2018 Project Descriptions

Women's Rights, Fetal Rights |
Jocelyn Viterna (Sociology)

Project Mentors: Jocelyn Viterna, Professor of Sociology

Once won, new legal rights are seldom lost again. This is a truism in social science research. And yet over the past two decades, a handful of the world's nations (including the United States) have begun to systematically reverse women's reproductive rights. Why have women's gains in reproductive rights been overturned in some countries but not others? And what are the consequences of these rights reversals for women's lives?

Between 1989 and 2009, six Latin American nations took legislative action to tighten their already-strict abortion laws. Previously, these nations allowed legal abortion only in limited circumstances--typically when the life of the mother was at risk, when the pregnancy was the result of rape, or when the fetus had deformities incompatible with life. With the passage of new legislation, however, these countries removed all abortion allowances. No abortions, no exceptions; not even when a woman's life is endangered by her pregnancy. This backslide in abortion policy has in turn had powerful consequences on women's lives. On one hand, women regularly die because the absolute abortion ban forces doctors in public hospitals to withhold a lifesaving abortion in situations where a pregnancy threatens a woman's life (for example, when a pregnancy puts too much strain on a woman's heart and causes heart failure). On the other hand, in at least two cases--El Salvador and Mexico--the absolute abortion ban has also been accompanied by a swift uptake in the prosecution and incarceration of women for abortion and the so-called "homicide" of fetuses. In El Salvador, for example, more than 30 women have received 12-40 year prison sentences for the "attempted homicide" or "homicide" of their fetuses. Most tragically, the legal, medical, and forensic records make clear that these women serving 30 year prison sentences suffered naturally occurring stillbirths; they did not take any action to abort their pregnancy. Perhaps not surprisingly, the women targeted by these extra-judicial prosecutions, and the women left to die in public hospitals, are overwhelmingly very poor, and very marginalized.

I have spent the last four years investigating these trends in Latin America, and most centrally, in El Salvador. Specifically, I have been working to win new trials for wrongfully-incarcerated women, and I have been interviewing Salvadoran doctors to learn how the total abortion ban affects their ability to practice medicine. In the process of writing about Central America, however, I have come to realize that what is happening there seems to be linked transnationally to similar processes unfolding elsewhere around the world. To illustrate: it appears that we are now seeing a sharp increase in the number of women incarcerated for crimes against a 'fetus' here in the United States as well!

My goal for this summer is to investigate whether and how the reversal of reproductive rights I've already documented in Latin America is connected with other cases around the globe. This investigation will take two approaches. First, I will ask the student who works with me to compile several existing data sets into one new data set that captures how women's reproductive rights have changed over time in nations around the world. This dataset will be made available to the public. Second, I will ask the student to investigate whether and how these pro-life "successes" (as identified in the data set) correlate with the outreach mobilizations of a few well-known transnational pro-life organizations (like Human Life International), likely by analyzing their publicly available tax documents and annual reports. Third, if there is time, I will ask the student to work closely with me to choose a small subset of countries (5-7) in which we will complete deeper investigations about reproductive rights, likely utilizing archives and newspaper articles. If the situation calls for it, the professor and the student may also travel to one of these nations to do on-the-ground investigations. If the summer research project goes well, and if the student is interested, there may be opportunities to continue the research project into the academic year.

Students working with me will start each weekday in a team meeting to discuss what work was accomplished the prior day, and what work will need to unfold in the hours ahead. Students will gain specific research skills (creating an international data set, developing a coding system, analyzing archival evidence, using qualitative software like N-Vivo, writing clear and effective reports, analyzing legislation and court documents), and more generally, will learn how to investigate an academic puzzle through the development of potential hypotheses and the collection of data to test those hypotheses. Depending on where the research takes us, and depending on the students' interests, the student(s) may also gain experience applying for approval from the institutional review board, and traveling to conduct interviews abroad. Fluency in a foreign language (Spanish, French, Kinyarwanda, Swahili, Tagolog, Polish, etc) would be helpful and could also influence the direction of the research. My style is perhaps best described as an apprenticeship; I like working side-by-side with students as we develop the initial steps collaboratively (e.g., let's read through these reports together and collaboratively develop a coding schema), until we arrive at the point where there are clear steps to be taken individually (e.g., let's each take and code three interviews, and then we'll re-group to discuss how the coding schema worked and what modifications should be made).

In addition to learning foundational research skills, students will be listed as co-authors on the data set they construct, and on any reports or papers they help write over the course of the summer.

Skills Needed: The ideal student will be a critical thinker, a careful reader, a clear writer, and will care about getting the details right. All other research tasks (creating coding systems, coding with N-Vivo, etc) will be learned on-the-job. Fluency in another language is a definite plus, but not necessary.

