The power of Grandmothers

This NPR Goats and Soda story is too good not to share. In Swaziland, a successful (or least appears to be successful–can we evaluate it?) program to reduce child abuse via grandmothers is my new favorite thing. By empowering children to approach and talk to grandmothers about their experiences, and giving grandmothers the tools to get kids the assistance they need,

They use stories with animals to give kids friendly characters so that children can easily recall pathways for help. Grandmothers go door-to-door to introduce themselves and their role in the community, normalizing the practice and making sure everyone knows it is a visible role.

The program has helped address one of the challenges of child abuse — helping children understand that certain kinds of treatment are not OK and should be reported to a grandmother or another trustworthy adult.

I love this. My current research in Tanzania is partially focused on identifying people in communities who can play this sort of role, to model new social norms, or to be neutral arbiters and safe havens. I’d love to see it evaluated and replicated.

The Quant-y, Measurable-y Stuff

This post is Part 3 of a five-part series on measuring Al Mokha’s impact in Yemen. (It is also available on Al Mokha’s blog, as are Part 1 & Part 2)

Erin and a Tanzanian woman smiling with an orange dirt background
Author (on right) in Nyarugusu, Tanzania, during a project for the International Rescue Committee 

As a development economist with interests that are a little outside the norm, I spend a lot of my day thinking about how to measure unmeasurable things. How prevalent is a certain belief? And how does it affect people’s behavior? Can one violent event, or experience, be objectively seen as worse or more violent than another? And if so, what determines that violence—scope, tenor, frequency? How do we fix it?

So, when Anda told me he wanted to start thinking more about impact and measurement at Al Mokha, I jumped up and down with glee. From the moment he and I first talked development and coffee in Cambridge almost a year ago, I’d been questioning, “cool, but how do you measure that?”

As an aside, when I tell someone I’m an economist, there are a few common tropes that people immediately fall back on. There’s often an immediate question of whether I can do their taxes (I could probably hack it, but you’re better off with a CPA) or help them pick stocks (I am the last person you want doing that).

And then there’s the belief that I must work with GDP and trade or exchange rates.

I don’t do any of those things, but I do think that GDP is a great place to start thinking about measuring Al Mokha’s impact. Al Mokha doesn’t want to just sell coffee; it wants to fundamentally change Yemen’s economy by revitalizing a sector that has been decimated. And if you can revitalize a sector that at one point dominated an economy, you should be able to see change at the GDP-level.

GDP word graphic with jumbled mess of economics terms

GDP is deceptively simple. When you hear about it on the news, GDP is a single number, maybe a few numbers if someone talks about year-over-year growth and growth in a particular quarter, but usually just one number. That number, however, is made up of literally thousands of numbers, of transactions between people and firms and countries and governments. In many countries, as in the US, there is an entire government agency devoted simply to measuring GDP, to making the rules about what counts in GDP and what doesn’t.

Yemen has been seriously suffering in the GDP department. According to the Yemeni Ministry of Planning, Yemen’s GDP is projected to shrink by almost 35% in 2016. Compare that to a paltry 1% shrinkage seen during the Global 2008 Recession in the US.

If the (or one) goal is to grow Yemen’s economy rather than to shrink it, what would it take to increase Yemen’s GDP via coffee? The most straightforward way would be to increase the size of the coffee sector. Economists like to think about questions like this as changing extensive or intensive margins. The extensive margin would be to increase the number of acres planted. But we can increase the size of the coffee sector without planting one more coffee plant, if the price of coffee beans from Yemen simply goes up, or yields are higher due to better fertilizer use or some other factor (the intensive margin).

"Coffee Up" word graphic with U a coffee cup

So, we started by asking a simple question, how do we raise GDP? And with one logical step, we already have four or five different outcomes we could try to measure (either separately or in tandem with GDP):

  1. acres of coffee planted
  2. kilograms of beans harvested in total
  3. coffee prices
  4. coffee yields (kilos of beans per acre)

And inside each one of those is also a list of measurable outcomes. For instance, do higher coffee prices lead to higher incomes for farmers? Or middlemen? And which is desirable (or is it both)? We went from one seemingly simple number to lots of measurable outcomes. Al Mokha is trying to figure out which ones it cares about.

Yemen coffee farmer picking coffee cherries

Finally, we also have to think about what may come out of this project that isn’t so good. “Unintended consequences” is kind of a loaded term right now if you’re following elections in the United States, but economists use the term a lot. We also call them externalities. While Al Mokha’s goals are to promote economic growth and diminish the appeal of violent insurgency, some may see pumping money into a failing state as foolish. It’s possible that any extra income going to farmers could be siphoned off by insurgent groups. Or that kids will be taken out of school to work in the fields. Or that creating demand for coffee will reduce agricultural space devoted to food, which is problematic in a country that is considered extremely food insecure by the World Food Program.

