A new and important paper by Anne Case and Christina Paxson finds the following:
We document the impact of the AIDS crisis on non-AIDS related health services in fourteen sub-Saharan African countries. Using multiple waves of Demographic and Health Surveys (DHS) for each country, we examine antenatal care, birth deliveries, and rates of immunization for children born between 1988 and 2005. We find deterioration in nearly all of these dimensions of health care over this period. The most recent DHS survey for each country collected data on HIV prevalence, which allows us to examine the association between HIV burden and health care. We find that erosion of health services is highly correlated with increases in AIDS prevalence. Regions of countries that have light AIDS burdens have witnessed small or no declines in health care, using the measures noted above, while those regions currently shouldering the heaviest burdens have seen the largest erosion in treatment for pregnant women and children. Using semi-parametric techniques, we can date the beginning of the divergence in health services between high and low HIV regions to the mid-1990s.
Case and Paxson are unable to pin down a mechanism for why this is happening. They suggest it is not driven by an erosion of wealth (though the data they use is somewhat lacking in measures of health beyond asset ownership) or reduced demand for medical care by HIV+ mothers. On the other hand, they cannot rule out adverse impacts of HIV/AIDS on the supply of health care workers and/or the diversion of resources to those with HIV/AIDS, perhaps at the expense of other aspects of medical care. As such, the authors rightly point out that there is more work to be done and that this work needs to be done very soon.
I'll have more to say about this in a forthcoming post.
Monday, May 25, 2009
Friday, May 22, 2009
The Human Development Index
Justin Wolfers put up an interesting post on the Freakonomics blog today about the Human Development Index (HDI), a summary statistic that combines information on life expectancy, schooling outcomes and income per capita. He cites a post by Andrew Gelman, which makes the claim that the Human Development Index for the 50 U.S. states provides little information above and beyond what state income per capita tells you.
Some thoughts:
1) The HDI was first developed in 1990 as a good faith effort to move beyond income as a measure of development. The idea was to capture other aspects of society and wellness that we might think are important in making people content and happy. The fact that there is little information beyond what a simpler measure of income per capita tells you is discouraging on the one hand, as it might imply that we aren't measuring non-monetary aspects of development well. On the other hand, the finding might suggest that wealth creation brings about the other aspects of development we care about, even if these other aspects are important ends in and of themselves.
2) It is important to ask what the content of the HDI is, how it is calculated, and its utility in informing policy. On this note, I'm a bit surprised that neither Wolfers or Gelman cite a seminal paper by my mentor from Duke, Allen Kelley. Written in 1991, Kelley notes the close correspondence between national income per capita and HDI, but also goes into the sausage factory of how the HDI is constructed (its a bit arbitrary) and what the statistic may or may not be able to tell us. The subtitle of his piece, "handle with care," gives you a good sense of his skepticism. The fact that the same skepticism remains warranted nearly 20 years later is more than a bit disturbing.
3) Finally, the blog fivethirtyeight.com is fantastic: the authors use data and elegant statistics to delve deeper into various political and social issues that are often taken for granted. It was first recommended to me by Joachim Hero, and I am now linking it in the sidebar for your enjoyment.
Some thoughts:
1) The HDI was first developed in 1990 as a good faith effort to move beyond income as a measure of development. The idea was to capture other aspects of society and wellness that we might think are important in making people content and happy. The fact that there is little information beyond what a simpler measure of income per capita tells you is discouraging on the one hand, as it might imply that we aren't measuring non-monetary aspects of development well. On the other hand, the finding might suggest that wealth creation brings about the other aspects of development we care about, even if these other aspects are important ends in and of themselves.