The Gender and Race of Armed Citizenship | Caroline Light (Women, Gender, & Sexuality)

Project Mentors: Caroline Light, Director of Undergraduate Studies and Lecturer on Studies of Women, Gender, and Sexuality

How do our nation's shifting rules around firearm ownership and carry influence the performance of citizenship? How are the categories "armed citizen" and "law-abiding citizen" coded - sometimes implicitly - with ideals of gender, race, class, and sexuality? What kinds of assumptions frame the appeal to lethal self-defense as a constitutional right, the belief that "good guys with guns" will protect us from "bad guys with guns"? How often and under what circumstances do women, trans gender, and gender-nonconforming people become criminalized when defending themselves from intimate partners and others occupying relatively dominant subject positions? Finally, why do so many believe that guns will protect the weak from the mighty when social scientific research suggests otherwise?

The research for this project will be conducted using multiple different methods and through in a number of different "archives." We will begin by gathering/surveying existing data on gender, race and gun homicide, and we will address the reasons why some vital data sets are missing or difficult to access. We may start by reading some shared texts (and exploring existing data sets) on the intersections of gender, race, sexuality, and violence more generally.

For quantitative data, we will explore the National Violent Death Reporting System (NVDRS) and the CDC's Web-based Injury Statistics Query and Reporting System (WISQARS). Researchers will gain familiarity with these databases and will use survey-making software, like Qualtrics, to code the qualitative data on homicides nationwide.

An average day may have researchers constructing surveys to code "big data" sets for insight into (for example) the circumstances under which women use firearms in self-defense against violent intimate partners. Or researchers might scan state by state legal codes to develop a map of differing "castle laws" (the laws that allow for violent self-defense, without retreat, in certain places). Researchers may also investigate local court homicide proceedings to determine how the criminal justice system adjudicates different people's claims of self-defense.

We will start the summer with frequent meetings where we check in and plan each day's work. Depending on the researcher's level of familiarity, we might start with some shared readings to set the conceptual stage (e.g. Readings on "intersectional" violence, an introduction to the databases we'll be working with, some basic overviews of the kinds of questions we want to ask). My hope is that the researcher will gradually gain confidence in asking difficult questions of the existing literature, and that these questions may help direct our later research.  Once we have gotten started and established our work expectations, we might meet weekly to check in.

Skills the researcher can expect to develop may include intersectional feminist and critical race analyses alongside some basic quantitative coding skills. The researcher will gain familiarity with (and hopefully confidence in) asking big questions of often taken-for-granted "conventional wisdom" about violence and vulnerability and will learn to look for gaps or silences in the "evidence" by which we address questions of public and private safety.

Skills Needed: Ideally, a researcher should have an open mind and an abundance of independent thinking and questioning skills. An ideal researcher should be at least somewhat familiar with the basics of gender and ethnic studies, perhaps taken one or two relevant courses in feminist/queer and ethnic or African American studies. Above all, the ideal researcher will possess an interest in looking beyond the surface rhetoric by which our culture (and our legal system) allows/justifies violence in particular circumstances, and why.


Language Acquisition |
Jesse Snedeker (Psychology)

Jesse Snedeker, Professor of Psychology

Language is not one representation but many. A spoken utterance can be characterized as a string of phonemes, a nested set of prosodic phrases, a series of lexical items, a hierarchically-organized syntactic tree, a configuration of semantic relations, or the impetus for inferences about the speaker's intentions. A fundamental challenge for the psychology of language is to understand the relations between these representations: the degree to which they are distinct, the ways in which they constrain one another, and the role that these connections play in language acquisition. My lab explores these questions with a primary focus on semantic representations and their relation to syntax and pragmatics. Semantic representations are central to cognitive science because they provide a window into our generative conceptual capacity. Language allows us to combine concepts from diverse cognitive domains. Understanding how meaning is encoded in language is central to understanding conceptual combination.

Our approach to these questions is experimental and developmental. The study of semantics has been based largely on the judgments of trained linguists. Where these judgments are unclear, controversial or uninformative, theories diverge. By using a broader range of methods with diverse populations, we can gain additional insight into the processes that give rise to meaning and the representations they create. Developmental work is critical for two reasons: 1) Adult language processing is complex and interactive, by observing language at an earlier state we may gain a deeper understanding of its architecture; 2) Developmental studies allow us to explore the relation between language and conceptual development. If the semantics of external languages build on a prior language of thought, then we would expect many aspects of semantic structure to develop early and constrain language acquisition. In contrast, if external language is the sole mechanism of domain-general conceptual combination, then we might expect conceptual and linguistic development to be closely yoked.

BLISS fellows will be given the opportunity to work on one of several projects investigating how children acquire and process language, ranging from how children learn the meanings of novel verbs to how children generalize grammatical and semantic representations across sentences to how children understand the referents of proper names. Fellows will be assigned a project based on their interests and will be involved in all major steps of its lifecycle: preparing study stimuli, conducting literature searches, recruiting participants, coding, entering, and transcribing data, and presenting their results. This will allow fellows to work closely with their mentors to make a contribution to the design of the study and the interpretation of its results.

