What You’re Feeling is Grief

60815764515__C2B82688-53CC-4A0E-AC79-85E815828994I had an early morning meeting this weekend with a colleague at a data collection firm we’ve been working with. He said he was taking next week off and was headed to Costa del Couch. We all laughed, but I noted internally that I’ve had conversations with several colleagues this week about taking time off or leave during these weird times–from childcare or our jobs or searching for jobs–and while it may seem counterintuitive to take a break when we can’t escape the way we normally do, I wanted to share with the world that it’s really important to take time away right now.

I was talking with a psychologist friend recently about pulling our pneumonia survey and delaying our Nigeria intervention and my concerns about our partners and staff and he stopped me and said, “Erin, are you talking to anyone about this?” He went on to say, “what you’re feeling is grief.”

It struck me in that moment that while I have been saying for many weeks now that our world will not be the same after this, saying it is not the same as feeling it and acknowledging it. We’re all experiencing huge changes in the way we’re able to live our lives right now. With or without family, with or without friends, all of us without our normal level of travel and cultural and personal connections and sense of purpose that comes with our work, a decreased ability to do work, a decreased ability to support our partners and staff in other countries, a need to engage more deeply with being alone or being intimately with a small group of people, and sometimes both. There’s a good chance we might not be able to travel internationally again this year. There’s a good chance we may see our offices and friends for only a few weeks at a time before having to go back to lockdown again and again. Our favorite restaurants and gyms and daycares and museums are closed and some may never re-open. There’s a good chance that by the time you read this, our projections of what the next few weeks or months will look like will change dramatically (now my standard COVID caveat).

I want to acknowledge that there is a profound loss associated with all of these changes, and it’s perfectly okay to feel that, however you’re experiencing it.

To anyone who is struggling to articulate their feelings these days, anyone who is feeling particularly agitated or lethargic or sad or frustrated or confused (I really think I covered the gamut here), it’s worth reading the following article in Harvard Business Review. I found it helpful to have someone tease out the contours of grief during these weird times and as encouragement to take a step back and a break, where possible. I would encourage everyone to take some time to feel things right now, because we are indeed grieving, even if we haven’t lost anyone to this disease yet—though some of us are experiencing that, too.

And with that, wishing everyone a Chag Sameach, Happy Easter, or good rest tonight and this weekend. Sending all the comfort and hugs I can muster across the interwebs. Please do reach out if there’s anything I can do for you.

Online amusement in the age of coronovirus

Let’s face it, social distancing is sad. #CancelEverything is sad. Isolation and quarantine are sad. But as long as we’re testing the capabilities of zoom and other videoconferencing software and agreeing to stay at least 6 feet away from people, we might as well see and hear beautiful things. The world’s symphonies, museums, aquariums, and artists are providing access to their joy online. This is a running list, please let me know if you hear of others and I will update.

While these are all free, please consider making donations to your favorite museums and arts spaces, and artists and musicians to lessen the burden on all in this difficult time.*


  1. You’ll get a form in the mail this week. Fill it online before April 1 with some really basic information on you and your family. This helps in so many ways.
    1. We get quick, accurate information to better assess how to allocate public services and political representation.
    2. It means a stranger doesn’t have to go knocking on your door in a month looking for you and possibly spreading this virus further.


  1. Berlin Philharmonic is offering free access to its online concerts. Use the code BERLINPHIL when you check out.
  2. The Metropolitan Opera will be HD streaming a recorded opera every evening at 7:30pm ET on their website starting March 16.
  3. The Seattle Symphony will be sharing videos, live-streams, and other broadcasts every evening at 7:30pm PDT.
  4. The Vienna State Opera has opened its video archives and will be playing ballets and operas online at 7pm or 5pm CET (registration required).
  5. Yo-Yo Ma is occasionally posting #SongsofComfort on twitter. This man is a national treasure. If you’ve never had the opportunity to hear him play, do try. I bawled like a baby when I got to see him at Tanglewood a few years ago.


  1. The National Gallery of Art has several exhibitions online and also has kid activities for download.
  2. Here’s a list of 11 other museums from around the world that you can visit online.