2) It is important to ask what the content of the HDI is, how it is calculated, and its utility in informing policy. On this note, I'm a bit surprised that neither Wolfers or Gelman cite a seminal paper by my mentor from Duke, Allen Kelley. Written in 1991, Kelley notes the close correspondence between national income per capita and HDI, but also goes into the sausage factory of how the HDI is constructed (its a bit arbitrary) and what the statistic may or may not be able to tell us. The subtitle of his piece, "handle with care," gives you a good sense of his skepticism. The fact that the same skepticism remains warranted nearly 20 years later is more than a bit disturbing.
3) Finally, the blog fivethirtyeight.com is fantastic: the authors use data and elegant statistics to delve deeper into various political and social issues that are often taken for granted. It was first recommended to me by Joachim Hero, and I am now linking it in the sidebar for your enjoyment.
Thursday, May 21, 2009
Links: Obama on Global Health, Kindergarten, and the Indian Marriage Market
1. Karen Grepin provides a nice discussion on Obama's comprehensive strategy on global health. While you are there, check out her interesting piece on advocacy and neglect in the global health arena.
2. In an interesting new paper, Elizabeth Cascio finds that the introduction of kindergarten programs in the 1960s and 70s led to reduced drop out and institutionalization rates among whites but not blacks. The differential effect, she posits, might have something to do with crowding out federally funded programs helping the poorest black five year olds. I'm not so convinced about this as the mechanism, and I think there is a study waiting to happen that looks at health effects, as well.
3. A fun new paper by Abhijit Banerjee and co-authors looks at the Indian marriage market. From the abstract:
This paper studies the role played by caste, education and other social and economic attributes in arranged marriages among middle-class Indians. We use a unique data set on individuals who placed matrimonial advertisements in a major newspaper, the responses they received, how they ranked them, and the eventual matches. We estimate the preferences for caste, education, beauty, and other attributes. We then compute a set of stable matches, which we compare to the actual matches that we observe in the data. We find the stable matches to be quite similar to the actual matches, suggesting a relatively frictionless marriage market. One of our key empirical findings is that there is a very strong preference for within-caste marriage. However, because both sides of the market share this preference and because the groups are fairly homogeneous in terms of the distribution of other attributes, in equilibrium, the cost of wanting to marry within-caste is low. This allows caste to remain a persistent feature of the Indian marriage market.
2. In an interesting new paper, Elizabeth Cascio finds that the introduction of kindergarten programs in the 1960s and 70s led to reduced drop out and institutionalization rates among whites but not blacks. The differential effect, she posits, might have something to do with crowding out federally funded programs helping the poorest black five year olds. I'm not so convinced about this as the mechanism, and I think there is a study waiting to happen that looks at health effects, as well.
3. A fun new paper by Abhijit Banerjee and co-authors looks at the Indian marriage market. From the abstract:
This paper studies the role played by caste, education and other social and economic attributes in arranged marriages among middle-class Indians. We use a unique data set on individuals who placed matrimonial advertisements in a major newspaper, the responses they received, how they ranked them, and the eventual matches. We estimate the preferences for caste, education, beauty, and other attributes. We then compute a set of stable matches, which we compare to the actual matches that we observe in the data. We find the stable matches to be quite similar to the actual matches, suggesting a relatively frictionless marriage market. One of our key empirical findings is that there is a very strong preference for within-caste marriage. However, because both sides of the market share this preference and because the groups are fairly homogeneous in terms of the distribution of other attributes, in equilibrium, the cost of wanting to marry within-caste is low. This allows caste to remain a persistent feature of the Indian marriage market.
Wednesday, May 20, 2009
Imports and Development
It's always fun when a friend or colleague of yours is mentioned in The Economist. Last week was no exception, when the venerable periodical covered a recent piece on trade co-authored by Amit Khandelwal, a former Yale economics graduate student now at the Columbia Graduate School of Business as an assistant professor.