The typical day in the life of a BLISS fellow in our lab varies depending on the specific project to which they are assigned and their progress over the summer. However, most fellows can anticipate spending 2-3 hours each day running participants, 2-3 hours recruiting participants, and the remaining time in the lab on tasks such as reading literature relevant to their project, attending research meetings with their mentor, or coding and entering data.

This project will be part of the summer internship program that our laboratory, together with others in the department, organizes every year. The BLISS Fellow(s) will be paired with a graduate researcher, and will be involved in all major steps of psycholinguistic research. They also participate in a weekly Reading Group to talk about 1-2 journal articles with other interns, while 1-2 research mentors moderate the discussion, in weekly Lab Meetings, and in weekly meetings with their graduate student mentor and Dr. Snedeker. In short, BLISS Fellow(s) get a chance to experience firsthand how scientific knowledge is actually produced, potentially helping them to decide whether to pursue graduate studies and a career in science, or not.

Skills Needed
An active interest in working with children, some background in linguistics and psychology, a high degree of independence, problem-solving skills and the ability and interest to quickly acquire new skills.

Abstract Thought in Humans (and Other Animals) | Susan Carey (Psychology)

Project Mentors: Susan Carey, Henry A. Morss, Jr. and Elisabeh W. Morss Professor of Psychology

We study the human capacity for abstract conceptual thought (a unique phenomenon on earth). We currently have two different research thrusts:

  • Q1. Is the capacity for abstract, language-like, combinatorial thought unique to humans (cross-species studies), and does it arise in human development only upon learning language (studies with human infants)?
  • Q2. What is the role of executive functions (cognitive control, inhibition, working memory—many frontal lobe functions) in the construction of theoretical knowledge?

BLISS fellows hear descriptions of 10 to 12 ongoing projects, and choose a particular project to collaborate on, under the joint mentorship of a graduate student or postdoc, and me (Professor Carey). They master the relevant background literature, participate in the design of the new study, learn all aspects of running the study, coding and analyzing the data, and bringing the data to bear on adjudicating between theoretically important hypotheses. They are part of a larger internship program in my lab, involving undergraduates and recent graduates from all over they world, and participate in reading groups and activities for the whole internship program, and present their findings in a poster session at the end of the internship program.

The ultimate goal of the internship is to engage students in bringing empirical data to bear on theoretically important issues, to see the process of making a scientific argument. This experience will inform students’ career choices (i.e., does he or she want to become a research scientist, and if so, are the questions within cognitive science of interest?). Whatever the outcome of these decisions (at this stage of life, it’s important to discover what one does not want to do as part of discovering what one does want to do), this experience make students more critical consumers of popular science reporting in the press, and of course work in the social sciences.

The proximate goal of the internship is to engage students in the fun of the collaborative activity that is science!

Skills Needed: An interest in cognitive science (some background in philosophy, computer science, experimental psychology, linguistics,or history of science).

The Neurodevelopment of Flexibility in Learning and Decision-Making in Adolescence | Leah Somerville (Psychology)

Project Mentors: Leah Somerville, Associate Professor of Psychology; Juliet Davidow, Postdoctoral Fellow

This project is aimed at exploring how and when adolescents can use previously learned information in new ways to solve problems that they haven't previously encountered before. To unpack this, we are investigating how adolescents learn and remember different kinds of information. We are also probing how the changing brain influences these different cognitive processes in children, adolescents and adults.

Because this study will involve testing children and adolescents (ages 11-17 years old) summer is a very important time to collect data, while many potential participants have more ability to come into the lab to take part in studies. We need BLISS fellows to help out with the collection of brain imaging (fMRI) data. This will involve being present for testing child, adolescent, and adult participants at the Center for Brain Science. Research assistants will help prepare the scanner environment before the testing begins, administer cognitive tasks and standardized clinical measures, help in the control room while the participants undergo brain imaging, and help with pre-testing and post-testing tasks including preparing testing materials, and data entry. Research assistants will help with the "quality control" or checking of the data, which will involve learning how to reconstruct, view, and assess the brain images.

The faculty mentor meets with the research assistants every week, through one-on-one meetings to discuss project progress as well as through lab meetings where all the lab's ongoing research is discussed, journal articles are work-shopped, or general research skill tutorials are held. This project is being led by a postdoc in the lab who has a record of successful undergraduate mentorship and extensive experience with brain imaging testing procedures and data analysis. She will primarily oversee the day-today activities of the research assistant. The BLISS fellow will learn many practical skills that can translate to work outside of the lab as well including critical thinking skills, basic computer software use and some programming skills. This would be a great introduction to the lab for a student who might be interested in doing a thesis project in the future.