  1. The Monterrey Bay Aquarium has webcams available of its many different sea environments.
  2. Two baby eagles were born at Chincoteague National Park last month and they have a webcam! They post occasional videos and the eaglets are adorable.
  3. The Denver Zoo has a baby rhino and the videos they are posting are amazing.
  4. Big Bear Valley Eagle Nest Cam


  1. LibriVox offers a ton of free audiobooks for a range of books whose copyrights have expired, read by volunteers. I’ve been listening to Evangeline and it’s great
  2. Kindle and Apple Books also offer books for free whose copyrights have expired.
  3. If you have a library card, you can get Kindle Books through Overdrive. Here are instructions on how to set it up.
  4. You can download 300,000 books from the NYPL for free.
  5. Reach, Incorporated, a DC-based literacy non-profit, has its teen authors reading their books on their youtube channel and keeps updating with new content.
  6. Audible has a free 30-day trial for audiobooks and has made some titles free.


  1. Down Dog App is an at-home yoga and barre app that’s free to download and use through April 1.
  2. The Ballet Physique streaming studio is a Denver, CO-based barre studio with great instructors and has a 14-day free trial.
  3. Les Mills has a bunch of free HIIT workouts on their website for quick, hard cardio workouts.


  1. Scholastic has a bunch of kid activities for download while schools are closed, from pre-K to Grade 6+.
  2. FreeCode Camp has 450 online courses you can take for free.
  3. National Geographic also has a lot of kids’ homeschooling resources, Pre-K to 12.


  1. St. John’s Episcopal Church in DC is live-streaming church services

*Edited and restructured as the list gets longer. Keep sending them!

Women’s Employment in Afghanistan

A few years ago, a colleague at Towson got in touch and asked if I had any data from that refugee project I had worked on. He had an honors student with a lot of promise, he said, and wouldn’t it be cool to work on a paper together? I, having little time or incentive (as a then-independent consultant/researcher/quasi-academic/whatever) to publish, loved the idea of getting some of that hard work into the public sphere. When that student graduated and we had one paper published, we kept going, writing another paper with the refugee data, and then cajoled another student into working with us, too, on an idea about terrorism and women’s employment in Afghanistan.

As if I couldn’t be happier to work with this crew, the wins keep coming. I’m very excited to announce that this paper with Seth and Lauren Cahalan came to fruition and has been accepted at Oxford Development Studies! In light of a heavily publicized paper on time to publication in economics, it’s worth noting this paper saw its fair share of rejection, but ultimately was about 2.5 years from idea to paper acceptance. We’re very excited.

We ask two main research questions in the paper:

  1. Is the number of terrorist attacks and casualties associated with women’s employment?
  2. Is that relationship different for men?

Theoretically, we hypothesized it could go either way. If terrorist attacks are more likely to directly affect male mortality, then perhaps women need to enter the workforce to provide for their families. On the other hand, if more attacks make the perceived security situation worse, male decision makers may be less likely to permit women to work.

Overall, both men’s and women’s employment go down at about the same rate (though disproportionate against the baseline) when the number of attacks goes up, supporting the fear and security hypothesis. But women’s employment actually goes up when there are more casualties, supporting the replacement hypothesis. Ultimately this increase is small, and only holds up in rural areas for women in non-agricultural work

I love this paper for lots of reasons besides stellar coauthors. It was so neat to watch Lauren build her Stata skills through the careful matching of datasets and then really dig into the econometrics. Topically, it continues a really fun strand of my work on women in the labor force with the added context of conflict-affected space.

The abstract is below, but I think there are few important things to come out of this paper.

The first is that causality is hard here. We look at lagged variables to try to show how changes in number of events and casualties changes labor force participation down the line, but it’s likely that attacks aren’t random and may even be related to the number of women in the workforce.

The second is that there are some really interesting rural/urban differences, much of which appears to be driven by which sectors are already accepting of women, but could be driven by underreporting of agricultural work by women as we see in other parts of the world.

