Holzman’s Article
Pages
Categories
Archives
Does the Minimum Drinking Age Save Lives?
In the early 1980s, researchers began to compare the rates of teen traffic fatalities in states with a drinking age of 21 versus those with one of 18. The resulting statistics startled legislators and prompted them to pass the Federal Uniform Drinking Age Act (FUDAA), which withheld federal highway funding from states who kept their drinking age below 21. Although this law essentially gave the federal government control of an issue previously delegated to states, the Supreme Court ruled in favor of the act, maintaining that the high national rate of teen accidents called for federal intervention. Since then, the assumption that a higher drinking age reduces alcohol-related traffic fatalities has remained unchallenged. But as Jeffrey Miron and Elina Tetelbaum, of Harvard and Yale, respectively, argue in their new study, “Does the Minimum Drinking Age Save Lives?”, this widely held public belief may be based on a misinterpretation of the available data.
When judging the effectiveness of FUDAA, researchers have traditionally focused on the period immediately following its enactment in 1986 and have failed to control for confounding variables such as improved safety features in automobiles and different state speed limits. The teen traffic fatality rate per mile driven has been falling since the late sixties, nearly 20 years before the federal drinking age was set at 21. The decline immediately following the FUDAA was a part of a much larger trend. While there was a significant initial drop in fatalities after the enaction of FUDAA, the rate of decrease slowed within two or three years to what it had been before the act was passed.
It is important to note that federal statistics used to justify the drinking age measure traffic fatalities in general – not necessarily alcohol-related collisions. The lower rate of fatalities may be due to more numerous hospitals and better medical treatments, instead of a fall in the rate of collisions. As Miron and Tetelbaum note, many other confounding variables plague the data: the states that showed the most significant drop in teen traffic fatalities after FUDAA were those that voluntarily implanted the law before the 1986 deadline. These states were also those most likely to enforce other traffic and safety measures such as lower speed limits. When the researchers controlled for these and other variables, the correlation between a minimum drinking age of 21 and fewer teen fatalities disappeared. In fact, in nine states it appeared that the higher drinking age had temporarily increased the fatality rate. There was no overall trend, the authors conclude, and no demonstrable long-term benefit from the FUDAA.
The purpose of the Federal Uniform Drinking Age Act, when it was written in 1984, was to reduce teen traffic fatalities. Many claim it has achieved its goal. However, a less myopic view of the data, one that takes other variables and long-term trends into account, reveals that its impact on fatalities has been slight – if not non-existent. The FUDAA also represents an intrusion by the federal government on states’ rights, and can additionally be linked to an increase in binge drinking among underage drinkers. These unfortunate externalities might be more tolerable if the FUDAA were indeed effective at lowering fatalities among young drivers, but Miron and Tetelbaum convincingly argue that it isn’t. Even if one is not completely persuaded by their argument, it remains clear that the issue deserves to be re-examined.
Jeffrey Miron is the Director of Undergraduate Studies for the Economics Department at Harvard University. Elina Tetelbaum is a member of the Yale Law School Class of 2010. The full text of the article can be read at http:// papers.ssrn.com/sol3/papers.cfm?abstract_id=1000359Professor Dirk Bergemann
Professor Dirk Bergemann, Douglas and Marian Campbell Professor of Economics
Education: Goethe University Frankfurt, B.A. Economics (’89), Fulbright Scholarship to U Penn, Economics M.A. (’92) and Ph.D. (’94)
Professional Career: Assistant professor at Princeton, Professor at Yale for the past 14 years
Hobbies/Interests: Tennis and squash, modern art and music, and travel
Research Focus: Game Theory applied to Strategy, Auctions, and Market Design
We all have a mental image of an auction. A buzzing hotel conference room, bidders holding up numbered cards as they shout out a price; auctioneer standing at the podium with gavel in hand, ready to sell off at the highest price. Who knows, maybe it’s all for a cliché painting.
Now scrap that image because as Dirk Bergemann sees it, the Internet and companies like Google are changing everything.
Professor Bergemann has researched game theory for most of his professional career. As he puts it, “much of economics is only fascinating when each party has asymmetric information, leading to the questioning of actual deal values.” But as he’s finding in his research, the World Wide Web has redefined our concept of auctions and game theory application.
“The Internet is interesting… because of the large number of individuals interacting with each other and the highly decentralized and private nature of information. Yet it has the ability to aggregate all of this information in an incredibly fast, rapid, and efficient rate at a scale you couldn’t have thought about before,” Bergemann tells us as we sit in his Hillhouse Avenue office. “Think of recommender websites like Netflix and Amazon; you not only see the product but you also see ratings which customers have given. So this is an example where you have information aggregation; even if I don’t know the right product for me, I now have the ability to retrieve that information which would otherwise have not been available. In other words, I can now make a more informed choice.”