Skills Needed: The only skills that we expect students to bring to the lab are attention to detail, punctuality, and a respectful, professional, attitude. All the lab related skills will be learned through hands on activity.

Cost-Effectiveness of Strategies for Preventing Suicide in the Military | Matthew Nock (Psychology)

Project Mentors: Matthew Nock, Professor of Psychology; Heather S. Pixley, Research Manager

As clinical psychology researchers, we are motivated by a goal to reduce suffering. As such, our natural instinct is to expend all our resources to help every individual achieve the best life possible. The problem with this instinct is that resources are limited--all of the time, energy, and money we use to help one individual necessarily displaces resources that could be used elsewhere, possibly to help even more people to a greater degree. Economic analysis, of which cost-effectiveness analysis is one type, is a tool for determining how we can use our resources to do the most good possible for the greatest number of people.

Due to growing concerns about suicide rates among service members, the Department of Defense (DoD) has funded an enormous epidemiological project called the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) that aims, in part, to develop algorithms for predicting which soldiers are at highest risk of suicide. Researchers working on the STARRS project have already found some success in identifying subsets of soldiers who are at elevated risk. But now the Department of Defense wants to know what can be done to intervene to prevent suicide attempts and deaths among those at high risk. The goal of the current project is to address just that question using the tool of cost-effectiveness analysis. We aim to comprehensively quantify the costs and benefits (to the DoD, to soldiers and families, and to society) of a variety of preventive interventions so that the DoD can select the optimal strategy or set of strategies. This project could have considerable influence on the DoD's policies and consequently could save lives.

The primary role of a research assistant will be to find estimates of the costs and benefits of a variety of interventions, and to carefully document how the research assistant arrived at those estimates. For example, in addition to finding the best available evidence for the efficacy of each prevention strategy, we will need to know things like the cost of recruiting and training a new soldier when one is injured or killed; the likelihood that someone who has attempted suicide will do so again and the frequency at which they will do so across their lifespan; details of the military healthcare system such as the typical cost of provider time; and even information about how funding allocation decisions are made within the DoD. Finding this information will require considerable creativity, flexibility, and resourcefulness. The RA will not only need to be able to read scientific papers and identify the relevant statistics, but also figure out the appropriate sources for other information, which may require consulting others such as librarians. The RA will gain intimate familiarity with the suicide prevention literature, the healthcare system, and cost-effectiveness analysis, knowledge that will enable the RA to have a positive impact in his or her future career.

The student will be located in the Nock Lab and mentored by Professor Nock, Heather Pixley (the lab manager) and 2 or 3 Postdoctoral Fellows who are also working on the project; there will be regular project team meetings, as well as the opportunity to attend Lab Meetings, Lectures, and other relevant events.

This project will provide a unique experience in that clinical psychology does not normally consider cost effectiveness in such detail, and it will give the student a new perspective on the healthcare world in general.

Skills Needed: Thoroughness and attention to detail; moderate statistical literacy; knowledge of database usage and literature search strategies; willingness to "think outside the box" and consult a variety of individuals to find information.



Performance of evidence-based adult psychotherapies compared to usual clinical care | John Weisz (Psychology)

Project Mentors: John Weisz, Professor of Psychology; Katherine Corteselli, Graduate Student; Spencer Evans, Postdoctoral Fellow

Evidence-based psychotherapies (i.e., those showing beneficial effects in randomized trials) are prominent in current psychological science. They have been developed to improve on usual clinical care of patients. But do they actually outperform usual care overall, and is the answer to that question different for different treated problems--anxiety, depression, and antisocial behavior, for example? We will seek to find out via a meta-analysis of randomized trials in the scientific literature over the past 20 years in which adult patients were randomly assigned to an evidence-based psychotherapy versus usual clinical care. Our previous published meta-analyses on this topic, focused on child and adolescent psychotherapy, published in The American Psychologist and in JAMA Psychiatry, have been widely cited, and we believe this new meta-analysis focused on adult psychotherapy will generate even greater interest among clinical scientists and other readers interested in psychotherapy effects.

Psychotherapy research has produced many empirically-tested evidence-based psychotherapies (EBPs) for adult psychopathology, developed to improve upon usual clinical interventions. Advocates argue that these should replace usual care; but do the EBPs produce better outcomes than usual care? We will seek to answer that question via a meta-analysis. We will begin with a systematic search spanning the past 20 years, using standard search engines, for psychotherapy trials in which patients were randomly assigned to an EBT or to usual care, with outcomes assessed via psychometrically sound measures.Effects will be calculated for each study, and a synthesis of the effect sizes across studies will be used to gauge the overall benefit of EBTs relative to usual care, and to identify moderators of treatment benefit.

