Buckle up your children? Endogeneity and car seats

I was in an online discussion about car seats yesterday and recalled a paper by Freakeconomist, Steve Levitt, that for 2-6 year olds car seats are as effective at preventing deaths as seat belts.  As Levitt points out you can't just look at car seat use since people who tend to use them are wealthy and people who take less risks (likely to have better safety features on their car and get in fewer serious accidents).  In other words the choice to use a car seat is endogenous to other factors that also influence the likely hood of crash. A more recent follow up paper by Doyle and Levitt show with a different data set  a similar result in terms of serious injury 9no difference between safety seat and belt), but they also show car seats for 2-6 year olds reduce less serious injuries.

Classical economics suggests that you shouldn't tell people what to do, since you might not know their preferences. Assuming Levitt's results are correct what is the trade off between a reduction in less serious injury versus the expense and hassle of a car seat. Most economists would point out we all take risks, not driving is likely less risky than driving.

But the law and social norms say I should put my 2 year old daughter in a car seat, and I haven't really researched it so as a risk averse dad, I'm going to stick with the car seat.

This Midnight Metro Ride Brought To You By....

The Washington Nationals fans do care if they ever get back home even if it means missing the last few innings. Last week a Nationals game against second place rival the Atlanta Braves started an hour late due to rain, then went extra innings. Unfortunately, the DC Metro [subway] which many fans including me use to get to the game closes at midnight on weeknights. So late in the game the Nationals announced leave now or find another way home, and many fans had to leave early and missed a thrilling finish.

The Nationals could have prevented this by paying $30,000 to the DC metro authority for each hour the metro ran late. The Nationals didn't think it was worth it after the last time they paid for a late metro and only a few hundred fans used the service.

But, what if the Nationals make the playoffs with games starting later and lasting longer due to TV and slower play the need for late night metro is even greater. If the Nats won't pay up, why not find a sponsor here are a few suggestions to pay who would get the free press.

1. Taco Bell -- Ride the metro to your forth meal
2. The State of Georgia Tourism Board --- Take the midnight train to Georgia
3. Obama Campaign --- Trains You Can Believe In
4. Romney Campaign ---- Believe in Metro
5. CSX -- Our trains can move a ton of Nats fan in a single hour
6. Ben's Chilli Bowl (note train is direct to U Street) -- From half street to half smoke

Happy to take more. Actually I think Taco Bell would be a good fit.

Don't read this post Comcast and missing markets waiting for the Cable Guy

Last week Justin Wolfers tweeted a complaint about Comcast giving him a 13 hour window to wait for someone to install his cable. Very few people want to wait all day for the Cable guy, and I'm sure Justin a man who wrote a check to hire movers for a friend who asked for helping moving (bravo!)  isn't one to want to wait around the house all day.

So my guess is that Justin could hire someone. Perhaps find a local responsible college student who for say $100 will wait for the Karl Hungus and let him fix the cable. My first thought is does the market provide a solution.

It looks like task rabbit has people offering to wait for the cable guy. My guess is there is a limited market for this since there is imperfect information how does Justin who is new to town find someone responsible he would trust in his house.

More importantly, why is Comcast letting the person who waits get all the rents. If Justin is willing to pay someone $100 to wait all day, why wouldn't he pay $100 Comcast to get a guanteed appointment. Comcast  then would get all the benefit.

But please don't tell Comcast about my idea (unless of course they are willing to give me a 5% cut)....

Of course my guess is that Comcast has thought about this but rejected it for other reasons (they can't always do it or they worry about how it would look)

ER visits down in MA after Health Care Reform

John C. Goodman on Tuesday in a Wall Street Journal Op Ed models the potential problem of providing health insurance to people if the supply of doctors are constrained he figures that if more people have health insurance then more people will see the doctors causing longer wait times. This is particularly a problem in emergency situations as he writes

"When people cannot find a primary-care physician who will see them in a reasonable length of time, all too often they go to hospital emergency rooms. Yet a 2007 study of California in the Annals of Emergency Medicine showed that up to 20% of the patients who entered an emergency room left without ever seeing a doctor, because they got tired of waiting. Be prepared for that situation to get worse."

The question is did ER visits go up, and the answer from a recent paper by Sarah Miller forthcoming in the Journal of Public Economics (her paper can be found here) shows that actually the data indicate that the health care reform in Massachusetts decreased ER visits by 5-8% and particularly decreased non-emergency visits.