The research in question looks at the effects of imports on aspects of economic development. As the Economist piece points out, in policy discussions on trade, there is this belief that exporting is good for the home country's development, but importing is not. Khandelwal et al's piece shows that, in the case of India, imports have had some positive benefits. Quoting from the news article:
As part of those reforms, India slashed tariffs on imports from an average of 90% in 1991 to 30% in 1997. Not surprisingly, imports doubled in value over this period. But the effects on Indian manufacturing were not what the prophets of doom had predicted: output grew by over 50% in that time. And by looking carefully at what was imported and what it was used to make, the researchers found that cheaper and more accessible imports gave a big boost to India’s domestic industrial growth in the 1990s.
This was because the tariff cuts meant more than Indian consumers being able to satisfy their cravings for imported chocolate (though they did that, too). It gave Indian manufacturers access to a variety of intermediate and capital goods which had earlier been too expensive. The rise in imports of intermediate goods was much higher, at 227%, than the 90% growth in consumer-goods imports in the 13 years to 2000.
Good stuff.
The research in question looks at the effects of imports on aspects of economic development. As the Economist piece points out, in policy discussions on trade, there is this belief that exporting is good for the home country's development, but importing is not. Khandelwal et al's piece shows that, in the case of India, imports have had some positive benefits. Quoting from the news article:
As part of those reforms, India slashed tariffs on imports from an average of 90% in 1991 to 30% in 1997. Not surprisingly, imports doubled in value over this period. But the effects on Indian manufacturing were not what the prophets of doom had predicted: output grew by over 50% in that time. And by looking carefully at what was imported and what it was used to make, the researchers found that cheaper and more accessible imports gave a big boost to India’s domestic industrial growth in the 1990s.
This was because the tariff cuts meant more than Indian consumers being able to satisfy their cravings for imported chocolate (though they did that, too). It gave Indian manufacturers access to a variety of intermediate and capital goods which had earlier been too expensive. The rise in imports of intermediate goods was much higher, at 227%, than the 90% growth in consumer-goods imports in the 13 years to 2000.
Good stuff.
Saturday, May 9, 2009
Antidepressants and Suicide
Whether antidepressant use increase suicide risk in the short-term is an ongoing debate in the clinical medicine and health policy worlds. A few years back, based on some evidence that antidepressant use was correlated with a higher risk of suicide, the FDA issued a "black box" warning, forcing manufacturers to acknowledge the increased risks on packaging and materials related to the drugs. The public responded predictably: antidepressant use dropped notably after the warning. (See this 2006 article for more on the issue)
The biomedical model that links antidepressant use to suicide is the following. Depressive symptoms involve both mood and reduced activity. Antidepressants, it is thought, start working by increasing activation before mood. As a result, the hypothesis is that, in the short term, people who have suicidal thoughts may actually carry it out because they are now "activated."
But is there another explanation that could explain the link between anti-depressants and suicide? An important possibility is selection: anti-depressants are taken by people with depressive symptoms, who are more likely to commit suicide. The fact that the association between anti-depressant use and suicide only exists in the short-run could be explained by this selection model as well: those who would commit suicide would do so, and those who are left may have been unlikely to do so in the first place or were prevented from doing so by the medication.
The overall literature on anti-depressants and suicide gives some support to the selection hypothesis. First off, the relationship between use and suicide seems to vary from study to study and across countries. We would not expect this if the biological model were correct. Second, the "black box" warning provides an interesting time series test. In several countries, the use of anti-depressants dropped after the public was informed about the potential risks, and the incidence of suicides actually increased. This runs counter to what we would expect from the biological mechanism model.