Students participating in the project will gain experience in all the primary aspects of a meta-analysis. We will work together to conduct a systematic search for studies that meet meta-analytic criteria, and we will create a screening system and a coding manual for coding the relevant characteristics of each study. The students will practice using the coding manual until they achieve acceptable inter-coder reliability, and then code the studies that have been identified through the search process. The resulting codes will be entered into a database, and study analyses will be carried out, guided by statistical experts in my lab, but with the students participating. Writing up the findings will the the last step.

A "day in the life" for the students will look busy! Each of the tasks described in the previous paragraph will involve considerable learning followed by hands-on activity that includes--in phases--searching the scientific literature via standard databases, developing the coding manual, training and practicing to achieve inter-coder reliability in the codes contained within it, coding the studies identified through the search, creating a meta-analysis database that will be used for study analyses, assisting the statistical experts in carrying out the analyses, and assisting the mentors in writing up the findings for possible journal submission.

I will follow the mentoring plan I have followed with previous BLISS students, blending individual and group meetings with conference calls and frequent emails. In addition, a great deal of preparing, reviewing, providing feedback, and re-writing is required to prepare the coding manual, and this becomes a highly interactive process, as we learn together which codes and which code definitions are working and which are not, and then decide together which changes are required. I will provide direct guidance--together with my co-mentors--throughout the search process, to ensure that the appropriate databases are searched, and the appropriate search terms applied, and to develop the screening criteria needed to rule studies in or out. The co-mentors and I will work closely with the BLISS student(s) in developing and refining the coding manual, carrying out the waves of practice and testing needed to establish inter-coder reliability for all the codes, and then guiding application of the coding process to all the included studies. We will also work together to design the meta-analytic database to be used in the statistical analysis, and in the waves of writeup required to produce the meta-analysis manuscript describing the goals, procedures, and findings.

Skills needed:Two or more courses in abnormal or clinical psychology, ideally including some content related to psychotherapy and psychotherapy research. Previous coursework in statistical analysis of study data. Some previous involvement in a previous meta-analysis would be especially valuable.

Harvard Decision Science Laboratory Internship | Julia Minson (HKS Public Policy)

Project Mentors: Julia Minson, Assistant Professor of Public Policy; Alki Iliopoulou, Lab Manager; Gabe Mansur, Research Computing Specialist; HDSL researchers

The Harvard Decision Science Laboratory is a university-wide bio-behavioral research facility. Unlike a traditional lab with one investigator, our lab supports many researchers from all across Harvard, exploring their behavioral research projects. We are an emerging hub for the conduct of experimental research in the behavioral and decision sciences.

Researchers from all across the university, and at all levels of the scholarly community—from senior faculty to undergraduate researchers—use our facility to investigate how emotion, neuroscience, and cognitive processes combine to shape human judgment and decision-making. In addition to working on Professor Julia Minson's research on judgment, collaboration, and social influence, a BLISS fellow will assist the lab staff with research tasks for other investigators involved in the lab; this includes screening, recruiting, and communicating with subject participants in lab experiments; setting up and running experiment sessions; coding data; providing feedback on research ideas; as well as other tasks as needed. Day-to-day at HDSL can be very different depending on current needs—you may be at the front desk checking in and out subjects or in the lab room running experiments or working on programming or testing an experimental protocol.

Specifically the BLISS fellow may have opportunities to work with experimenters in the conceptualization and design of their experiments; in programming these designs using standard tools for subject interaction (for example, MediaLab, z-Tree, E-Prime, MatLab, Qualtrics); in conducting the experiments and interacting with subjects in the lab; and in compiling and analyzing experimental data. In addition, as part of their basic training in the tools of the lab, the fellow will be trained on the use of physiological monitoring systems for behavioral experiments. This “Physio Boot Camp” involves training in the correct placement of sensors, assuring the clarity of a signal for data collection, skills for respectful interactions with subjects, and use of software tools (specifically BioLab) for the analysis of physiological data.

Working at HDSL will provide the BLISS fellow with a uniquely comprehensive introduction to experimental methods and tools. HDSL BLISS fellows become acquainted with a wide variety of investigators, research questions, and experimental methods. The experience is one particularly suited to students interested both in learning experimental technique and eager for an exposure to a variety of fields, with a view to building skills and exploring fields of potential interest through direct engagement with researchers.

Skills Needed: Undergraduates from all disciplines are welcome to apply as we support a diverse pool of investigators. We do recommend that students have taken an introductory course in economics and/or psychology. Interest in decision science in general as well as willingness to jump into the variety of different projects that are likely to be running in the summer is important. The great opportunity of working for HDSL is that students here can learn skills they want to acquire. We provide structured training in the development of experimental methods and experience in conducting experiments.

Geography's Impact on Ideology and Agenda-setting | Ryan Enos (Government)

Project Mentors: Ryan Enos, Associate Professor of Government

Geography is of central importance to the study of politics. Space often structures important divisions in political, cultural, and social life, from political ideology to religious affiliation. The distribution of people across space has important consequences for political outcomes and behavior, and geographic variation helps researchers measure and understand complex political phenomena. BLISS fellows will work with Professor Enos and his team on a diverse set of projects exploring the relationship between space and politics, while developing practical skills related to data collection and experimental social science.

