(1) Favorite Book - "Three Cups of Tea," chronicling the story of Greg Mortensen, an American mountaineer who, through a series of incredible events, took up the cause of building schools in Northwestern Pakistan and Afganistan, is a great way to close out the year. Mortensen's work (upwards of 70 projects in the region) in education, public health and skill building has probably done more (per dollar or per unit effort) to fight the likely root causes of terrorism - poverty and extremely high opportunity costs to obtaining education and skills - than anything else in recent memory.
A good story to remind us how to go about "going about" next year (and subsequent ones, as well).
(2) Favorite Film - A good hard look at crime, terrorism, civil liberties and how far "good" can go before crossing the line, along with some ridiculously sharp acting by Christian Bale, Heath Ledger and Aaron Eckhart, compels me to put "The Dark Knight" at the top of my film list. Just because the Academy is too stodgy to give "popcorn flicks" Best Picture nominations doesn't mean that this phenomenal movie shouldn't get a nod. (Other favorites: "The Visitor," "Wall-E," "Slumdog Millionaire," and "Milk").
(By the way, "Least Favorite Trailer of 2008" goes to the one for the forthcoming Meryl Streep film "Doubt." Not only was it bizarre the first time around, I've been exposed to it before every movie I've seen in the last six months. Do they ever plan on releasing this film, or is the strategy to just keep showing the preview over and over for some extended period of time?)
(3) Favorite Academic Discourse - Is happiness contagious? James Fowler and Nicholas Christakis argue that it is: you will be happier if you have happier friends, or if your friends have friends who have happy friends. Fowler and Christakis have also argued in the past that obesity and smoking are contagious, as well.
In the same issue of the BMJ containing the happiness piece, Ethan Cohen-Cole and Jason Fletcher (the latter of Yale University) argue that Fowler and Christakis' methodology to examine peer or social-network effects is faulty in that they do not fully control for the common environment shared by all the individuals in the network (after all, you and your friends may be happy because all of you were exposed to a city-wide balloon race or something like this). Using the Fowler-Christakis setup, they are able to show that things that shouldn't be contagious - acne, height and headaches - appear to be susceptible to peer effects just like happiness. When fully controlling for the common environment, these spurious causal effects disappear. (Yup, "Favorite Falsification Test of 2008").
Guess which piece got all the press attention. Justin Wolfers at Freakonomics has a good commentary on the discourse and the manner in which it was reported.
(4) Favorite Post - I really had a good time writing this one.
(5) Favorite Historic Moment - That Obama guy winning.
And, finally, best wishes for the New Year! I'll see you in 2009.
Tuesday, December 30, 2008
Monday, December 29, 2008
Harvey Milk Documentary on Hulu
I just found out that you can watch the documentary, "The Times of Harvey Milk" on hulu.com. Here is a link to the film.
I recently saw "Milk," starring Sean Penn, and found it to be extremely interesting, inspiring and ridiculously well acted. After watching the film, I was told by a few people that I might have gotten more out of the Penn version by watching the aforementioned documentary first. Regardless of what you choose to do, you should definitely see the film and check out the documentary.
I recently saw "Milk," starring Sean Penn, and found it to be extremely interesting, inspiring and ridiculously well acted. After watching the film, I was told by a few people that I might have gotten more out of the Penn version by watching the aforementioned documentary first. Regardless of what you choose to do, you should definitely see the film and check out the documentary.
Cell Phone Free
Last week, I decided to treat myself to a smartphone and ordered a Blackberry Storm. While my new unit was being activated and sent over, the SIM card on my older phone was deactivated. As I haven't yet received the Blackberry, I have been cell phone-less for the last five days or so.
Not having a phone has been an interesting experience. Some thoughts:
(1) I am much more productive now with work: I need to get a draft of my third dissertation paper done by the first week of Jan (check back in over the next few days for a summary of this work), and I've made more progress in the last week or so than over the previous four weeks.
(2) My social life has taken a tremendous hit: I sent out an e-mail to some friends and family giving them my home phone number. In spite of me being an excellent conversationalist, nobody has called my alternate number. I attribute this to the fact that having a cell phone diminishes the returns to memorizing phone numbers to zero. As such, the fixed cost of learning a new number or entering into your address book becomes high enough for people to wait until your old number comes back online to call. At least, this is what I am telling myself.