Afghanistan is one of the most dangerous countries for women and has the sixth lowest women’s employment rate globally. The low participation rate represents a large loss of potential economic activity and raising it could have large effects on growth. Security concerns are a key underlying barrier preventing women from working, but there is little work estimating the magnitude of a mechanism behind these effects. We address this gap in the literature by estimating the relationship between terrorism and women’s employment. We link a representative household survey, the 2015 Demographic and Health Survey (DHS), to the Global Terrorism Database (GTD), which catalogues terrorist attacks, locations, and fatalities. We find that the number of attacks per month in a given province is negatively associated in the following month to both men’s and women’s employment, yet the relative magnitude is larger for women due to their low employment rate. Conversely, we find that fatalities from these attacks are positively associated with women’s employment in non-agricultural sector in rural areas. This research illuminates a potential link between women’s employment and terrorism, thus adding to the ever-increasing knowledge of the costs of conflict.


Seth, Lauren and I had lunch on Saturday at an Afghani restaurant in DC and it was a reminder of how much fun these have been for me. Seth has turned these initial two projects with students into a veritable researcher training ground, with Slack groups and lab meetings and a steady stream of excellent students who are asking interesting questions, learning the ins and outs of econometric analysis, and generally killing it. If you have the desire to do research with undergrads–including and especially practitioners–check out the work Seth has done to create lessons learned from this group. Or! Bring him a research idea and ask him who the next Savannah or Lauren is; I’m sure you’ll get a good one!

Workforce and Skills Measurement Guidance

Some of my favorite problems to tackle are those of measurement. For instance, how do we think about social norms, but in a quantitative way? And, is it even possible to create a universal (or even multiple-country) indicator for women’s empowerment (been thinking about this one A LOT lately). Over the past year, I’ve also spent a lot of time thinking about how to measure skills, and specifically, how USAID should measure changes in skills as the causal effect of its workforce development programs all over the world.

Turns out, this question is not even remotely trivial, but I’m very excited to share some recently released guidance on exactly how USAID plans to incorporate some of the latest and greatest advances in measuring many (but not all) types of skills in its youth programming around the world. This is officially joint work with Catherine Honeyman of World Learning (a delightful collaborator if I do say so myself), but also benefited enormously from a wide consultative process with stakeholders across USAID in DC and around the world as well as implementing partners and peer research organizations working on issues of skills development for youth.

As background, USAID works with youth programming in 60 countries and workforce development programs in 30 countries. All of these programs are working to impart technical, vocational, social-emotional, digital, reading, and mathematics skills (and more) to various program participants. And all of them (yes ALL of them) are required to report back to USAID on their project in various ways.

Alongside, there has been a ton of work recently on how to measure various skill development. As we learn more about the importance of “skills” to labor market outcomes like perseverance or grit and hard-working or dependable, the more necessary it becomes important to measure them in a way that has meaning across different spaces. After lengthy reading and consultation, we identified two skills or groups of skills that were selected to be “standard foreign assistance indicators”–percent individuals with improved soft skills and improved reading skills–and three that have been designated “supplemental indicators,” percent of individuals with improved math skills, percent with digital literacy skills, and percent passing a context-relevant technical skills assessment. More precise wording is in the table, with a much longer justification in the how-to note itself.

Screen Shot 2019-10-30 at 5.47.13 PM

In short, though, it’s really tough to standardize measures across all those countries and all those contexts.

I won’t lie to you; we did not find all the answers. However, I think the note provides an excellent jumping off point for ongoing conversations on how to measure important workforce development programming outcomes and highlights important gaps in our understanding of how to measure these outcomes in a consistent and useful way. Please reach out if you have ideas!

We also put together a fun (measurement is FUN, I SAY!) FAQ that you can read here.

Paper Acceptance Day!

When I was working for myself, I had a lot of conversations with development practitioners that started with an element of intuition, and ended with one of frustration:

“We think X thing is happening, but we’re not really sure…” or “we think we could do this thing we’re doing better…” and then “…but we don’t have the technical skills on staff,” or “…we’ve asked like 85 consultants to do this and no one wants to help us.”

Usually, these requests came in fields that I (at the time) was certainly no expert in. Livelihoods analysis, tertiary child protection systems, community-based disaster recovery management programming, etc. They almost always meant diving into new literatures and exploring problems of measurement and data use in ways that were novel and particular and had to somehow hew to the very real limitations faced by those implementing these programs: namely, time, but also expertise and sheer number of people.

These were also the projects that I have loved taking on, taking ideas that are firmly based in practitioner experience and helping make sense of how to use data and econometrics and innovative measurement and descriptive tools in ways that are practical and implementable. Getting to work with inquisitive and innovation-oriented development professionals is one of the things I love best about my job.