A recent paper (forthcoming in the Journal of Health Economics) provides what I think is the most careful analysis of the causal relationship between anti-depressant use and suicide, taking explicitly into account the potential selection bias issue. The authors, Jens Ludwig, Dave Marcotte and Karen Norberg, utilize an instrumental variables (IV) approach:
In this paper we present what we believe to be the first estimates for the effects of SSRIs on suicide using both a plausibly exogenous source of identifying variation and adequate statistical power to detect effects on mortality that are much smaller than anything that could be detected from randomized trials. We construct a panel dataset with suicide rates and SSRI sales per capita for 26 countries for up to 25 years. Since SSRI sales may be endogenous, we exploit institutional differences across countries that affect how they regulate, price, distribute and use prescription drugs in general (Berndt et al., 2007). Since we do not have direct measures for these institutional characteristics for all countries, we use data on drug diffusion rates as a proxy. We show that sales growth for SSRIs is strongly related to the rate of sales growth of the other major new drugs that were introduced in the 1980s for the treatment of non-psychiatric health conditions. This source of variation in SSRI sales helps overcome the problem of reverse causation and many of the most obvious omitted-variables concerns with past studies. Our research design may also have broader applications for the study of how other drug classes affect different health outcomes.
Using this strategy, they find that a 12% increase in anti-depressant sales is associated with a 5% decrease in suicides. Interesting stuff.
While the main innovation in the paper is the use of instrumental variables, this may also the main weakness. First, as discussed in previous posts, in order for the IV approach to work, the instruments should only affect the outcome through the exposure of interest. The authors in this paper go through some trouble to establish the validity of their IVs. Its all carefully done and compelling, but, depending on your priors about institutional differences in pricing strategies, you may still have qualms about the IV.
The other issue with IVs, is that the effect it computes applies to those people (or here, groups of people) that are most affected or sensitive by the instrument (see this earlier post for more on this). Thus, it is very important to note that the finding in this paper does not rule out the possibility that anti-depressant use might have adverse impacts on some populations. I think this is of particular interest to clinicians, and there are new methods in econometrics that can help uncover heterogeneity in treatment effects (see this paper on the heterogeneous impacts of treatment on breast cancer, utilizing methods developed by Heckman and co-authors).
The biomedical model that links antidepressant use to suicide is the following. Depressive symptoms involve both mood and reduced activity. Antidepressants, it is thought, start working by increasing activation before mood. As a result, the hypothesis is that, in the short term, people who have suicidal thoughts may actually carry it out because they are now "activated."
But is there another explanation that could explain the link between anti-depressants and suicide? An important possibility is selection: anti-depressants are taken by people with depressive symptoms, who are more likely to commit suicide. The fact that the association between anti-depressant use and suicide only exists in the short-run could be explained by this selection model as well: those who would commit suicide would do so, and those who are left may have been unlikely to do so in the first place or were prevented from doing so by the medication.
The overall literature on anti-depressants and suicide gives some support to the selection hypothesis. First off, the relationship between use and suicide seems to vary from study to study and across countries. We would not expect this if the biological model were correct. Second, the "black box" warning provides an interesting time series test. In several countries, the use of anti-depressants dropped after the public was informed about the potential risks, and the incidence of suicides actually increased. This runs counter to what we would expect from the biological mechanism model.
A recent paper (forthcoming in the Journal of Health Economics) provides what I think is the most careful analysis of the causal relationship between anti-depressant use and suicide, taking explicitly into account the potential selection bias issue. The authors, Jens Ludwig, Dave Marcotte and Karen Norberg, utilize an instrumental variables (IV) approach:
In this paper we present what we believe to be the first estimates for the effects of SSRIs on suicide using both a plausibly exogenous source of identifying variation and adequate statistical power to detect effects on mortality that are much smaller than anything that could be detected from randomized trials. We construct a panel dataset with suicide rates and SSRI sales per capita for 26 countries for up to 25 years. Since SSRI sales may be endogenous, we exploit institutional differences across countries that affect how they regulate, price, distribute and use prescription drugs in general (Berndt et al., 2007). Since we do not have direct measures for these institutional characteristics for all countries, we use data on drug diffusion rates as a proxy. We show that sales growth for SSRIs is strongly related to the rate of sales growth of the other major new drugs that were introduced in the 1980s for the treatment of non-psychiatric health conditions. This source of variation in SSRI sales helps overcome the problem of reverse causation and many of the most obvious omitted-variables concerns with past studies. Our research design may also have broader applications for the study of how other drug classes affect different health outcomes.