Some of the above are measurable, albeit with a lot of effort. We can examine what (little) data is coming out of Yemen from the government and international organizations. We can poll farmers who are supplying coffee about how much they are planting and earning. We can add up all the coffee production in the country and show how it changes year over year, and from there do a little back-of-the-envelope calculation on GDP change. We could show how higher prices of coffee offset negative welfare impacts (ala this paper on Quinoa in Peru by Bellemare, et. al.). Not that any of these is particularly easy in practice, but there’s at least a path to understanding them.

Yemeni man holding basket of bright red coffee cherries

Furthermore, lots of outcomes Al Mokha hopes to achieve are not measureable. I’m going to save that for my next post, the nebulous and qualitative, and leave you with this: deciding what to measure and how to measure it is an ongoing process. GDP is a great place to start. It is a number that is widely recognized as a measure of economic success, and an area that is visibly and severely suffering in Yemen right now. GDP-level change is the kind of change we’re hoping to accomplish.

So what does a change in GDP actually look like? I did a quick back-of-the-envelope calculation: doubling Yemen’s current coffee production (of 19,800 tons in 2013) will increase GDP somewhere between 0.30 and 3.73 percent. That may sound small, but it’s not. That’s $109 million – $1.3 billion.

It’s clear there’s huge potential to increasing coffee production at an economy-wide level and big picture, we’re focused on that. But there’s all those components, too. We’re going to keep digging down, into yields and acres planted, into where the money ends up and how it is spent, and into how it changes outcomes for families. We believe that impact is important. Al Mokha is trying to sell coffee, but with a purpose, and a seemingly simple goal to sell coffee will have many disparate impacts all over the economy, from the very biggest measurement, to the smallest.

Calculation, for those nerdy enough to care: First, note that good numbers from recent years are hard to come by, so these are estimates, but the leave us with answers that are generally on the same order of magnitude. If coffee production is 19,800 tons (2,200 pounds a ton, 2013 estimate from the Yemen Ministry of Agriculture), and farmers receive $2.50 per pound, doubling production would mean an extra $109 million in the economy, which is about 0.30% of 2013 GDP estimate of $35 billion (World Bank). However, GDP only counts finished goods, so if we take the US consumer price of Al Mokha coffee, at $30/pound, you end up with an extra $1.3 billion in the economy, a 3.73 percent change.

The reality is probably somewhere in the middle in terms of what would get counted in Yemen’s GDP as it is unlikely that U.S. consumer prices or green coffee bean prices are the relevant metric, beans will be priced differently according to quality, and ultimate benefit to farmers and Yemen depends on much value-add is done in-country and how much is exported. But even lower estimates of prices and current production yield effects on the same order of magnitude. If we assume exporters are getting $6/lb., that yields $261 million for a 0.75% increase in GDP.

Those interested in trying Al Mokha’s coffee can shop at www.almokha.com

Note: I am Al Mokha’s advisory board member in development economics. I have no financial interests in Al Mokha and have received no compensation for this post.

Things I felt like sharing today

  1. I’m reading about lots of obscure methods of explaining the variation in multiple variable combinations, like principal components analysis and partial least squares regression. Or at least, they are obscure to an economist; I imagine they are taught more frequently in other fields. I had to laugh at this Stata help note on partial least squares: “This program is provided for educational purposes. It is difficult to recommend the PLS composites for any serious empirical work (see Rönkkö, McIntosh, and Antonakis (2015).” #movingonthen
  2. Talking with a doctor friend last weekend, I can’t speak to how true this is, but replace ‘clinician’ with ‘teacher’ and ‘hospital’ with university and you have a conversation I’ve had many times: “We don’t have great clinicians in (xyz speciality) at (supposedly one of the best hospitals in the country). They don’t incentivize you to be a good clinician. You’re rewarded based on your research dollars and sometimes patients suffer. Only schools like (by some estimation the best medical school in the country) produce good clinicians because they have the resources to devote to them, but the school doesn’t care about you if you want to be a good clinician. They’re just biding their time until they can push you out.”

World Refugee Day

I haven’t been writing much the past couple of years, or rather, I haven’t been writing in this space. Much of my day-to-day work is writing academic papers, reports, proposals,etc. Today is June 20, World Refugee Day, and I want very much to tell you about my experiences working with refugees over the last six months, but I’m just not there. So, instead, I’ve put together a short list of links. Taken together, they give a picture of refugee life that is not as visible in day-to-day coverage of refugees, though perhaps not as coherently as if I had put together a narrative. It will come, one day.