(3) My risk of getting cancer has probably remained constant: There is a growing and very controversial body of work on whether cell phone use (or exposure to mobile towers) increases ones risk of getting various tumors. Most of these studies suffer from omitted variables bias and the fact that cell phone use being widespread only for 10-15 years or so, the follow up period is just not long enough to make many of the purported positive results to actually be plausible.
(4) My risk of getting in a car accident may not have changed, either: This one is based on a really clever piece of research by Saurabh Bhargava and Vikram Pathania trying to get at the causal effects of cell phone use of traffic accidents. From their abstract:
Previous research in the laboratory and by epidemiologists has compared the danger of cell phone use while driving to that of illicit levels of alcohol. This paper investigates the causal link between driver cell phone use and crash rates by exploiting a natural experiment - the discontinuity in marginal pricing at 9pm on weekdays when cellular plans transition from peak to off-peak pricing. We first document that this pricing threshold induces a 20 to 30% jump in call volume for two samples of callers. We then document the corresponding change in the fatal and all crash rate. Using the years prior to the introduction of two-tier pricing as a control, as well as weekends as a second control, we find no evidence for a relative rise in crashes after 9pm on weekdays from 2002-2005. The upper bounds of our estimates rule out increases in all crashes larger than 1.0% and increases in fatal crashes larger than 1.3% - these upper bounds reject the increases implied by most existing studies. An analysis of regional trends in cell phone ownership and crashes, legislation banning driver cell phone use, and differences in urban and rural ownership confirm our basic result. We discuss possible explanations and present a behavioral model to reconcile this counterintuitive finding with existing research.
Not having a phone has been an interesting experience. Some thoughts:
(1) I am much more productive now with work: I need to get a draft of my third dissertation paper done by the first week of Jan (check back in over the next few days for a summary of this work), and I've made more progress in the last week or so than over the previous four weeks.
(2) My social life has taken a tremendous hit: I sent out an e-mail to some friends and family giving them my home phone number. In spite of me being an excellent conversationalist, nobody has called my alternate number. I attribute this to the fact that having a cell phone diminishes the returns to memorizing phone numbers to zero. As such, the fixed cost of learning a new number or entering into your address book becomes high enough for people to wait until your old number comes back online to call. At least, this is what I am telling myself.
(3) My risk of getting cancer has probably remained constant: There is a growing and very controversial body of work on whether cell phone use (or exposure to mobile towers) increases ones risk of getting various tumors. Most of these studies suffer from omitted variables bias and the fact that cell phone use being widespread only for 10-15 years or so, the follow up period is just not long enough to make many of the purported positive results to actually be plausible.
(4) My risk of getting in a car accident may not have changed, either: This one is based on a really clever piece of research by Saurabh Bhargava and Vikram Pathania trying to get at the causal effects of cell phone use of traffic accidents. From their abstract:
Previous research in the laboratory and by epidemiologists has compared the danger of cell phone use while driving to that of illicit levels of alcohol. This paper investigates the causal link between driver cell phone use and crash rates by exploiting a natural experiment - the discontinuity in marginal pricing at 9pm on weekdays when cellular plans transition from peak to off-peak pricing. We first document that this pricing threshold induces a 20 to 30% jump in call volume for two samples of callers. We then document the corresponding change in the fatal and all crash rate. Using the years prior to the introduction of two-tier pricing as a control, as well as weekends as a second control, we find no evidence for a relative rise in crashes after 9pm on weekdays from 2002-2005. The upper bounds of our estimates rule out increases in all crashes larger than 1.0% and increases in fatal crashes larger than 1.3% - these upper bounds reject the increases implied by most existing studies. An analysis of regional trends in cell phone ownership and crashes, legislation banning driver cell phone use, and differences in urban and rural ownership confirm our basic result. We discuss possible explanations and present a behavioral model to reconcile this counterintuitive finding with existing research.
Thursday, December 18, 2008
Nudges and Matrimonial Websites
The Freakonomics blog has a great post on the behavioral economics behind minimum-payment requirements for credit cards. Citing some recent research by Neil Stewart, the post explains how providing a minimum threshold may induce people to actually make a smaller payment than they otherwise would have. This behavior comes about because of our susceptibility to anchoring: we tend to base our decisions on even arbitrary values we are given as starting points (see here for some examples).