A little about the study from a quick read. One initial way to do such a study would be to compare ER visits in the years or months just before and after the reform. This would be a good first pass, but not convincing enough, because if we see that ER visits were higher in 2005 the year before the reform than say 2006 it might be from other causes than the reform (weather, the economy, the Red Sox's Bullpen?). So what Miller does is look at how rates of people in each county who were uninsured in 2005 influence influence ER visits in that year and future years. In 2005 before the reform the rate of people uninsured does not influence the rate of ER visits. After the reform the 2005 uninsured rate is negatively related to ER visits in years after the reform. This is good evidence since the reforms was likely to have a bigger influence in places with more uninsured people.  

I don't want you to reach the conclusion that health insurance reduces health care or doctor use, since other research on a program in Oregon suggests that people use more health care when they have insurance. I did want to point out though I learned all of this by following different economists on twitter.

Female and Male Wage Gap and Endogeneity Problems

Some friend of mine requested I comment on an article about the pay gap between men and women written by Ramesh Ponnuru for Bloomberg

Ponnuru is an editor at the conservative/free market leaning magazine the National Review. I start with this because the free market model of wage discrimination goes something like this. If men and women are equally productive workers, but some firms choose to pay women less than men then women will go to firms that pay them equally (or those firms can pay women slightly less making the women friendlier firms more profitable and all the discriminating firms will go out of business). Of course this assumes that not all firms are discriminating and that those discriminated against have the ability to open up their own business. For example if we take advertising agencies like Sterling/Cooper/Draper/Price it is likely that both African Americans and Women were discriminated against in terms of pay and job prospects in 1960s New York and did not have the resources to open ad agencies thus their salaries and opportunities were held down.

One point that the article makes is that the often quoted statistic that women make three-fourths the average wage that men do, but when you compare women and men in the same profession or number of hours that the wages are much closer (looking around estimates range from 85-95 depending on the industry).  In other words Ponnuru believes that the gap is overstated because women choose jobs with lower salaries and fewer hours.  He believes other things still create a gap.

However, as one economist (Sarah Miller*) pointed out the choice of a job is endogenous to the wage. This is a very key term in economics and needs explaining. If we try to control for job choice when looking at the wage gap we might underestimate that gap if women avoid fields were they are underpaid relative to men or experience worse discrimination.

So what is an economist to do? Typically we look for things that would influence women's decisions that are not determined by wages one example  might be policy changes that are not determined by the wage. This article by Jane Waldfogel in the Journal of Economic Perspectives, while a bit old is very readable (like most articles in the journal). It suggests that countries with better maternity leave policies have lower wage gaps, but that could be endogenous in that perhaps those countries with low wage gaps would be more likely to produce these types of policies. However, Japan and the United States passed new legislation and by comparing women who had children before and after the new policy and their wage gap, the hypothesis that better maternity coverage lowers the gap is supported.

Granted I'm only citing one paper, but I get the sense from googling around that a pay gap still exists, in part that is due to job choice, but figuring out how much is very hard.

* I discovered Sarah Miller's research via twitter yesterday, very interesting stuff on ER use when health insurance is provided and Prosper loans, I look forward to reading it and writing about it.

Measuring the Value of Olympics to Participant Countries

Bill Easterly makes a joke about trying to measure the impact of participating in the Olympics influence on winning medals. For countries who didn’t earn medals in London a study would show participating has no influence on medals earned. At the core of the joke is you should carefully choose what you are measuring. International aid programs might increase school enrollment or the number of businesses but it might not increase income or the number of PhD earned.

What is the goal of countries like Chile who didn’t win any medals at the Olympics? Was it a complete failure if a country goes home without any medals?  I don’t think the Jamaican bobsled team (made famous by the movie Cool Runnings) would consider their participation and last place finish in the 1988 Olympics a failure.  Even the bobsled team used that event to build toward eventually having a Jamaican win a bobsled medal at the 2006 Olympics. So perhaps participating with little chance of winning increases future chances of medals.

The next question is what is the value to a country of an Olympic medal? For a smaller country like say Costa Rica, there was a lot of national pride gained by Claudia Poll’s 1996 Gold medal and when I was there in 2000 the country was very excited to watch her swim again. I think an economic value could be put on this pride. Typically economists would use a method called Contingent Valuation where they survey people how much is a medal for your country worth to you or even ask people to donate or pay taxes to a national Olympic committee. This paper by Brad Humphreys and others suggests Canada's big investment in 2010 paid off 2-3 times the cost based on surveying people's value of Olympic medals. 