  • 65.3 million people were displaced by the end of 2015. That’s about 1 in 113. More stats from UNHCR.
  • In Kenya, hundreds of thousands of people are in danger of being forcibly removed from refugee camps. My colleague and friend, Stephanie Schwartz, on why this is a really bad idea. Forced repatriation from refugee camps has a long and messy history in the Great Lakes Region of Africa.
  • In Sulaymaniyah, refugees and displaced Iraqis are participating in Refugees Got Talent, a celebration on #WorldRefugeeDay. You can see performances and hear stories on their facebook page.
  • Why we shouldn’t use the terms refugee and migrant interchangeably and some other FAQs.
  • Refugees are almost certainly at greater risk of sexual assault, sexual exploitation, and gender-based violence, and have fewer tools to address such problems in the face of limited mobility, poverty, proximity to abusers, and lack of access to services.
  • In Nyarugusu, two of my enumerators are trying to finish master’s degrees in Tanzanian universities. They tell their stories about the desire for education on this fundraising page. The fundraiser is technically closed, but they’re still short a few hundred dollars to be able to graduate. Let me know if you are interested in helping out.
  • UPDATE 6/21/16: Owen Barder on distinguishing migrants vs. asylum seekers and why it’s so hard. Key (mouthful of a) quote: “The intrinsic interconnectedness of persecution, conflict, economic collapse, environmental degradation, and natural disasters makes it impossible to separate these motives for migration.”

Introducing my work at Al Mokha: Peace, love, and coffee on earth

This post is Part II of a five-part series on measuring impact and progress at Al Mokha. A version of it is also available on Al Mokha’s blog. Part I is available here.

For my friends and readers, this post is something of an announcement that I’ve joined the advisory board of a start-up coffee company. Al Mokha is coffee with a purpose, premised on a big idea: that revitalizing a 500-year old industry can be a vehicle for peace and prosperity, that strategic investments in certain markets can be an effective foreign policy tool.

Coffee, Erin?, you might say. Didn’t you write not too long ago about how you had your first cup of coffee ever? Like, literally ever?

It’s true; I don’t drink coffee (though I did have my second cup ever in Tanzania in February–you try telling an Ethiopian and an Italian that you don’t drink coffee). So, I can’t even really tell you whether I think the coffee is tasty. But, I do know that other people really like this coffee. And I think this is exciting. So, bear with me.

Erin Fletcher sitting at laptop with hut in background
Author conducting research for the IRC in Nyarugusu, Tanzania

In Part I, Anda–the founder of Al Mokha–wrote about the perils of trying to be a salesperson / academic, and how that duality plays out in the business: investors want to see sales growth whereas economists (me!) want to see real, measureable change in things like poverty levels, in coffee production, in anything we can quantify with data.

So we’re working towards that. In this post, Part 2, I introduce myself and in parts III- V we tackle some tough questions.

Well, who am I? If you visit Al Mokha frequently, you know Anda and you’ve heard mention of some of his advisors (as he spins tales of sinking all his time into a startup). I’m the nerdy PhD obsessed with data and development. I have a doctoral degree in economics. I spend most of my time reading and writing papers on violence against women and children and female labor force participation. I spend a lot of time thinking about very economist-y things like incentives.

For a while, I did this thinking and writing at Harvard. That’s actually where Anda and I met. Over beers at Shay’s in Cambridge, he was quick to tell me he didn’t want to talk about work (and thus didn’t want to talk about Al Mokha). Faced with a real-live economist, though, he couldn’t keep quiet long, and we spent the evening having a super nerdy academic discussion about development economics, foreign policy, and reducing violence. We talked about famous development scholars like Angus Deaton, and making lists of other economists he should read, like Marc F. Bellemare.

These days, I do similar work for international organizations. I’ve worked with UNICEF to model what a fully functioning child protection system would look like in Zimbabwe. I’ve advised the United Kingdom’s Department For International Development (DFID) and the Nike Foundation on social norms. And in a room full of experts on evidence for the UN Population Fund (UNFPA), I’ve found consensus on evidence in programming for adolescent girls. I’ve written and administered surveys to better understand poverty, disease, and violence in Nepal for Engineers Without Borders, and Kurdistan and Tanzania for the International Rescue Committee (IRC). These may seem like disparate topics, but I think what they reveal is that I like thinking big—about varieties of programs, about outlandish ideas, about what the best model is to solve a problem in development.