This result is clearly of interest to behavioral economists: decades of research at the intersection of psychology and economics has led some observers to suggest that these tools be used in policy. That is, the irrational behavioral foibles of humans can be used to "nudge" them into making the right decisions (the term comes from a new book by Richard Thaler and Cass Sunstein).
Are there other avenues where nudges may be effective? And are there (other) examples where well-intentioned nudges may be counterproductive? A good illustration for both queries comes from Indian matrimonial websites. Like all social networking, a given user's experience on a martrimonial website depends on the activities of his or her peers. Inactive individuals do not contribute anything to the marketplace, while active ones generate all sorts of positive externalities. As such, it is in the best interest of the website company to get people active, either by introducing clarity to the user interface, making the website more fun, recruiting tons of singles, or by nudging existing users to make more contacts.
Regarding the latter, one particular website keeps track of whether a given user responds to an expression of interest from another user and the number of times that user accesses the website. Based on their activity patterns, slow or shy users are then sent an e-mail urging them to use the website by saying they are (1) missing out and (2) being rude to others by not responding. While perhaps not based in behavioral economics per se, the idea is to use some form of shame and opportunity cost argument to get people more involved.
But does this nudge work? I could easily see something like this being counterproductive. After all, nobody likes a nag, and I would bet there are plenty of people who would stop using the website just to avoid being made to feel guilty. Furthermore, messages (1) and (2) are conflicting: one is positive and the other negative. This might just confuse people.
I think this sort of thing is ripe for an experiment. However, this would require some alternate nudge option. Any ideas?
This result is clearly of interest to behavioral economists: decades of research at the intersection of psychology and economics has led some observers to suggest that these tools be used in policy. That is, the irrational behavioral foibles of humans can be used to "nudge" them into making the right decisions (the term comes from a new book by Richard Thaler and Cass Sunstein).
Are there other avenues where nudges may be effective? And are there (other) examples where well-intentioned nudges may be counterproductive? A good illustration for both queries comes from Indian matrimonial websites. Like all social networking, a given user's experience on a martrimonial website depends on the activities of his or her peers. Inactive individuals do not contribute anything to the marketplace, while active ones generate all sorts of positive externalities. As such, it is in the best interest of the website company to get people active, either by introducing clarity to the user interface, making the website more fun, recruiting tons of singles, or by nudging existing users to make more contacts.
Regarding the latter, one particular website keeps track of whether a given user responds to an expression of interest from another user and the number of times that user accesses the website. Based on their activity patterns, slow or shy users are then sent an e-mail urging them to use the website by saying they are (1) missing out and (2) being rude to others by not responding. While perhaps not based in behavioral economics per se, the idea is to use some form of shame and opportunity cost argument to get people more involved.
But does this nudge work? I could easily see something like this being counterproductive. After all, nobody likes a nag, and I would bet there are plenty of people who would stop using the website just to avoid being made to feel guilty. Furthermore, messages (1) and (2) are conflicting: one is positive and the other negative. This might just confuse people.
I think this sort of thing is ripe for an experiment. However, this would require some alternate nudge option. Any ideas?
Tuesday, December 16, 2008
Yale and the Financial Crisis
Check out the detailed consequences of Yale's 25% drop in the endowment value over the last six months or so here. The situation at other American universities also looks somewhat glum.
Be Careful With Natural Experiments
Those of you who visit this space regularly know a thing or two about my obsession with causal effects. Answering many questions in health economics requires a strategy to understand the causal effect of one variable on another, and recovering such effects demands clever strategies or tools that go beyond simple multivariate models of some X on some Y. The cleanest way to get the causal effect of X on Y is to randomize X. This strategy has been used widely in laboratory and clinical medicine, and is now being exploited in a growing number of influential papers in economics and political science.
In many cases, however, it may not possible to randomize X or the question of interest involves some program or event that occurred in the past. In these situations, researchers looks for other sources of variation in X that are effectively random - the natural experiment. A good (and now famous) example from health economics involves the impact of early life events on health and socioeconomic position later in life. A great deal of early work in epidemiology found links between the disease environment faced by an individual at birth and this individuals health later in life. This link could be causal: fetal health influences organogenesis and development that goes on to influence adult health. At the same time, alternate explanations are possible: children born to poor parents become poor themselves, which affects their health. To complete the story, poor parents may tend to reside in poor, diseased areas.