I also love when we get to share these projects. I’m super pleased to announce that one such project (or rather, our write-up of the work) has been accepted at World Development!

With Richard Barad, a GIS specialist, and Chris Hillbruner, then Deputy Chief of Party for FEWS NET, we piloted a method to use large-scale household and remote-sensed data to describe livelihoods in a way that was compatible with participatory methods and comprehensible to practitioners in the field. Quick summary:

  1. Collect HEA-compatible variables from large scale household data like the LSMS from the World Bank (agricultural activities, livestock ownership, etc.)
  2. Spatially interpolate these data!
  3. Identify relevant remote-sensed data that pertains to livelihoods (soil quality, ground cover, distance to markets, passable roads, etc.)
  4. Export all the data to grid squares and perform principal components analysis and cluster analysis
  5. Make pretty maps! Go Forth! Discuss! Use to understand food emergencies!


I stress that it’s both because a lot of people we spoke with are pushing forward some neat and innovative methods to do something similar, but in a way that was difficult to implement. What we propose is certainly still in beta, but we think can be put in place quickly to inform emergency situations and to validate other types of data collection.

Abstract below, and a yet-to-be-finalized copy is here and tweetstorm is here.

Understanding livelihoods patterns is a key component of food security and poverty analysis. The Household Economy Approach (HEA) is a leading method of conceptualizing, organizing, and analyzing information on livelihoods systems that is widely used within the food security analysis community. This approach is typically informed by data collected using qualitative methods. However, the increasing availability of large-scale household survey datasets presents an opportunity to explore the degree to which these data can be used to strengthen HEA analysis. Here, we present the results of a novel pilot study that uses large-scale household survey data to create livelihoods products for Nigeria, using a combination of spatial interpolation, principal component analysis, and cluster analysis. We show how these techniques can leverage existing data to create low-cost maps of quantitatively described livelihoods that are stable over time and conceptually consistent with products derived using traditional methods.  We also outline future research for how to incorporate these outputs into practitioner analysis.

*Paper acceptance day was actually yesterday

Child Marriage and Integrated Programming

I spent part of this morning (very early this morning) speaking to a joint meeting of experts on child marriage from UNFPA and UNICEF from the South Asia region. Earlier this year, I coauthored a directed literature review of white papers and scholarly literature from 2014 to 2018 on child marriage in South Asia and was asked to share some thoughts and findings.

The paper is forthcoming, and I’ll be sure to post it when it’s available, but presenting the findings very briefly this morning to my colleagues forced me to think about big picture lessons. In doing so, my thoughts coalesced on one important point. Namely, as the research around child marriage has evolved, so too has the thinking about how to address the problem, changing from a single-minded, narrowly focused assignment of poverty as the root cause and thus poverty alleviation as the solution, to nuanced understanding of the heterogeneity of the experiences of child brides and the deep cultural, social, political, legal, and normative institutions that shape both the practice and responses to programming designed to stem it.

Viewing child marriage as a systems problem, one that is subject to feedback loops and responsive in ways we may not fully expect or plan for (externalities or unintended consequences) is an increasingly recognized and necessary way to view programming. In addition, the increasing recognition of the subnational heterogeneity and increasing availability of quality data and studies on particular groups and their child marriage institutions is an important step in better placing the problem in its social and political contexts. I’m excited to see continued work in this area that recognizes, for instance, how cash transfers may increase secondary school attainment and delay the very youngest child marriage, but also increase marriage at age 18, or how outlawing child marriage may influence sex selection, or how to make girls’ empowerment and agency interventions effective and safe for girls in an environment with strong, persistent patriarchal norms, or how norms that appear to be changing and flexible may be more pliant in the face of macro-economic changes and individual-level shocks. Developing methods and evaluation tools to test the efficacy of integrated programming is a frontline problem for child marriage programming, as well as other social programming.

All said, there are so many exciting areas of research, and I love being deeply ensconced in these debates with scholars and practitioners. Paper to come shortly!


Some Personal/Professional News

Fall is marked by change, and while the aspens are turning yellow all over Colorado, it seems it’s my turn, too. I am very excited to announce that I am joining Results for Development as an Economist/Researcher as of today! R4D guides development projects around the world in the health, nutrition, governance, and education spaces. I will be working with their Rapid Feedback Monitoring, Evaluation, and Rapid Learning Team to bring the technical side of evidence generation and evaluation design in-house, but also supporting projects across the organization in formative research and data analysis with a keen eye towards policy and program design.