At Al Mokha, this is also my job. Anda and I share a vision of development that is big. It is, however, nestled in the details, of course: I read articles about agriculture and development, but with each paragraph, I’m highlighting what I think is relevant to Al Mokha’s model, whether it helps or hurts the plan to revitalize Yemen’s coffee sector. We talk about how new research supports or hurts the underlying motivation and theory behind the business. What makes this big? What makes this different? Currently, we’re also spending a lot of time thinking about measurement, which is really what this blog series is about.

Can coffee actually stem terrorism? Can we foster a big enough market for Yemeni coffee to actually have an effect on poverty and people’s general wellbeing? Is it even possible to build up an export business in what people are calling a failed state?

Old Kurdistani tank and other army vehicles
War memorial in Kurdistan (taken by author, 2015)

I don’t know if coffee can bring peace to earth, or even to Yemen. The evidence on how economic activity affects terrorism and violence is mixed. Some say that agriculture and people getting training and jobs can reduce violence, but the programs that have been tested are complex and require a lot of inputs. It’s also messy and full of endogeneity problems. I’m not really sure what the best way to measure peace is. I also don’t know if boosting the coffee sector is the right thing to do for Yemen’s economy. Could money and effort be better spent on technology or some other industry?

But there’s also a lot of promise in the idea. Yemen’s embassy in DC and Yemen’s Ministry of Agriculture, for example, see coffee as an engine for rehabilitating Yemen. This particular coffee is really good. And creating a stable market for a commodity could reduce price volatility for coffee, encouraging farmers to plant more. Okay, I already got more technical than I intended for this post, but I promise we’ll come back to price volatility.

So, who’s right? And what does it take to show that Al Mokha works? What does “work” even mean in this context? What can we learn? I have lots of questions, and a few might have come to your mind as well. There are probably many more I haven’t even thought of. But that’s super exciting to me; there’s so much to find out.

Over the next three weeks, I’m going to write more about Al Mokha’s impact and how we are going to measure that. It’s something we expect to evolve as the company does and we hope to get some of you involved in our research and thinking about how we go about it.

It might get a little nerdy. You’ve been warned. Stay tuned!

Those interested in trying Al Mokha’s coffee can shop at www.almokha.com

Note: I have no financial interests in Al Mokha and have received no compensation for this post. As a member of the advisory board, I occasionally receive coffee samples, which I hand out to friends and family as I only learned how to brew a cup of coffee last month.

The plight of comfort women

Japan and South Korea announced a deal today whereby Japan will pay $8.3 million to the government of Korea and apologize for its use of comfort women during World War II. The money will go to create a fund for victims, according to the AP, though I haven’t seen details of what that will look like.

I’ve only recently become acquainted with comfort women stories, but from what I’ve seen (first from my visit earlier this year to the Nanjing Memorial Massacre Hall), I’ve found comfort women stories very weirdly impersonal. They seem portrayed as crimes against the state or about honor, and never feel like they are really about the women. The euphemism “comfort women” is especially divorced from reality. It feels like this collective failure to protect women from outsiders leads to embarrassment and emasculation of men, rather than concern for actual women who were trafficked, enslaved, and raped repeatedly over a period of years.

Gut bacteria and pregnancy

I’ve managed to get two papers out over the past two months on obesity and pregnancy with my wonderful co-author, Susan Averett. One is available online at Maternal and Child Health Journal; the other is forthcoming in an edited volume, Applied Demography and Public Health. The basic gist of both papers is that maternal obesity is probably not as bad for kids as we tend to think. Using maternal fixed effects we show that kids of moms who were obese pre-pregnancy are not doomed to have worse outcomes than their siblings who were born to the same mothers but when they weren’t obese. There just aren’t that many differences in terms of adverse birth outcomes or obesity in early childhood.

One thing we can’t control for, that I’ve been discussing with a friend recently, is how gut bacteria plays a role in all this. This friend* is pregnant, very close to her due date, and, as she approaches her due date, has been religiously eating dates, largely based on the results of one study in Jordan that women who ate dates were more likely to go into spontaneous labor, labored for fewer hours, had higher mean cervical dilation, etc.

My immediate response was, “well, how can we be sure the women eating dates were different somehow than those who weren’t?” And my next move was to find the paper and point out this line:

As date fruit consumption is a part of the cultural beliefs of the population under study, it was difficult to get patients who would commit to not taking date fruit at all (control group), therefore the study group had more patients than the control group.

The authors are actually careful to say that the study should be repeated using a randomized control trial, but I don’t know that we learned that much from this as the control group is clearly different in this obvious way from the intervention group (they’re willing to give up dates). It is interesting, though, to think about how foods change our gut bacteria and how that might affect other health issues.

*Addendum: This friend would like to point out that she likes dates and was aware of the limitations of the study. “It can’t hurt!” she says.