BLISS fellows will be responsible for collecting and organizing a diverse set of geographic data for a range of topic areas related to ideology and social identity, including the urban-rural divide, the dynamics of state and county-level policies, and the distribution of religious organizations and affiliations across space. Research assistants will help collect survey data and historical and contemporary administrative data and will also help in implementing experiments to test the effects of geographic space in a controlled environment. 

BLISS fellows will learn skills in data management, data collection, and analysis and will learn about an important field of academic research.

Skills Needed: None, though statistical training or GIS mapping skills are a plus.

Class-in-Group: Intra-racial Dynamics in Metropolitan Politics and Policy | Jennifer Hochschild (Government)

Project Mentors: Jennifer Hochschild, H.L. Jayne Professor of Government and Professor of African and African American Studies

How do the large "clunky" categories of black/white/Asian/Latino etc. get mobilized or modified in political and policy disputes in American metropolitan areas? In particular, do the interests and viewpoints associated with being well-off or poor affect intra-group dynamics, as well as relations across groups?  Do the policy arenas of policing, school reform, housing disputes, and city budget choices show the same racial and class dynamics across metro areas?

We have completed four site visits, involving a week-long trip to the site (New York, Chicago, Los Angeles, and Atlanta), to interview experts about, respectively, the criminal justice system, public pension reform, school reform, and urban development. Interviews have been transcribed, along with initial coding. The BLISS intern will continue our research on the four policy issues through published documents, census and other reports, and perhaps further conversations with experts.  

We also have a completed survey and I hope that the BLISS intern can do some basic statistical analyses of the results. There is likely to be additional coding of the interviews, and coding of newspaper articles and other media, around the themes of intra-group consensus or disagreement, and initer-group coalition or tension.

The BLISS intern will work closely with me and other research collaborators. My BLISS student from summer of 2016 continues to work on the Class-in-Group research. We will have frequent conversations, ad hoc meetings, exchanges of notes and information, etc. I will also work out a more specific research plan for analysing the survey, so that those proceed along a clear and fairly fixed path. The BLISS intern will gain experience in researching particular policy arenas and locations, setting up and participating in interviews, analyzing survey data, and moving from evidence to the first stages of writing an article or book chapter.

Skills Needed: Ideally, the student will have data analytic skills so can work with the survey data--so some hands-on statistical training would be very helpful. That is not essential however. Facility with JSTOR, Lexis-Nexis, and other electronic tools would be great.

How Education's Effect on Political Participation Varies Across Regime Type | Horacio Larreguy (Government)

Project Mentors: Horacio Larreguy, Associate Professor of Government

How the education's effect on political participation varies across regime type: Evidence from sub-Saharan Africa
In what constitutes a book project, we want to understand how education shapes political participation in sub-Saharan Africa, and particularly how it does so across regime types (i.e., autocracies, anocracies, developing and consolidated democracies, etc.). There are good reasons to believe that the direction of education's effect on political participation in developing democracies may vary depending on the extent of democratic institutions. When such institutions are strong, education could increase political participation because it is a venue for political change. However, under electoral authoritarian regimes, educated voters may instead deliberately disengage, since their participation might not only have limited effect in the political sphere but may also legitimize the leader by signaling support for the regime. Understanding if and when education increases political participation is crucial for comprehending the potentially critical role that it might play in democratic consolidation, which remains a contentious issue in the literature.

An important challenge in studying how the effect of education on political participation varies across regime types is that educated individuals often differ along other dimensions that could also directly affect political participation. To address this endogeneity issue, we exploit large-scale school construction programs to generate plausibly exogenous variation in education. Specifically, we compare the political behavior of individuals that were just young enough to benefit from the increase in access to schooling relative to those that were just old enough that they missed it. The region of focus will be sub-Saharan Africa where there are several countries that experienced large-scale school construction programs, and there is significant variation in regime type across countries.

So far, and largely aided by the superb research assistance of an undergraduate student within the BLISS program this past summer, we have concluded the analysis in five countries in sub-Saharan Africa that vary significantly in the strength of their democratic institutions: Zimbabwe, Tanzania, Nigeria, Kenya and Senegal. As expected, while across all countries educated citizens are also more interested in politics and supportive of democracy, we see stark differences in the effect of education on political participation in these countries. In Zimbabwe and Tanzania---with dominant party regimes---we see that education decreases political participation. In contrast, in Nigeria and Kenya---fairly competitive consolidating democracies---we find that education leads to large increases in basic civic and political engagement. Lastly, in Senegal, Africa's oldest and probably one of its strongest democracies, there is no effect of education on political participation. The goal is now to extend this analysis to two other countries---Botswana and Zambia---and to conduct a meta-analysis polling all countries.