It is unclear how much a country like Chile or Costa Rica should invest in Olympians, but economic tools can help find the answer. My guess is that the answer is not zero, unless like the United States the market takes care of this investment through sponsorships. 

Hagerstown Stadium Construction

Hagerstown Maryland has a population of 40,000 and a Single A baseball team called the Suns. You might remember the Suns as the place that Bryce Harper played last year. My co-author and Towson colleague Tom Rhoads told me about how Hagerstown is trying to build a new stadium to avoid the team from moving to Winchester Virginia. You can read our previous research on minor league baseball stadium construction here, but in short we find a new stadium adds about 400,000 fans to attendance over 10 years at the A level that the suns play at. Surprising to me is the cost of the stadium listed at 30 million dollars with 10 million coming from local governments. My previous look into construction cost shows that most single A stadiums cost about half that much.  A good example might be Lancaster Jethawks who saw a large uptick in attendance and then it fell over time.

Typically when these stadiums are being discussed the team and proponents of construction  write large studies claiming that the project will pay for itself and of course Hagerstown has done just that by hiring Ripken baseball. I have only skimmed the report but my quick thoughts. This is meant as a quick comparison and not a criticism of the Ripken study, I have not read it closely enough yet.

1. The report claims that Hagerstown will see large impact on income and jobs. Nearly every study economists have done looking for the impact on the local economy of a new stadium or a team has shown zero effect on income or employment. The reasoning is that although spending on baseball or other sports might go up with a  new team or stadium spending on other entertainment might go down as families choose baseball over movies or eating out. A recent study by Nola Agha shows for some levels of minor league baseball that local incomes increase with new teams or stadium, but not the level Hagerstown plays in.  Hagerstown proximity to Fredrick which also has a minor league team might mean families substitute going out in Hagerstown for going to Fredrick (just 1/2 hour away) to see baseball if the Hagerstown team moves, so for Hagerstown not having a team could have a bigger impact than a town where there is no close other team.

2. The estimated impacts on attendance seem a bit high in the report. The report does use stadiums built in downtowns, while our study uses stadiums built between 1992-2006 so things may have changed. We find that attendance increased roughly 50% for the first 3 or so years, while their low estimates are 60% increases. Additionally, the report suggests attendance stabilizes after the 4th year, while our research shows that roughly 4 years after a stadium is built attendance is 33% higher than before and 10 years later it only about 15% higher. So their estimates in terms of attendance don't seem unreasonable.

3. In a case study of Ceder Rapids an A team with similar attendance we estimate that the return on investment of ticket revenues is roughly 2% this doesn't include concessions. What is very interesting is that with state help the bond to borrow the money will be just above 3% so with parking and concessions it should have a positive return for the team, but look to point number 1 about if this makes sense for the city.

I'm still agnostic if Hagerstown should help fund a stadium, which is strange for an economist as usually one of the tenants of sports economics is that public subsidies are a bad idea. I would be hesitant to support it since the down side of these situations  (construction cost overruns, lower than project attendance) in my anecdotal view seem to happen more than the upsides. However, given the low cost of borrowing and the potential for baseball fans to drive to Fredrick, this just might make sense.

Sponges Worthy or IUD and Diff-in-Diff

On my subway ride today I read an article on technology used in Seinfield episodes that we no longer use (movie phone, answering machines) and of course contraceptive sponges are no longer the method of today. Which is too bad because Avinash Dixit has created a model which helps show who is sponge worthy.

A good article in Slate helps us see the latest in birth control, particularly related to teens. The article first talks about falling teen birth rates and concludes that abstinence only education vs receiving information on birth control is not likely the cause. The author uses a simple method economists often refer to as difference in difference (diff-in-diff or DID for short). Suppose we see that Maryland had a teen pregnancy rate of 8% in 2000 and after 10 years of a new sex ed program that started that year we seen the pregnancy rate is 5%. We don’t know if the program is the cause in the reduction of teen births perhaps other factors. What we can do is compare Maryland to a similar state to one with similar other trends that had abstinence only. If we see in the other state that the teen pregnancy rate also fell 3 percentage points it is more likely that other factors besides the type of sex ed are at play. The article argues that all states have seen similar declines in teen pregnancy regardless of education. So using diff-in-diff we see no impact of sex ed.


The real change seems to be the new IUDs, that prevent pregnancy better than the pill for the average teen (since only 1 in 4 teens remember to take the pill each day and 10% of teens on the pill became pregnant in one year according to one study).