ekfletch

Betsey Stevenson and Justin Wolfers on why we study families

I’m often asked why my research is economics and not sociology. Justin Wolfers and Betsey Stevenson give one answer as part of a longer Q&A on their research:

Your other areas of research focus include marriage, divorce, and family. Why would these areas interest economists? Or business leaders?

Dr. Stevenson: Economics is about how people make decisions optimally, given that they’re facing constraints. That framework can be applied anywhere, not just to things that are about dollars and cents and the economy. Families and labor markets are intimately connected, and to understand one, it’s helpful to understand the other. That’s because decisions about labor force participation and about what kinds of jobs to take and what kind of hours to keep are made within the context of family lives. What happens in families affects the way people make those kinds of decisions. And what happens in labor markets affects the decisions people make about families. Economists are also interested in families because we have come to realize that there are many parallels between family and labor markets.

Dr. Wolfers: The first place that people notice the similarities between family and economics is in what some have called the marriage market, which looks a whole lot like the labor market. People search for partners the same way they search for jobs. When you find a spouse or a job that looks like a good fit, you take it. And you must make a decision about how much time to spend searching for the perfect spouse or the perfect job before accepting a job or a spouse.

Weighing in on the soda ban

I’ve been only nominally present on the internet lately due to family stuff I won’t bore you with, but the last few days have seemed especially filled with vitriol towards Bloomberg’s soda ban.

I understand the detractors to be of a particular political bent, but I’ve been surprised by both the magnitude of the response and the apparent blinding nature of the issue. People I generally consider intelligent and levelheaded have, in my mind, totally missed the boat to have a larger conversation on policy.

Will Wilkinson, in the Economist, provides an account (invoking Jonathan Swift a bit) for his stance against paternalism, but sets up false dichotomies.

GIGANTIC sugared soft drinks are disgusting. Let’s just get that out of the way. Can we also agree that the high-calorie drinks rich people like to consume—red wine, artisanal beer, caramel frappuccinos, mango smoothies with wheatgrass and a protein boost—aren’t at all disgusting? At any rate, we yuppie pinot-drinkers know how to look after ourselves. In contrast, the wretched classless hordes, many of them being of dubious heritage, lack the refinement of taste necessary to make autonomy unobjectionable. Those who abuse their liberty, filling the sidewalks of our great cities with repulsive shuffling blimps, can’t expect to keep it, can they?

All those high-calorie drinks that rich people consume are consumed by rich people for a reason. Well, several, probably. They taste good (except wheatgrass, yuck), they confer some sort of status on the drinker (conspicuous consumption), and they’re expensive. Have you seen the price of Bordeaux lately? 2010 Chateau LaTour is $1500 a bottle. Why, you ask, despite there being a glut of many other wines on the market? It’s likely because Bordeaux is one of the few foreign wines that has been introduced to China. And the Chinese love their red wine. But we (the US government) don’t intervene and say we have to make sure that wine grapes are affordable and vintners stay in business, so let’s incentivize more wine grape growing in France or prevent the Chinese from demanding wine.

But I digress. These drinks are expensive because the market recognizes both their inherent qualities and conspicuous consumption qualities, identifies demand and supply, and provides at the equilibrium price. Every one of my principles students could show you a graph to that effect.

The difference is that we do intervene with corn, a primary ingredient in sugary sodas, which artificially holds down the price. Sugary sodas are not just cheap because of supply and demand; they’re cheap because government intervention, particularly subsidies for growing corn, keeps corn abundant and cheap.

Again, my principles students could all give you a list of reasons why those subsidies are in place: we care about food security and being able to provide our own food in case of a crisis; we want to preserve a rural way of life; we want to make sure farmland is used for farming and not housing developments. But also, subsidies become entrenched, often far beyond their usefulness. Farmers are used to the guaranteed income and don’t want to give them up. Companies who buy corn want corn cheap, so they lobby to keep the subsidies in place. Pop includes high fructose corn syrup as an ingredient, a cheap alternative to sugar. They want to keep the price of corn low so they keep their profit margins while keeping their product cheap.

Subsidies are hard to get rid of because they have the property of concentrated benefits/diffuse costs. A few people benefit a lot from subsidized corn (and indeed we all benefit a little from low prices on foodstuffs that have subsidized corn as an ingredient), and we all pay a little through our taxes to keep those prices low. Subsidies, in theory, should stay in place as long as the benefits to society outweigh the costs. And perhaps the soda ban shows that the costs (increased obesity as a result of soda consumption–though perhaps a tenuous link) are greater than the benefits to society as outlined above. Or Bloomberg’s just a paternalistic whack.