To get around this issue, Douglas Almond, in an influential 2006 paper, utilized the influenza pandemic of 1918, which struck without warning, over a short period of time, and had large, notable effects. Being exposed to influenza in utero can be thought of as a random shock (a natural experiment), and Almond took advantage of this property to derive the causal effects of health in utero on outcomes later in life.
However, while the "influenza strategy" is as close to a slam dunk as you could possibly get in observational research, other things that may seem like natural experiments a priori may not be as definitively good. In fact, such variation may even lead researchers astray.
A new NBER working paper explores this issue in detail. Kasey Buckles and Daniel Hungerman consider the case of season of birth, which has been shown to be associated with a variety of health and socioeconomic outcomes later in life. These associations have been attributed to fetal exposure to different weather conditions or differential exposure to arbitrary age-cutoffs (in sports or in schooling). At first glance, season of birth appears to be a great source of exogenous variation for a slew of different causal questions: after all, individuals don't have any control over when they are born and it seems like something that would be left to chance. However, Buckles and Hungerman convincingly argue that this is not the case:
In this paper we consider a new explanation: that children born at different times in the year are conceived by women with different socioeconomic characteristics. We document large seasonal changes in the characteristics of women giving birth throughout the year in the United States. Children born in the winter are disproportionally born to women who are more likely to be teenagers and less likely to be married or have a high school degree. We show that controls for family background characteristics can explain up to half of the relationship between season of birth and adult outcomes. We then discuss the implications of this result for using season of birth as an instrumental variable; our findings suggest that, though popular, season-of-birth instruments may produce inconsistent estimates. Finally, we find that some of the seasonality in maternal characteristics is due to summer weather differentially affecting fertility patterns across socioeconomic groups.
This is a neat paper and serves as a good warning to those interested in finding natural variation to identify causal effects. Another excellent paper on the same subject, by Mark Rosenzweig and Ken Wolpin, goes through a variety of other potentially fallacious natural experiment examples and is a must read for anyone doing empirical work.
In many cases, however, it may not possible to randomize X or the question of interest involves some program or event that occurred in the past. In these situations, researchers looks for other sources of variation in X that are effectively random - the natural experiment. A good (and now famous) example from health economics involves the impact of early life events on health and socioeconomic position later in life. A great deal of early work in epidemiology found links between the disease environment faced by an individual at birth and this individuals health later in life. This link could be causal: fetal health influences organogenesis and development that goes on to influence adult health. At the same time, alternate explanations are possible: children born to poor parents become poor themselves, which affects their health. To complete the story, poor parents may tend to reside in poor, diseased areas.
To get around this issue, Douglas Almond, in an influential 2006 paper, utilized the influenza pandemic of 1918, which struck without warning, over a short period of time, and had large, notable effects. Being exposed to influenza in utero can be thought of as a random shock (a natural experiment), and Almond took advantage of this property to derive the causal effects of health in utero on outcomes later in life.
However, while the "influenza strategy" is as close to a slam dunk as you could possibly get in observational research, other things that may seem like natural experiments a priori may not be as definitively good. In fact, such variation may even lead researchers astray.
A new NBER working paper explores this issue in detail. Kasey Buckles and Daniel Hungerman consider the case of season of birth, which has been shown to be associated with a variety of health and socioeconomic outcomes later in life. These associations have been attributed to fetal exposure to different weather conditions or differential exposure to arbitrary age-cutoffs (in sports or in schooling). At first glance, season of birth appears to be a great source of exogenous variation for a slew of different causal questions: after all, individuals don't have any control over when they are born and it seems like something that would be left to chance. However, Buckles and Hungerman convincingly argue that this is not the case:
In this paper we consider a new explanation: that children born at different times in the year are conceived by women with different socioeconomic characteristics. We document large seasonal changes in the characteristics of women giving birth throughout the year in the United States. Children born in the winter are disproportionally born to women who are more likely to be teenagers and less likely to be married or have a high school degree. We show that controls for family background characteristics can explain up to half of the relationship between season of birth and adult outcomes. We then discuss the implications of this result for using season of birth as an instrumental variable; our findings suggest that, though popular, season-of-birth instruments may produce inconsistent estimates. Finally, we find that some of the seasonality in maternal characteristics is due to summer weather differentially affecting fertility patterns across socioeconomic groups.