Using this strategy, they find that a 12% increase in anti-depressant sales is associated with a 5% decrease in suicides. Interesting stuff.
While the main innovation in the paper is the use of instrumental variables, this may also the main weakness. First, as discussed in previous posts, in order for the IV approach to work, the instruments should only affect the outcome through the exposure of interest. The authors in this paper go through some trouble to establish the validity of their IVs. Its all carefully done and compelling, but, depending on your priors about institutional differences in pricing strategies, you may still have qualms about the IV.
The other issue with IVs, is that the effect it computes applies to those people (or here, groups of people) that are most affected or sensitive by the instrument (see this earlier post for more on this). Thus, it is very important to note that the finding in this paper does not rule out the possibility that anti-depressant use might have adverse impacts on some populations. I think this is of particular interest to clinicians, and there are new methods in econometrics that can help uncover heterogeneity in treatment effects (see this paper on the heterogeneous impacts of treatment on breast cancer, utilizing methods developed by Heckman and co-authors).
Tuesday, May 5, 2009
Staying Off Facebook with Help from Behavioral Economics
A few days ago, I got pretty annoyed with some stuff on Facebook and wanted to stay away from it for a while. Unfortunately, logging into Facebook is too easy. Actually, somehow I'm always signed on on my laptop and blackberry. That, coupled with the fact that I addictively check my newsfeed and spend a lot of time in front of my computer, made it difficult to stay off of the site for long.
So, I decided to take a more "drastic" step: deactivation. While I guess you can never really leave Facebook, you can take your profile offline. Deactivation means that others cannot find you, message you or whatever.
Since I've deactivated, I haven't been on Facebook for a couple days and things are just great. But, at first blush, it might be a mystery to some as to why deactivation would work. After all, to reactivate, I'd just have to log back into the website (yeah, it's just that easy). So how could this have any effect on my Facebook behavior? I think there are a few reasons:
(1) I like marginal costs that are essentially zero. Deactivation raises the marginal cost of going on Facebook just a little bit, which might work to totally devalue what in my head should be a free experience (see here for a good discussion of this phenomenon in another context).
(2) Deactivation works for me as a self/pre-commitment device (see here for a broader discussion). I realized I'd feel a lot worse repeatedly deactivating and reactivating rather than just navigating from Facebook to another page and back. In the former case, I'd feel like more of a flake or diva for signing off a service and going back on, whereas that kind of behavior is more easily justified when you are already part of the service.
So behavioral economics has helped me get around my Facebook conundrum. Interestingly, various behavioral economics inspired "nudges" almost stopped me from establishing my pre-commitment device. When you go to deactivate, you are shown pictures of your five of your friends (from your jointly tagged pictures) with captions like "Mike will miss you," all below the question "Are you sure you want to deactivate?" Furthermore, below the pictures, you are asked to provide a reason for why you want to deactivate, and for all of the choices except "This is temporary. I'll be back" Facebook gives you a pithy statement about why you might want to reconsider.
Finally, see you on Facebook...at some point in the future.
So, I decided to take a more "drastic" step: deactivation. While I guess you can never really leave Facebook, you can take your profile offline. Deactivation means that others cannot find you, message you or whatever.
Since I've deactivated, I haven't been on Facebook for a couple days and things are just great. But, at first blush, it might be a mystery to some as to why deactivation would work. After all, to reactivate, I'd just have to log back into the website (yeah, it's just that easy). So how could this have any effect on my Facebook behavior? I think there are a few reasons:
(1) I like marginal costs that are essentially zero. Deactivation raises the marginal cost of going on Facebook just a little bit, which might work to totally devalue what in my head should be a free experience (see here for a good discussion of this phenomenon in another context).