Is banning pop in larger than 16-oz bottles the right answer? Probably not. Can Bloomberg single-handedly change US farm policy? Absolutely not. So he does what is within his control. Is it paternalist? Yes, totally. But if we were really worried about paternalism, why didn’t I hear all of these people crowing about laws that seek to limit abortion rights and intimidate mothers and prevent access to birth control? Make a distressed rape victim listen to a lecture about her child’s beating heart, allow her doctor to withhold information from her, and make her wait three more days before having an abortion? Sure! But increase the cost of consuming something whose cost is artificially low and presents potentially harmful negative externalities? The nerve.

Claire Potter makes a similar argument in the Chronicle of Higher Education.

Jobs, poverty and teen child-bearing

Several weeks ago, I printed out an NBER working paper on teen childbearing by Melissa Schettini Kearney and Phillip B. Levine. I had every intention of reading it then, but it just wasn’t going to happen at the end of this totally crazy semester. Since then, a few things forced my hand. I finished the semester (yay for surviving my first year of professoring!), the paper has been accepted for publication in the Journal of Economic Perspectives, Matt Yglesias put together a nice little review of the article in Slate, and a friend emailed me rather incensed by Yglesias’ review. From a quick scan of the JEP version, it doesn’t appear too much different from the NBER version, but my comments refer to the NBER version.

Yglesias’ review presents Kearney and Levine’s research as novel and surprising, but I think that misses the point. While the authors do a good job of aggregating statistics from several data sources and findings from different papers, the primary contribution of this paper is not novel, but rather confirming what we already know: that teen pregnancy is higher in the US than other places and; that poverty likely causes teen pregnancy more than teen pregnancy causes poverty. Past studies, cited in the paper, have shown that teen pregnancy has little to no effect on outcomes when you control for poverty, or within-family characteristics, and in some cases, may even result in better outcomes than if the teen hadn’t become pregnant. This is a significant theme in Edin and Kefalas’ ethnographic study, Promises I Can Keep, which I discussed here, and other research in fields such as sociology and demography.

Ultimately, the economics community thought it was an important paper as it went to a very prominent journal, but I really just see it as a good synthesis of what we know.

In related, and I think more exciting research, the link between poverty and teen child-bearing may be even tighter than suggested Kearney and Levine’s paper, though not in the way that the Kearney and Levine paper posit. A working paper by three Duke Sanford professors, Elizabeth Oltmans Ananat, Christina Gibson-Davis, and Anna Gassman-Pines examines the link between job losses and teen pregnancy.

I’m so predictable. I love this paper because even the anticipation of poverty, or joblessness, more specifically, predicts teen pregnancy rates. The authors show that when mass layoffs are announced in a North Carolina (before the layoffs actually occur), that county sees a subsequent corresponding reduction of births to teenagers in that county, but only for Black teenagers. The mechanism appear through both reduced pregnancy rates and reduced birth rates, which suggests that teens are both practicing safer sex and having more abortions when job prospects in their counties suddenly become dimmer.

There were a few places I thought the paper could improve, and the first one is my primary concern. Even though the authors find a statistically significant effect, I’m curious about the mechanism for how this affects teenagers. What evidence is there to show that teenagers would be affected by these job losses? Why aren’t they just in school and ignoring them? Initial information about their education level, school attendance, when they enter the workforce, etc, would be useful, to sell the story. I think the age and education of teens would be a big factor here. Wouldn’t you see a bigger effect for teens closer to graduation? Or a smaller effect in counties where teens are more likely to go to college (say wealthy Orange county, where Chapel Hill is located)?

The ability of inhabitants to migrate and commute is also problematic and suggests a (you guessed it!) spatial auto-correlation issue that I imagine is present. The authors claim they are underestimating the effects of job losses by ignoring migration and spillovers, but I wonder whether there are spillover effects that could be estimated through job loss in surrounding counties, rather than just say it’s a lower bound. Also, if spatial auto-correlation is present, that’s going to affect the standard errors, not just bias the estimates.

A minor, but I think incredibly important interesting, result is that the job losses also resulted in fewer black mothers reporting a father’s name on the birth certificate. The magnitude of the effect is approximately half of the effect of that on the pregnancy rate itself, which is pretty large. I think this result actually goes a long way towards answering my first question: Why do we think teenagers would be affected by this? If the story is that teens are being more careful about sex or having more abortions when their job prospects are low, is it really their own unemployment they fear, or also their partner’s? Teenage parents are less likely to be married than their older counterparts, so who is supporting them through their pregnancy? Paying for prenatal visits? Do teens feel they’re going to be working and raising their children? My own work shows that black mothers at any age are more likely to receive a promise of financial support and Edin & Kefalas suggest that the promise is key to the marginal have a baby (or at least stop trying not to have one) for mothers of low socio-economic status. I think this relationship could be teased out a little more.