This is a neat paper and serves as a good warning to those interested in finding natural variation to identify causal effects. Another excellent paper on the same subject, by Mark Rosenzweig and Ken Wolpin, goes through a variety of other potentially fallacious natural experiment examples and is a must read for anyone doing empirical work.
Thursday, December 11, 2008
Bad News on the Other Side of the Birth Weight Distribution?
Two recent articles explore the long-run effects of maternal weight gain and/or high weight at birth on long-run outcomes. Much of the health economics literature has focused on understanding returns to birth weight at the lower tail of the distribution, and several studies have found notable effects of birth weight on wages, schooling and cognition later in life. The downsides to high birth weight have been explored in the medical literature, which has typically focused on obesity and diabetes risk among children born to diabetic mothers (who birth larger than average - or macrosomic - infants).
A recent study in Obstetrics and Gynecology illustrates that increased gestational weight gain among mothers is associated with higher risk of obesity among their offspring. Unlike past studies, the study sample here is not limited to mothers with diabetes. While causality is difficult to establish here - mothers that gain weight during pregnancy may have similar preferences/constraints regarding food, which can be passed on to the next generation via other means besides biology - the results are interesting and deserve further research attention.
In a recent NBER working paper, Resul Cesar and Inas Rashad look at the association between birth weight and cognitive outcomes among a sample of children and teens and young adults followed in two different panel studies. The main result is that birth weight is associated with lower cognitive test scores at both the low and high part of the distribution: heavy infants suffer deficits, too.
I'm guessing this literature will likely explode in the next few years, especially with the ever growing obesity "epidemic" and its obvious consequences for maternal weight during pregnancy (see here for an earlier piece on this). Hopefully, this literature will address two major areas:
(1) Causal effects - Spatial variation in food prices, dynamics from recessions, the introduction of public programs and other shocks can be used to identify maternal weight gain during gestation and child birth weight. Utilizing this variation can help eliminate from the picture alternative interpretations of the association between maternal weight gain, child birth weight and later child outcomes.
(2) Mechanisms - Even if you can recover causal effects, what is the pathway? A link between high birth weight and lower cognition could be biological in nature or may operate through reduced and less effective investments in childhood and adolescents due to diminishing social returns to body weight: kids may learn less effectively if they are being made fun of at school for their appearance. These explanations have vastly different policy implications and should be teased apart.
A recent study in Obstetrics and Gynecology illustrates that increased gestational weight gain among mothers is associated with higher risk of obesity among their offspring. Unlike past studies, the study sample here is not limited to mothers with diabetes. While causality is difficult to establish here - mothers that gain weight during pregnancy may have similar preferences/constraints regarding food, which can be passed on to the next generation via other means besides biology - the results are interesting and deserve further research attention.
In a recent NBER working paper, Resul Cesar and Inas Rashad look at the association between birth weight and cognitive outcomes among a sample of children and teens and young adults followed in two different panel studies. The main result is that birth weight is associated with lower cognitive test scores at both the low and high part of the distribution: heavy infants suffer deficits, too.
I'm guessing this literature will likely explode in the next few years, especially with the ever growing obesity "epidemic" and its obvious consequences for maternal weight during pregnancy (see here for an earlier piece on this). Hopefully, this literature will address two major areas:
(1) Causal effects - Spatial variation in food prices, dynamics from recessions, the introduction of public programs and other shocks can be used to identify maternal weight gain during gestation and child birth weight. Utilizing this variation can help eliminate from the picture alternative interpretations of the association between maternal weight gain, child birth weight and later child outcomes.
(2) Mechanisms - Even if you can recover causal effects, what is the pathway? A link between high birth weight and lower cognition could be biological in nature or may operate through reduced and less effective investments in childhood and adolescents due to diminishing social returns to body weight: kids may learn less effectively if they are being made fun of at school for their appearance. These explanations have vastly different policy implications and should be teased apart.