(2) Deactivation works for me as a self/pre-commitment device (see here for a broader discussion). I realized I'd feel a lot worse repeatedly deactivating and reactivating rather than just navigating from Facebook to another page and back. In the former case, I'd feel like more of a flake or diva for signing off a service and going back on, whereas that kind of behavior is more easily justified when you are already part of the service.
So behavioral economics has helped me get around my Facebook conundrum. Interestingly, various behavioral economics inspired "nudges" almost stopped me from establishing my pre-commitment device. When you go to deactivate, you are shown pictures of your five of your friends (from your jointly tagged pictures) with captions like "Mike will miss you," all below the question "Are you sure you want to deactivate?" Furthermore, below the pictures, you are asked to provide a reason for why you want to deactivate, and for all of the choices except "This is temporary. I'll be back" Facebook gives you a pithy statement about why you might want to reconsider.
Finally, see you on Facebook...at some point in the future.
Monday, May 4, 2009
The Cost of Political Opposition
Dissent is an important part of public discourse in any setting. In a truly democratic regime, one would expect dissent to carry little cost (though I expect it might in hard to observe ways). But what about in autocratic regimes? What is the price of opposing the ruling party?
In a recent working paper, Chang-Tai Hsieh, Edward Miguel, Daniel Ortega and Francisco Rodriguez try to address this question in the context of the Hugo Chavez led Venezuela. In their own words:
In 2004, the Chávez regime in Venezuela distributed the list of several million voters whom had attempted to remove him from office throughout the government bureaucracy, allegedly to identify and punish these voters. We match the list of petition signers distributed by the government to household survey respondents to measure the economic effects of being identified as a Chavez political opponent. We find that voters who were identified as Chavez opponents experienced a 5 percent drop in earnings and a 1.5 percentage point drop in employment rates after the voter list was released. A back-of-the-envelope calculation suggests that the loss aggregate TFP from the misallocation of workers across jobs was substantial, on the order of 3 percent of GDP.
That political opposition in an autocratic regime can invite economic retribution is not that surprising, but the 3% of GDP number kind of is. It's just a great illustration of how the incentives of the public and autocratic leaders are not aligned: one would hope that a 3% loss of GDP would have dissuaded Chavez from going after his opposition.
All in all, a really interesting, if not very sad, read.
(Ed: Marginal Revolution has an interesting take on this paper, as well)
In a recent working paper, Chang-Tai Hsieh, Edward Miguel, Daniel Ortega and Francisco Rodriguez try to address this question in the context of the Hugo Chavez led Venezuela. In their own words:
In 2004, the Chávez regime in Venezuela distributed the list of several million voters whom had attempted to remove him from office throughout the government bureaucracy, allegedly to identify and punish these voters. We match the list of petition signers distributed by the government to household survey respondents to measure the economic effects of being identified as a Chavez political opponent. We find that voters who were identified as Chavez opponents experienced a 5 percent drop in earnings and a 1.5 percentage point drop in employment rates after the voter list was released. A back-of-the-envelope calculation suggests that the loss aggregate TFP from the misallocation of workers across jobs was substantial, on the order of 3 percent of GDP.
That political opposition in an autocratic regime can invite economic retribution is not that surprising, but the 3% of GDP number kind of is. It's just a great illustration of how the incentives of the public and autocratic leaders are not aligned: one would hope that a 3% loss of GDP would have dissuaded Chavez from going after his opposition.
All in all, a really interesting, if not very sad, read.
(Ed: Marginal Revolution has an interesting take on this paper, as well)
Wednesday, April 29, 2009
Shaan's Blog and Swine Flu
My friend, former tennis partner and former Yale MPHer Shaan Chatturvedi has just started blogging about his experiences in Guyana, where he is currently a working for the CDC Global AIDS Program. His most recent post, on the swine flu outbreak, is fantastic and promises of good things to come from his blog. Do check it out!