All in all, it’s a good read, and presents an interesting counterpoint to the Levine and Kearney paper. L&K say poverty causes teen pregnancy, but the Duke paper says that teens are responsive to future job prospects, and respond by delaying (or at least trying to avoid) childbearing.

At first glance, the papers might seem incongruous, but it’s really a stock versus flows kind of issue. Other things equal, teenagers in poverty are more likely to become pregnant early due to a host of factors, but they still plan and have an idea about how they will care for the child. When that plan is disrupted, it appears it can affect some teens’ decision to bear children, on the margin.

Director of monetization: Economy 2.0?

About a year ago, a good friend of mine started developing an online gaming platform. The point was to create a place for several different game creators to host their games, increase their user base and allow for exchange of in-game currencies. We talked a lot about, and I thought even more about, how to create an in-game currency in a way that reflected use preferences, wouldn’t inflate or deflate too quickly, and ultimately, would earn my friend some real cash.

The platform is in beta now, and we never managed to formalize some of the things we’d hoped to test, but since then, we’ve exchanged lots of emails and articles about in-game economies, inflation, relative worth of found objects and more. Just last week, we had dinner and my friend asked whether I thought gaming companies employed economists to create in-game economies. Some, he contended, were incredibly realistic and well designed, others suffered from gluts of goods and all other sorts of problems. We got an answer fairly quickly. This month’s JOE (Job Openings for Economists) came out today and while flipping through it, I saw this notice, for a job at fiveoneninegames, looking rather conspicuous among the ads for financial analysts and visiting assistant professors.

The funniest part of all this, of course, is that I’ve never played one of these role-playing games. Bejeweled? Zuma’s Revenge? Sure, but I’m much more of a crossword and sudoku kind of girl. All the same, I’m really tempted to apply. Is that weird? The idea of having total control over an economy (even with a non random, selected set of participants in the game) sounds so appealing.

Okay, weird, I know. Back to running regressions.

The measure of a market (a really old one)

A year ago, about this time, I was on my way to Ottawa for the Canadian Economic Association and Canadian Network of Economic History Meetings. They coincided, so I presented papers at both and took the opportunity to adamantly assert that I was not an economic historian.

A year later, I have a book chapter, a working paper, a paper (almost–I’m just waiting for confirmation it was sent out) under review, and a paper idea percolating that all belong under the label Economic History. I’m trying to get the paper idea in shape to submit to the CNEH meetings again this year, with the deadline fast approaching.

While my coauthors and I were hard at work on the paper on financial portfolios in the early 18th century, the one that is (almost) under review, a seminar participant at Stanford asked my coauthor what an optimal portfolio would look like. We didn’t know. None of us is really a finance person. I got involved with the project because it had a gender component and the others on the papers are economic historians.

I took it upon myself to pick the brain of my colleague who teaches finance at Gettysburg and found myself quickly immersed in the heady world of portfolio optimization, betas, alphas, indices, Markov matrices, and so much more. From my understanding of the literature, the S&P500 represents the closest thing we have to an optimal portfolio, and so creating a similar index for the time period we’re interested in should provide the answer to the seminar participant’s question.

What I find particularly interesting about the index method of portfolio construction is that the S&P500 in particular is thought to provide an accurate picture (returns and growth-wise) of a balanced portfolio of all assets–not just financial instruments. If you were to put a big chunk of your money in a fund that purchased stocks along with the S&P market capitalization strategy, you would actually be bringing your portfolio out of balance by buying things like real estate or durable investment goods.

The idea, I believe, is that the stock market has “evolved” such that it captures the risk and reward of all those types of instruments–not just the stocks themselves. I find this assumption, particularly given the dramatic dips and peaks in the stock market we’ve seen over the past four or five years to be heroic, at best, but it becomes more problematic when we turn to 1700s finance.

The financial instruments available in the period for which I have data are incredibly few in number and even more limited in scope. Besides a bank or two, they are joint-stock, charted trading companies, whose fortunes lie entirely in the wind and the water and the ability of colonists to extract resources from the colonies. There’s no ability to invest in steel or textiles or the machines that make them. I don’t have information about real estate or other investments for most of the people in the sample, and I certainly don’t have their prices. So, our optimal portfolio can really only be for the available stocks, not for the entire gamut of instruments.