The effect of the Internet on game theory application seems consequential, but Google demonstrates just how much market structures can change. “Google is making basically 95% of its profits from auctions that they run for sponsored search. It’s sort of an amazing thing if you think about it, it basically says to advertisers ‘we are willing to list your link in a prominent place (at the top or on the side of the search page) if you are going to bid for the right to be listed there under that specific keyword.’ So Google runs a completely open real-time online auction where everybody can submit at any time a bid to have his/her link listed following a search on the Internet. This auction is real-time and ongoing, meaning at any time an advertiser (think Amazon or General Motors) can basically say ‘well, if this keyword or combination of keywords comes up, please put me on top and here’s how much I’m willing to pay for you to do that.’ “
So far, this doesn’t sound so different from the typical auction. So we asked Professor Bergemann, what was Google doing differently? “Think about a normal auction, it’s an open outcry auction where you can observe the bidders, the bids, and the winning bid. And sometimes of course the identity of the winner is not known, but you have all the bidding data” (in a sense, you’ll know how much higher you would have to bid).
“What Google is doing is telling you nothing of that at all, they just say: ‘you tell me the maximum price that you’re willing to pay, and I’m going to weigh your bid with some unknown weight which I compute internally. I guarantee you that you won’t have to pay more than what you bid (sometimes you have to pay even less) but that’s all the information that you get’.”
So right now there is no public transparency into how Google is weighs these bids? “No, and actually that’s exactly the interesting aspect of it. There are some reasons why Google may want to keep this private: Google uses algorithms, and their values are basically coming from the huge amount of information that they collect. So in some sense they have propriety rights on the information, but it’s still important to observe that it’s very different from the sort of auction that we are familiar with. What we would like to know is what are the properties of this auction. Is it an efficient auction? What are its revenues? What kind of implications does it have in terms of the welfare for small vs. large bidders, and so on. This is interesting empirically, as there are millions of keywords and there are auctions going on continuously for these keywords. In terms of real world bidding, this is an incredible trove of data.”
While Google may be one of the largest corporations worldwide, truly robust research into Google’s methods will surely have implications beyond its own activities. We asked Bergemann what these areas would be, and he aptly pointed to the role which asymmetric information played in the recent financial crisis. “Coming from a microeconomist/game theory point of view, what’s interesting is that most of the collapse of the market was due to a dramatic increase in sensitivity of information; people were very careful not to reveal their information. One of the outgrowths of the crisis is the need to research how to make markets informationally robust and [therefore] less sensitive”.
Given the advent of the Internet, Google was able to fundamentally change the balance of game theory. We asked Professor Bergemann whether he believed it was feasible to experience another paradigm shift of similar magnitude in our lifetimes. His response: “Absolutely. Google is less than 10 years old, and that is an incredibly short amount of time given the large amount of mobile data being provided to people at their fingertips. In the next decade, as the Internet expands and more data is collected and transferred, it will have dramatic implications for the transformation of everything we know. Simply put, the rate of change has dramatically increased.”
Looks like the days of the traditional auction are numbered.
Posted in Economics at Yale, Latest 1 Comment
Why do the Poor live in Cities?
Most modern American cities have undeniably high rates of poverty. In 2000, more than twice as many Americans living in central cities lived in poverty than those living in suburbs. In U.S. metropolitan areas, the poor live closer to, and seem to be attracted to, the city center, while wealthier residents of the same metropolitan areas tend to live in suburbs, far away from the city center. Why do the poor live disproportionately in cities?
This puzzle is a central question in urban econom- ics. Early work in urban land use theory explains the urban centralization of the poor by arguing that the rich buy more land and therefore choose to live where land is cheap. This makes sense as long as the income elasticity of demand for land is greater than the income elasticity of travel costs per mile – that is, as long as people are still willing to move outside the city, even given the longer commute. The model assumes that everyone uses the same mode of transportation and that the main cost of transport is time. In this case, the poor will live in cities if and only if the income elasticity of demand for land is greater than one.
However, this early model is fl awed, and has limited applicability to modern American cities. First, modern cities are not monocentric; rather, the vast majority of the jobs of a city tends to be located over three miles from the Central Business District (CBD). Second, the income elasticity of land for people living in apartments and detached homes is to be roughly 0.25, which is far less than one, the value necessary for the above model to hold.
The 2000 U.S. Census shows that the average poverty rate in American cities drops significantly, from about 20% to 7.5%, as you move from the CBD of a city to its suburbs. How can we tell that this connection between city residence and poverty comes from treatment – that is, cities make people poor – rather than from selection, where the poor disproportionately move to central cities? Here, the data support selection: although ghettos may exacerbate poverty, poor people move disproportionately to the center of the cit- ies, either when switching homes or moving to a new metropolitan area.
Given the high proportion of the urban poor who are Black, one might think that inner-city poverty is really just another example of the segregation of minorities. However, the authors found that poor Whites have roughly the same central city – suburb poverty gap as Blacks, so it is unlikely that race plays an important role in the centralization of the poor.
Read on as Glaeser, Kahn, and Rappaport explore the urbanization of the poor and the reasons behind this phenomenon. They consider historical ideas and formulate their own theory on the importance of public transportation in causing this pattern. Download the PDF here, exclusively at YaleEconomicReview.com!
Posted in Academic Pieces, Latest 2 Comments