For more on the swine flu, check out this interesting article by Dr. Carlos del Rio, the chair of the Global Health Department of the Emory School of Public Health. There is a lot of interesting stuff in there about different control measures and the reasons why swine flu mortality might be higher in Mexico than in the US.
For more on the swine flu, check out this interesting article by Dr. Carlos del Rio, the chair of the Global Health Department of the Emory School of Public Health. There is a lot of interesting stuff in there about different control measures and the reasons why swine flu mortality might be higher in Mexico than in the US.
Tuesday, April 21, 2009
Long-Run and Intergenerational Effects of Early Childhood Environments
Great NBER working paper this week on the long-run and next generation returns to early life conditions. Specifically, Eric Gould and co-authors look at consequences driven by the airlift of Yemenite immigrants into Israel. In their own words:
This paper estimates the effect of the childhood environment on a large array of social and economic outcomes lasting almost 60 years, for both the affected cohorts and for their children. To do this, we exploit a natural experiment provided by the 1949 Magic Carpet operation, where over 50,000 Yemenite immigrants were airlifted to Israel. The Yemenites, who lacked any formal schooling or knowledge of a western-style culture or bureaucracy, believed that they were being "redeemed," and put their trust in the Israeli authorities to make decisions about where they should go and what they should do. As a result, they were scattered across the country in essentially a random fashion, and as we show, the environmental conditions faced by immigrant children were not correlated with other factors that affected the long-term outcomes of individuals. We construct three summary measures of the childhood environment: 1) whether the home had running water, sanitation and electricity; 2) whether the locality of residence was in an urban environment with a good economic infrastructure; and 3) whether the locality of residence was a Yemenite enclave. We find that children who were placed in a good environment (a home with good sanitary conditions, in a city, and outside of an ethnic enclave) were more likely to achieve positive long-term outcomes. They were more likely to obtain higher education, marry at an older age, have fewer children, work at age 55, be more assimilated into Israeli society, be less religious, and have more worldly tastes in music and food. These effects are much more pronounced for women than for men. We find weaker and somewhat mixed effects on health outcomes, and no effect on political views. We do find an effect on the next generation – children who lived in a better environment grew up to have children who achieved higher educational attainment.
I find this paper noteworthy for several reasons:
(1) The authors have a credible and interesting source of variation, and the actual early life exposures they look at have immediate policy implications
(2) The authors explore a wide variety of different outcomes, including behavioral aspects. In my dissertation, I argued that long-run returns need to be taken into account when making resource allocation decisions. However, this is difficult if only a subset of long-run returns are known. This paper really hits this gap.
(3) Finally, and more self-serving, the long-run effects of sanitation and clean water jive well with my thesis paper on the National Clean Water Program in Mexico (see the next post).
This paper estimates the effect of the childhood environment on a large array of social and economic outcomes lasting almost 60 years, for both the affected cohorts and for their children. To do this, we exploit a natural experiment provided by the 1949 Magic Carpet operation, where over 50,000 Yemenite immigrants were airlifted to Israel. The Yemenites, who lacked any formal schooling or knowledge of a western-style culture or bureaucracy, believed that they were being "redeemed," and put their trust in the Israeli authorities to make decisions about where they should go and what they should do. As a result, they were scattered across the country in essentially a random fashion, and as we show, the environmental conditions faced by immigrant children were not correlated with other factors that affected the long-term outcomes of individuals. We construct three summary measures of the childhood environment: 1) whether the home had running water, sanitation and electricity; 2) whether the locality of residence was in an urban environment with a good economic infrastructure; and 3) whether the locality of residence was a Yemenite enclave. We find that children who were placed in a good environment (a home with good sanitary conditions, in a city, and outside of an ethnic enclave) were more likely to achieve positive long-term outcomes. They were more likely to obtain higher education, marry at an older age, have fewer children, work at age 55, be more assimilated into Israeli society, be less religious, and have more worldly tastes in music and food. These effects are much more pronounced for women than for men. We find weaker and somewhat mixed effects on health outcomes, and no effect on political views. We do find an effect on the next generation – children who lived in a better environment grew up to have children who achieved higher educational attainment.
