. Men only penalized female candidates for attempting to negotiate whereas women penalized both male and female candidates. Perceptions of niceness and demandingness mediated these effects. In Experiment 2, participants adopted candidates’ role in same scenario and assessed whether to accept the compensation offer or attempt to negotiate for more. Women were less likely than men to choose to negotiate when the evaluator was male, but not when the evaluator was female. This effect was mediated by women’s nervousness about negotiating with male evaluators. This work illuminates how differential treatment may influence the distribution of organizational resources through sex differences in the propensity to negotiate.
For more on this topic, you can see facts and discussion of her book "Women Don't Ask" (with Sara Laschever) at this website. A taste:
Women Don't Like to Negotiate
In surveys, 2.5 times more women than men said they feel "a great deal of apprehension" about negotiating.
Men initiate negotiations about four times as often as women.
When asked to pick metaphors for the process of negotiating, men picked "winning a ballgame" and a "wrestling match," while women picked "going to the dentist."
Women will pay as much as $1,353 to avoid negotiating the price of a car, which may help explain why 63 percent of Saturn car buyers are women.
20 percent of adult women (22 million people) say they never negotiate at all, even though they often recognize negotiation as
Women are more pessimistic about the how much is available when they do negotiate and so they typically ask for and get less when they do negotiate—on average, 30 percent less than men.
appropriate and even necessary.
From the archive: Do girls cause divorce?
There is evidence that they might. Steven Landsburg discusses the evidence and the debate in two columns in Slate. The reseach it is based on is by Gordon Dahl and Enrico Moretti. Thier paper can be found here. Here is the abstract to the paper:
This paper shows how parental preferences for sons versus daughters affect divorce, child custody, marriage, shotgun marriage when the sex of the child is known before birth, and fertility stopping rules. We document that parents with girls are significantly more likely to be divorced, that divorced fathers are more likely to have custody of their sons, and that women with only girls are substantially more likely to have never been married. Perhaps the most striking evidence comes from the analysis of shotgun marriages. Among those who have an ultrasound test during their pregnancy, mothers carrying a boy are more likely to be married at delivery. When we turn to fertility, we find that in families with at least two children, the probability of having another child is higher for all-girl families than all-boy families. This preference for sons seems to be largely driven by fathers, with men reporting they would rather have a boy by more than a two to one margin. In the final part of the paper, we compare the effects for the U.S. to five developing countries.
What do you think is going on?
Second, here's an interesting TED talk from a few years back on how we ended up with so many varieties of spaghetti sauce.
EC 303 Quiz 1 Solutions
Tuesday, October 13, 2009
EC 292 Assignment
Monday, October 12, 2009
EC 303 assignment
Thursday, October 08, 2009
Bet on your grades
While hanging out together one Sunday afternoon, I mentioned to my friend Steven Wolf that I had an exam the following day and that if I were to study I was sure to get an A. (At the time, I was a student at University of Pennsylvania.) But I was enjoying my Sunday afternoon, and I told Steven that I had no intention of studying. That's when, in order to provide me with motivation, we made the following agreement: If I got an A on the exam, he would give me $100, and if I didn't get an A, I would give him $20. Steven and I quickly realized that lots of other students might like this kind of motivation. To that end, we began developing what is now Ultrinsic Motivator Inc.They'll also let you hedge your grade risk by buying grade insurance. I love that you can now but insurance against my grading practices.
Return to blogging
A key rule of thumb to keep in mind when using descriptive statistics to make an argument is "no naked numbers." Descriptive statistics alone are largely meaningless. They only take meaning when you have something to compare them to. Occasionally, we establish context by comparing them to the numbers our gut tells are high or low. This can be dangerous. You can end up looking like a real idiot if someone else bothers to actually place your numbers in context.
Subscribe to Posts [Atom]