The project entails the continuation of the construction of a rich database and statistical analysis using Stata. The database includes

  • Afrobarometer surveys on political-participation,
  • various sources on school construction, and
  • census microdata

The econometric analysis uses this data implementing difference-in-differences and regression-discontinuity analyses. As explained earlier, such analyses compare the political behavior of individuals that were just young enough to benefit from the increase in access to schooling relative to those that were just old enough that they missed it.

The undergraduate will learn creative ways to undertake these tasks since there is no standard way of doing this. Effectively, a day as part of the project entails doing significant coding in Stata, presenting results using LaTex, as well as a lot of discussion with me to figure out the logic behind such coding and result presentation.

Beyond helping with the coding and developing coding skills, as well as gaining experience in presenting results, students will become familiar with the related literature, how the project fits in such literature and the importance of the project analysis towards filling a specific gap in our knowledge. Moreover, I really try to structure things so that students can be part of each of the project steps. Sometimes we experience delays, and thus I am open to continuing the collaboration through the alternative ways that Harvard provides to engage undergraduates in research activities (e.g., Gov92r, IQSS URS) so that to ensure that goal. I also put an important emphasis on the rationale behind their work. Once students comprehend the type of analysis we aim to conduct, it is much simpler for them to understand the structure that the data should have, and thus the coding they need to do. I follow a very hands-on approach with weekly or bi-weekly meetings, and hold an open-door policy with all my students.

Skills Needed: I expect to work with undergraduates with a strong quantitative focus that have ideally taken an intermediate statistics course and have slightly-beyond-basic training in Stata or related software that entails coding (e.g., R, Python). A quantitative mindset and some basic training in a software structured around coding will avoid spending the first few weeks getting up to speed. Some knowledge of LaTex is also preferred. Regardless, the student working on the project will leave with a proficient knowledge of Stata and LaTex.

Managerial Talent and Economic Performance: MacArthur's Economic Purge | Melissa Dell (Economics)

Project Mentors: Melissa Dell, Assistant Professor of Economics

Managerial Talent and Economic Performance: Evidence from Discontinuities in Douglas MacArthur's Economic Purge
Productivity, innovative capacity, and the ability to transition into new product spaces vary significantly across firms. These differences are central to understanding economic development and structural change, and this proposal focuses on one potentially important determinant: managerial capital. The context is post-war Japan, a setting that offers both a large-scale natural experiment and extremely rich data. During the occupation of Japan, Allied Forces implemented an economic purge of senior managers in large Japanese firms. Managers who had been employed between 1937 and 1945 at firms with a net worth of over one hundred million yen were subjected to mandatory removal and barred from managerial positions in other large firms. They typically moved to firms in newer, up-and-coming industries, which were not dominated by large conglomerates. We examine the impacts of managers using plausibly exogenous variation from the purge and rich archival dossiers for all Japanese formal sector firms.

Quasi-experimental variation from the purge will be combined with archival data to examine the impacts of managerial capital. Dossiers for all Japanese formal sector firms, collected at regular intervals for the decade following World War II, contain detailed balance sheets, information about the payroll and workforce, and the technologies used in production. They also include comprehensive CVs for all board members and managers. The study will estimate the impacts of the purge on the firms whose managers were removed and on the firms to which these managers moved. The latter requires constructing a control group consisting of firms to which managers just below the cutoff would have plausibly transitioned had they been purged. This can be done by using the archival data to estimate a model of manager transitions that includes rich individual, firm, and network level characteristics.

The purge may have lowered economic performance by depriving Japan's large firms of scarce talent, or alternatively may have cleared the way for younger, innovative leaders to gain a foothold. It could have also promoted innovation by channeling talent into newer, rising industries, which tended to be exempt from the purge. However, if managerial capital is primarily embedded in organizational norms rather than in individuals, the purge may have had limited effects.

The BLISS fellow will focus on data analysis and modeling. Depending on the skills and interests of the student, analyses may include natural language processing of archival documents, data plotting and regression in R or Stata, geospatial modeling and mapping using Python and ArcGIS, and data parsing using Python.

At the end of each day, the BLISS fellow will be expected to send the PI a summary of the work completed that day. At the beginning of each day, the BLISS fellow will discuss the work for that day and any difficulties met during the previous day with someone from the project team. The PI or an experienced research assistant will be available at all times to answer questions. During the summer, my team of research assistants works together in a collaborative environment that includes weekly group meetings for all team members, in which insights are shared across projects. There are also project-specific meetings between research assistants, graduate students, and myself on a near-daily basis.

This research experience will provide extensive opportunities for the BLISS fellow to hone his or her problem solving skills. The student will gain experience developing simple prototypes to test various data analysis approaches. There will be an emphasis on understanding the underlying mechanics of how the programs that the student writes function. This understanding can then be used to efficiently and accurately debug and improve the programs. The student will also gain experience manipulating large datasets. These are all skills that will be useful for writing research papers or a senior thesis using empirical data, regardless of the student's field of study. These skills will be particularly useful for students interested in pursuing a career in data science, analytics, or academic research and are more generally applicable to approaching problems in a systematic way.