Tuesday, December 9, 2008
Returns to Medical Care Among High Risk Infants
A really interesting NBER paper this week looks at health returns to medical interventions among high-risk (here, low birth weight) newborns babies. The difficulty in assessing the casual effects of medical care in this population is that worse-off infants may get more of it because they require it. On the other hand, unhealthy babies may be more likely to come from poor families, who lack access to health care. Either process makes it difficult to recover causal effects.
Douglas Almond, Joseph Doyle, Amanda E. Kowalski and Heidi Williams adopt an interesting strategy to get around this issue. Essentially, they utilize existing birth weight thresholds (those below 1500 grams are classified very low birth weight) and provider obedience to these discontinuous (and perhaps arbitrary?) thresholds. By comparing those babies just below 1500 grams to those just above it, the authors contend that they get around the targetting of medical care to worse off babies - after all, the difference of a few grams around the cutoff is likely random and unrelated to innate biological hardiness. In a sense, the level of treatment given is essentially random for babies born around this narrow threshold. More on their methodology (which Almond and Doyle employ in another very interesting paper) and results:
We estimate marginal returns to medical care for at-risk newborns by comparing health outcomes and medical treatment provision on either side of common risk classifications, most notably the "very low birth weight" threshold at 1500 grams. First, using data on the census of US births in available years from 1983-2002, we find evidence that newborns with birth weights just below 1500 grams have lower one-year mortality rates than do newborns with birth weights just above this cutoff, even though mortality risk tends to decrease with birth weight. One-year mortality falls by approximately one percentage point as birth weight crosses 1500 grams from above, which is large relative to mean one-year mortality of 5.5% just above 1500 grams. Second, using hospital discharge records for births in five states in available years from 1991-2006, we find evidence that newborns with birth weights just below 1500 grams have discontinuously higher costs and frequencies of specific medical inputs. We estimate a $4,000 increase in hospital costs as birth weight approaches 1500 grams from above, relative to mean hospital costs of $40,000 just above 1500 grams. Taken together, these estimates suggest that the cost of saving a statistical life of a newborn with birth weight near 1500 grams is on the order of $550,000 in 2006 dollars.
Aside from the policy relevance of the results and the innovative research design used in the study, I find it really interesting that the provision of health care is so sensitive to seemingly arbitrary guidelines. Does anyone have a sense of where the 1500 and 2500 gram cutoffs came from, and whether they've outlived their clinical relevance?
Douglas Almond, Joseph Doyle, Amanda E. Kowalski and Heidi Williams adopt an interesting strategy to get around this issue. Essentially, they utilize existing birth weight thresholds (those below 1500 grams are classified very low birth weight) and provider obedience to these discontinuous (and perhaps arbitrary?) thresholds. By comparing those babies just below 1500 grams to those just above it, the authors contend that they get around the targetting of medical care to worse off babies - after all, the difference of a few grams around the cutoff is likely random and unrelated to innate biological hardiness. In a sense, the level of treatment given is essentially random for babies born around this narrow threshold. More on their methodology (which Almond and Doyle employ in another very interesting paper) and results:
We estimate marginal returns to medical care for at-risk newborns by comparing health outcomes and medical treatment provision on either side of common risk classifications, most notably the "very low birth weight" threshold at 1500 grams. First, using data on the census of US births in available years from 1983-2002, we find evidence that newborns with birth weights just below 1500 grams have lower one-year mortality rates than do newborns with birth weights just above this cutoff, even though mortality risk tends to decrease with birth weight. One-year mortality falls by approximately one percentage point as birth weight crosses 1500 grams from above, which is large relative to mean one-year mortality of 5.5% just above 1500 grams. Second, using hospital discharge records for births in five states in available years from 1991-2006, we find evidence that newborns with birth weights just below 1500 grams have discontinuously higher costs and frequencies of specific medical inputs. We estimate a $4,000 increase in hospital costs as birth weight approaches 1500 grams from above, relative to mean hospital costs of $40,000 just above 1500 grams. Taken together, these estimates suggest that the cost of saving a statistical life of a newborn with birth weight near 1500 grams is on the order of $550,000 in 2006 dollars.
Aside from the policy relevance of the results and the innovative research design used in the study, I find it really interesting that the provision of health care is so sensitive to seemingly arbitrary guidelines. Does anyone have a sense of where the 1500 and 2500 gram cutoffs came from, and whether they've outlived their clinical relevance?
Subscribe to:
Posts (Atom)