I find this paper noteworthy for several reasons:
(1) The authors have a credible and interesting source of variation, and the actual early life exposures they look at have immediate policy implications
(2) The authors explore a wide variety of different outcomes, including behavioral aspects. In my dissertation, I argued that long-run returns need to be taken into account when making resource allocation decisions. However, this is difficult if only a subset of long-run returns are known. This paper really hits this gap.
(3) Finally, and more self-serving, the long-run effects of sanitation and clean water jive well with my thesis paper on the National Clean Water Program in Mexico (see the next post).
Friday, April 3, 2009
Correlation Does Not Imply Causation. And...?
Since I started grad school four years ago, I've noticed that the general public is much more attuned to idea of correlation not always implying causation. Of course, the indoctrination is not complete just yet, and there are plenty of instances where an association is mistaken for something more, but the fact that people are becoming better consumers of statistics is gratifying. I attribute this to the spate of popular press economics and statistics books/blogs in the last few years (though I might be in danger of confusing correlation and causation myself by saying this!)
The standards in empirical research reflect how seriously people are taking this motto: finding a clever instrumental variable or even experimental variation is no longer good enough. Papers without extensive "robustness" checks and falsification tests have less credibility now than they would have even five years ago. This, like the trend in the general public, is a good development.
However, with these positives come some more troubling tendencies. Specifically, I have a beef with the overuse of the causation-correlation dictum. Now, anybody can bring down a paper simply by saying "correlation does not imply causation" without having to provide a reason why this might be the case. For example, I am working on a paper looking at the long-run causal effects of birth year exposure to a clean water and sanitation efforts (I'll post a link to this paper in a month or so when a good draft is ready). I have a plausible identification strategy, and also include all sorts of controls, trends and falsification checks in my analysis to further establish causality. My results check out.
However, someone recently remarked told me that I should be concerned about omitted variables. When I pressed her on what these might be, she wasn't sure but commented that "there are always omitted factors."
Clearly, this isn't helpful. It's really easy to look/sound clever and point out that correlation does not imply causation: it is technically a true statement! But I think people who make this claim should talk about how it applies to the analysis at hand (i.e., have some kind of model or story that makes more explicit the nature of the potential biases and where they come from). Otherwise, the statement by itself is pretty uninformative and does little to advance our knowledge.
The standards in empirical research reflect how seriously people are taking this motto: finding a clever instrumental variable or even experimental variation is no longer good enough. Papers without extensive "robustness" checks and falsification tests have less credibility now than they would have even five years ago. This, like the trend in the general public, is a good development.
However, with these positives come some more troubling tendencies. Specifically, I have a beef with the overuse of the causation-correlation dictum. Now, anybody can bring down a paper simply by saying "correlation does not imply causation" without having to provide a reason why this might be the case. For example, I am working on a paper looking at the long-run causal effects of birth year exposure to a clean water and sanitation efforts (I'll post a link to this paper in a month or so when a good draft is ready). I have a plausible identification strategy, and also include all sorts of controls, trends and falsification checks in my analysis to further establish causality. My results check out.
However, someone recently remarked told me that I should be concerned about omitted variables. When I pressed her on what these might be, she wasn't sure but commented that "there are always omitted factors."
Clearly, this isn't helpful. It's really easy to look/sound clever and point out that correlation does not imply causation: it is technically a true statement! But I think people who make this claim should talk about how it applies to the analysis at hand (i.e., have some kind of model or story that makes more explicit the nature of the potential biases and where they come from). Otherwise, the statement by itself is pretty uninformative and does little to advance our knowledge.
Subscribe to:
Posts (Atom)