It is a phenomenal opportunity, but it is also one that necessitates leaving Colorado and moving to Washington, DC. I am heartbroken, obviously, to leave my family and my mountains.

I am also sad to be leaving behind my own consulting work. I have worked with amazing partners all over the world for the past three years on so many interesting questions. I am grateful to everyone who hired me, worked with me, and let me try out my nutty ideas. FEWS NET, the IRC, Bead For Life/Street Business School, Promethean Community, Al Mokha, are just a few of my awesome partners. I have learned so much from of you, and I think we have accomplished some great things, too.

One of the best things about this R4D job is that my actual job is pretty similar to what I have been doing. I’m handing over much of the administrative work, the sales-y parts, and the risk to a larger organization. (There might even be someone to help me with data cleaning once in awhile! What kind of PhD turns that down?!). In turn, I get more time to think deeply about econometrics, about research, about survey experiments, about evaluative design, about policy, about programming, about equity, about research ethics, about measurement. Can you tell how excited I am?!

I am. I am super excited to join this extremely talented and creative team and take on a new portfolio of work. So, let me know if you’re in DC. Or come visit!

(Also, because several people have asked, I am keeping my house–it’s for rent starting mid-December if you know anyone who wants to live in an adorable 1910s duplex in Washington Park West.)

Measuring Difficult Things

I’m super excited to be giving a talk on Wednesday at the One Earth Future Foundation, a local, homegrown, interdisciplinary think tank here in Colorado working on all sorts of issues related to sustainability and making sure that we can keep living on this planet.

Because I don’t get to give talks very often, I may have gotten a bit excited and planned to talk about, well, basically everything I have done the past few years (and frankly a lot that I’m hoping to do in the next few!). The abstract is below. Really looking forward to a fun discussion!

Measuring Difficult Things

Social norms, unobserved characteristics, and hidden statuses can exert significant effects on outcomes ranging from domestic violence to armed group participation to social capital and network formation. In quantitative analysis, not being able to account for these factors can lead to biases in our estimation and thus incorrect or incomplete policy prescriptions. Even when we are able to see that these factors exist, finding a way to include them in our estimation is often limited by issues of measurement. In this talk, I give an overview of some my research on how to measure difficult things, beginning with social norms and intergenerational effects in a study of refugees, list experiments, and how to generate representative samples of hidden populations. I explore challenges in measurement, offer lessons from the field, and describe prospective work that aims to continue deepening our understanding of how to approach measuring difficult things.

Just how much do adults influence their kids?: Nyarugusu edition

In my first paper with Seth Gitter and Savannah Wilhelm on refugees in Tanzania’s Nyarugusu Camp, we showed that, among child respondents, there appears to be a strong social norm around reporting violence at school. Children believe their peers will report victimization, kind of regardless of the things we might think would matter. Boys and girls think their peers would report at roughly the same rate, and the type of violence doesn’t seem to change their perceptions much either.

A natural question to come out of this is whether these attitudes and perceptions are coming from their parents. Luckily, I designed this survey exactly to get at that kind of question. A novel and important part of this study was to survey linked child-parent pairs, so we can compare children’s responses to their parents. We have a second paper, now a Towson University working paper, out to deal with just that question. Here we’re still using the vignettes (one-line short stories of hypothetical situations), so there is some measure of uncertainty given the randomization of the vignette characteristics, and we’re still concerned with reporting norms, not direct victimization, but we have some pretty interesting findings.

First, parents’ and kids’ answers are positively correlated, so while we cannot necessarily identify the mechanism, there does appear to be some intergenerational transfer of attitudes or social norms.

We use the error terms from the parent regressions (basically the variation in the observations that cannot be explained by the vignette characteristics) to represent this sort of nebulous “parent” effect. When we put these error terms into the kid regressions, there is a large and statistically significant effect on those terms. In other words, the unobservable stuff that determines a parent’s answers is clearly and significantly related to his or her kid’s answers, controlling for both the vignettes that children hear and their demographics.

But they’re not perfectly correlated. We wouldn’t expect them to be, given the vignette randomization, so we rely on grouped vignette characteristics to tell us a little more. Parents saw some violence type distinctions where kids did not; sexual violence (teasing or touching) was perceived as more likely to be reported than other types of violence.

Finally, the specific way the surveys were set up allow us to dig a little deeper into parent thinking. Parents were screened at the beginning of interviews and asked to identify and give consent for one of their children to be interviewed immediately after their own survey. Largely, this was the child in the age range who happened to be at home and we believe this to be relatively random (school is half day, most adults had several children in the age range of both genders, parents were chosen via random walk, etc.). Gender and age are balanced for the child sample.

So, given that the child was selected before any data collection took place, we examine the data to see if parents answer vignettes differently depending on gender of the victim, conditional on the gender of the child who was interviewed after them. Spoiler alert. They do. Parents who selected a boy child to be interviewed and heard vignettes about boys were more likely to think the hypothetical victim would report violence. This isn’t true about girls. We think this effect might be something like priming, that the child to be interviewed might be at the top of the parent’s mind while he/she is taking the survey.

Seth and I will be presenting this paper at the MIEDC conference at the University of Minnesota at the end of April. We’re super excited to share it with everyone!

Pop the Champagne: An ode to great coauthors

Yesterday, I took a quick career mentoring phone call with a post-doc at CUNY Futures Initiative. After I had given her the rundown of what I presume to be my extremely fascinating job and life, the post-doc asked me how I managed to keep one foot in academia with everything else I had going on.

I have great coauthors, I told her.

And it’s so true.

Yesterday, I received notification that the first paper using the data I collected with the IRC in Tanzania’s Nyarugusu refugees camp was accepted for publication at Migration Letters. Time to break out the champagne, but first, lemme go update my CV real quick…

K, I’m back.

As background, the IRC hired me in 2015 to conduct a two-country, two-stage investigation of social norms of violence in schools in refugee camps. We started with focus groups and key informant interviews with Burundian and Congolese refugees in Nyarugusu camp in Tanzania, displaced Iraqis in Arbat IDP camp in Sulaymaniyah, Iraq, and out-of-camp Syrian refugees in urban Sulaymaniyah. These results were used to inform a survey of 300 randomly selected child-parent pairs in Nyarugusu. We were unable to complete the second phase of research in Iraq.

The survey covers lots of things, but specifically in this newly accepted paper, with Seth Gitter at Towson University and his senior thesis student, Savannah Wilhelm (now all graduated and grown-up, *sniff*), we examine norms around reporting violence to authorities and characteristics of children that may predict reporting in Nyarugusu. I love these data for many reasons, not just because they’re my baby, but because I think they’re a unique and novel contribution to an understudied field.

We survey young kids in an environment where very few large-scale surveys have been done (we have ~300 child-parent pairs, though this paper only looks at the children’s responses); we talk about violence in ways that are designed to mitigate retraumatzation and reduce under-reporting; and the findings are really interesting. Gender, either of the respondent or the hypothetical victim, has no significant bearing on the willingness of kids to report violence. In an extremely unstable, vulnerable environment where kids have fled violence, or grown up with stories of their parents fleeing violence, there appears to be a normative expectation that kids report victimization, and they feel comfortable doing that at school.

I think that’s huge.

This is also the first paper for which I’ve gone through the whole process with a student co-author, which was extremely rewarding. I honestly cannot gush enough about how great it’s been to work with Seth and Savannah (and now Seth and Lauren, and hopefully more to come!). Abstract is below. A second paper using the parent responses as well is in the works and hopefully we’ll get to share that soon, too.

Reporting Violence Against Children: Social Norms in Nyarugusu Refugees Camp

There is substantial evidence to show that the two million children living in refugee camps are vulnerable to violence, although little is known about under what circumstances children will report that violence, complicating efforts to reduce vulnerabilities. We presented 300 children in a Tanzanian refugee camp (Nyarugusu) with hypothetical vignettes regarding a victim’s response to violence. Vignette characteristics were randomized (e.g. victim, perpetrator, and location) to test what factors influenced reporting. Respondents believe the victim was more likely to report violence at school or adolescent perpetrators. Surprisingly, we find no substantial difference based on victim or respondent’s gender.