I doubt that I’ll be the one to rewrite modern portfolio theory, and I do think this is the best place to start, but it’s not ideal. Story of an economics paper, I guess.

Spatial auto-correlation is not causation

There’s a strong tendency in human nature to draw distinctions along dichotomous lines. Good and evil, black and white, ugly and pretty. We all know that these distinctions only really work in children’s fiction, and even then tend to fall flat, but we try anyway. In teaching, particularly a new subject, those dichotomies are both useful and can lead to the downfall of a lesson.

In that vein, the instructor in my spatial econometrics workshop last week presented two significant data issues that a researcher might encounter in using spatial data: spatial heterogeneity and spatial dependence.

By way of definition: spatial heterogeneity is simply that there is something about an area or a piece of space that is different than the spaces around it. My dichotomizing, learning mind went immediately to the idea of observables. Clearly, if we are trying to include spatial information–location–in a regression, we know that the area has certain characteristics. As long as we explicitly control for these in our regression (and believe they are accurately measured), it doesn’t present much of a problem.

However, this is not always the case due to the level of analysis problem. In a general econometric specification, we control for the unit of spatial analysis that is relevant–county, Metropolitan Statistical Area (MSA), state, whatever it may be. By choosing the level and assigning a dummy variable, perhaps, we assume that all those characteristics are captured uniquely, but also that they are assigned independently to the spatial unit. Take for instance the distribution of the African-American population in the United States. Regression analysis that uses that variable as a covariate assumes that the number of African-Americans in Georgia is independent from the number of African-Americans in South Carolina, which makes little intuitive sense. Both were states with large plantation economies that employed Black slaves from Africa in production of goods. It makes sense that these two states, spatially proximate, would also have similar factors leading to their demographic makeup. Thus, spatial heterogeneity: areas in the South have higher Black populations than in the North.

The corollary to spatial heterogeneity is spatial dependence. Like spatial heterogeneity, we see patterns occur in certain variables, but rather than an outside, perhaps observable and easily measurable factor that accounts for the clustering, there’s something inherent about the place itself that causes proximate areas to change their realization of some variable. Think of housing prices. Housing prices are higher in places with certain amenities (close to transportation, mountains, whatever), but housing prices are also higher in areas with higher housing prices. Perhaps homeowners see their neighbors selling their houses for more and thus put them on the market for more. Or buyers see houses in the area with higher values and thus are willing to spend more. This spills over county and other lines, too.

Both of these problems, regardless of how strict that line is between the two, manifest in spatial auto-correlation. The variation we see in each variable for two spatially proximate observations is less than the variation for two independently observations because the information comes from the same place. Some of this we can control for, some of it we can’t, and some of it we can try to control for with the tools I’ll discuss in coming days.

Regardless, it’s important to remember that the realization of spatial heterogeneity and spatial dependence is the same mathematically. Statistically, we cannot differentiate between whether some unobservable variable caused everything to be higher, or whether each observation is exerting an effect on its neighbors (a butterfly flaps its wings…). So, even with acknowledgement of these problems, we have not established causation.

A familiar refrain is, thus, minimally modified: spatial auto-correlation is not causation.

A note on correlation and causation: (see Marc Bellemare’s primer for a more detailed explanation)

Anyone who has ever taken a statistics course is familiar with the refrain that correlation is not causation. It’s a common refrain because it’s something that is often ignored when statistics are cited in news articles and personal anecdotes. My favorite example of this is that ice cream sales and murder rates are highly correlated. Only the biggest of scrooges would believe that ice cream sales caused murder rates to increase. In the abridged words of Elle Woods, happy people don’t kill people. And in my words, ice cream makes people happy.

They do move together, though, which is essentially the definition of correlation. When ice cream sales go up, murder rates go up; when murder rates go down, ice cream sales go down. Not because one causes the other, but rather because of the seasonality of both variables. More homicides occur in the summertime, and more ice cream is sold in the summertime.

Spatial Econometrics: The Miniseries

Last week, I spent three days in a workshop (or short course) on spatial econometrics at the University of Colorado‘s interdisciplinary population center, the Institute for Behavioral Science. At the beginning of last semester, many of my methods students expressed interest in doing their research papers on a topic with a significant spatial component. I would have loved for them to incorporate spatial analysis, but it was a topic I had touched only tangentially and didn’t feel qualified to learn it at the same time as teaching that (incredibly demanding) course for the first(ish) time. In addition, having just attended the PAA meetings in San Francisco, I’ve been looking for ways to expand my econometric skills and incorporate spatial data into my work. It was really fantastic. I don’t know whether they’ll be hosting the event again next summer, but do keep a lookout if you’re interested. I thought it was extremely helpful. And fun (see nerdy tweets from last week about loving matrix algebra). Paul Voss, of the University of North Carolina’s Population Center, Elisabeth Root, and Seth Spielman were all great.