Skills Needed: Some experience with programming -- in a language such as Python, R, or C -- is required. This could be acquired by successful completion of CS50, or through past work experience. Completion of an introductory statistics or econometrics course is preferred but not required. Japanese language skills are not required.

Consistent Application of the Social Cost of Carbon | Jeffrey Miron (Economics)

Project Mentors: Jeffrey Miron, Senior Lecturer on Economics and Director of Undergraduate Studies

Do federal global warming and other environmental policies apply a consistent estimate of the Social Cost of Carbon?

The social cost of carbon (SCC) is an estimate of the monetized damages caused by a one-ton increase in carbon dioxide emissions in a given year. SCC estimates quantify the impact of a variety of damages caused by carbon dioxide emissions, including changes in net agricultural productivity, reductions in human health, property damages from increased flood risk, and changes in energy system costs. Policymakers rely on SCC estimates to implement a variety of environmental policies, such as green loans and subsidies, tax credits for wind and solar energy, and energy efficiency regulations.

Calculating the SCC is a complex process that requires identifying the relevant costs of carbon emissions, monetizing them, and determining how to balance the interests of the current generation against those of future generations. As a result, estimates of the SCC can vary widely, depending on the different models, assumptions, and discount rates used in calculations.

I am interested in whether estimates of the SCC are consistent across the policy landscape. For example, do some regulations implicitly rely on large values of the SCC, while others rely on smaller estimates? Are the costs of environmental policies balanced with a consistent estimate of their benefits? Does the policy framework as a whole coherently address the harms of carbon dioxide emissions?

To investigate these questions, I am interested in first examining current SCC estimates and reviewing their calculation. Then, I seek to identify implicit SCC estimates across policies to determine whether the environmental policy framework consistently balances the costs of regulation with the benefits of reducing carbon dioxide emissions.

The BLISS fellow will have the opportunity to become acquainted with an important and highly relevant area of policy research by participating in the search for and analysis of data and investigating the economic justification for environmental policies. Many of the skills developed during this research program are transferable to academic domains beyond economics.

I plan to meet at least weekly and typically more often with the undergraduate. These meetings would normally last 30-60 mins and involve reviewing completed work and discussing new assignments. My goal in the interactions will be to teach the student about economic research methodology more generally, in addition to discussing and learning about the specific project.

Skills Needed: Some knowledge of economics, statistics, and stata or R would be useful although not absolutely essential as long as the student is interested in learning these tools.


Best Practices for 3D Scanning of Ancient Artifacts | Peter Manuelian (Anthropology)

Project Mentors: Peter Der ManuelianPhilip J. King Professor of Egyptology and Director of the Harvard Semitic Museum, Joseph Greene, Deputy Director and Curator of the Harvard Semitic Museum; Adam Aja, Assistant Curator of Collections

How do 3D scanned models of ancient artifacts enhance teaching and research? What are the possibilities for broad dissemination, and what are the pitfalls of the 3D scanning process? Materials, resolution, person-hours... Come help us determine best practices while scanning a small but distinguished collection, the Egyptian artifacts in the Harvard Semitic Museum, for future research, teaching, and museum display.

The Harvard Semitic Museum is engaged in all aspects of enhanced collections management. We have recently acquired a $20,000 Artec Space Spider portable 3D scanner (resembling in shape and size a clothing iron). Tethered to a PC, the user aims the scanner at all sides of an ancient artifact to created a high resolution 3D model.

Our collection of Egyptian objects numbers 300 to 500 pieces, most of them small. This represents a unique opportunity to 3D scan an *entire* ancient collection in one of Harvard's museums. The student will receive basic training in scanning, study best practices, and assess the intellectual challenges involved: where does the technology excel, and where does it disappoint? How many extended uses can be devised; from rotatable online study, to "digital repatriation" back in original (virtual) contexts and find spots? We expect to make wide use of the resulting 3D model files, in the classroom, online, and in 3D printing for research and museum display.

There are tutorials for the Artec Space Spider scanner, and Museum staff will set up and train the student as well. S/he will learn best practices working with the collections in Harvard Semitic Museum storage, and will keep good tracking and archival notes of the technology and time involved per object, per material type, and per selection of resolution. Other researchers and cataloguers in HSM storage will be available for consultation. The Museum director, an Egyptologist, will inform the student about the significance of each type of artifact.

Skills Needed: Computer Science concentrators are by no means required, but those who are skilled with their hands, interested in technology, and excited by the study of material culture from the human past represent our ideal candidates. Students should have excellent sense of organization and project management skills, patience and careful handling of antiquities, and a methodical workflow.