I posted a short introduction to spatial econometrics last week based on my readings for the first class and am now excited to share some of the things I learned, so over the next few weeks, I’ll post some of my thoughts in a mini-series on spatial econometrics. This post will be updated with a list of posts in the series, so do follow along.

Experts, please keep me honest! This stuff is very cool, but I’m still a newbie.

Preliminary outline (subject to change):

An introduction to Spatial Econometrics
Spatial Autocorrelation is Not Causation
The Weights Matrix for Spatial Analysis
Some Notes on Terminology in Spatial Econometrics

Percolating

My short course this week at CU’s pop center was incredible and exhausting and incredibly exhausting. The easiest part, for me, was thinking about spatial models using matrix algebra, if that’s any indication of what we did all week. I’m fairly certain I forgot how to write STATA code and learned just enough R and GeoDa to be dangerous, but you can bet that’s not the last of me.

Next week should bring to fruition ideas and blog posts that have been percolating: teen pregnancy, more spatial econometrics (separately, although, that gives me an idea…), and some 1720s finance, as well as back to your regularly scheduled programming.

Have a safe and happy holiday weekend!

Replication, or the lack thereof, in Economics

My scientist friends have always been puzzled by my responses to questions about replicating studies in Economics. It’s just not done very often. In fields like astrophysics and biology, replication is almost as important, if not more important in some cases, as the novel finding itself, but not so in Economics. I’ve seen evidence that other social sciences are similar and there was some recent debate about the replication of psychology experiments and the failure to come to the same conclusions using similar methodologies. (There were other pieces on this, but this is one that I found today). In short, journals favor novel and interesting outcomes, so obvious or unsurprising results are far less likely to be published. The publication of the novel results leads to a power imbalance (she already published this, so she’s the expert and gets the soapbox). No one wants to fund or highlight research that’s already been done. Replications that confirm are boring and replications that challenge established findings have to be 110% on everything.

It’s really hard to challenge established findings. Look at how long (three years after publication) and how many papers it took for Emily Oster to admit her paper on missing women and Hepatitis B was wrong. Regardless, she still has a job and now tenure at Chicago. Or how many papers have been written challenging Donohue and Levitt’s abortion paper and they still stand by it.

I got a bit far afield, though. Economists are not generally in favor of duplication of effort. If someone’s doing it already, unless you can do it a lot better, you shouldn’t really do it. Hence persistent ideas of comparative advantage and gains from trade.

However, the recent spate of randomized control trials, particularly in development settings, has prompted more and more debate about the validity of these experiments and appears to have resulted in at least one group that’s eager to test and replicate in order to confirm (or deny?) the validity of certain projects.

Clearly, there are limits to what can be replicated using existing data, and limited funding to collect new data using similar methods.It’s unclear to me how they will choose appropriate experiments to reproduce or test, and as much faith as economists tend to put in a sample size of one, I’d bet we won’t be too happy with a sample size of two, but I think it’s a good start. The Development Impact Blog by the World Bank will keep up with the process of replication, so worth following if you’re interested. I know I’ll be watching.

h/t @JustinWolfers

Though kind of dated now, Daniel Hamermesh’s paper on replication in economics is here.

On anticipating divorce, again

Related to my post earlier this week, a new working paper shows that women in the US respond to increased divorce rates by working harder. Knowledge of high divorce rates appears to be enough to incentivize working harder in anticipation of even a probabilistic one-earner household. I haven’t had the chance to read the paper itself (I will, but 68 pages!?), but Ezra Klein discusses it here:

Why would this be the case? Researchers believe it’s because marriage provides “implicit social insurance” for women, who are still more likely to be the secondary income-earners in the U.S. and Europe. So in the U.S., where divorce rates are higher, “women have a higher incentive to obtain work experience in case they find themselves alone in the future,” they write. “European women anticipate not getting divorced as often and hence find less reason to insure themselves by working as much as American women.”

A longer treatment of the paper by the authors is on the VoxEU website.

Referenced: Chakraborty, Indraneel, Hans A Holter and Serhiy Stepanchuk (2012). “Marriage Stability, Taxation and Aggregate Labor Supply in the US vs. Europe”, Working Paper.

Share this:

Share this:

Share this:

Share this:

Share this:

Share this:

Share this:

Share this:

Share this:

Share this: