“With no lips, the teeth get cold”: What can SNA tell us about contemporary PRC-DPRK relations?

Background

North Korea (officially, the Democratic People’s Republic of Korea, or DPRK) sits at the strategic intersection between the People’s Republic of China (PRC, or China), the U.S., and its East Asian allies, the Republic of Korea (ROK) and Japan. (Russia is also a player, but only plays a bit role in this analysis.) In 2017 and 2018, events on the Korean peninsula reached a fever pitch of both tension and détente. The U.S.-led maximum pressure campaign put the U.S. and its allies on a near-war footing, and then, in an unlikely turn, an incredible thaw began. A sitting U.S. president met for the first time with a North Korean supreme leader, and at least some halting efforts towards the removal of nuclear capability from the northern half of the peninsula appeared to have begun. Major annual military exercises by the U.S. and ROK were shelved. Meanwhile, the ROK president had several meetings with the North Korean leader.

Many Western public figures believe that China has the power to “fix” the North Korea issue whenever it wishes, including the U.S. president. Based on a superficial examination, relations between the PRC and the DPRK appear to have improved in 2018. For the first time since taking power, the North Korean leader traveled to another country for a visit (it happened to be China), and high-level Chinese envoys were also received in Pyongyang. But on a deeper level, what is the state of relations? How could we tell using open source materials?

Research Questions

As one part of a larger research project, this effort proposes to examine person-to-person ties between the PRC and DPRK as one means of characterizing the nature of state-to-state relations. Using social network analysis (SNA) tools is meant to reveal the more profound significance of particular ties.

• Which ties are the most valuable/important/influential? Are there certain positions which seem to be the most influential, regardless of who occupies them?
• What degree of resiliency exists in the relationship? (i.e., is there but a single main link, or are there multiple links of near-equal strength that could be used as a fall-back should something untoward happen to the primary interlocutors?) What would be the implications of the potential severing of critical links?
• Do the connections remain now? If not, has a different channel replaced them?

Hypothesis

My primary hypothesis is that overall, relations between the DPRK and PRC have worsened since 2011. In the larger picture, I believe such a worsening of relations between the countries contributed to the DPRK’s decision to focus on the rapid development of its nuclear weapons capability during this time period. I also believe that the research will show that individuals are more important than roles in the PRC-DPRK relationship (i.e., position matters less than does who is the specific person in the job).

Previous Works and Data Collection

At least two analysts have examined power structures and person-to-person relations in the Chinese Politburo and high leadership. Keller’s work, described in Niall Ferguson’s 2018 book, The Square and the Tower, examines network position as a factor in coalitions which allow certain people to ascend to top leadership positions. Gregory’s 2013 study tests several candidate theories which aim to explain loci of influence on the Chinese Politburo and finds that a combination of institutionalism and personal ideology displays the best fit. Each researcher used SNA as a key tool in their analysis and their work points the way toward additional resources which can help inform this study.

Several databases exist which already contain information about high-level exchanges between the two countries dating back at least to 2003. China Vitae of the Carnegie Endowment for International Peace “tracks the appearances and travel of up to 200 leading Chinese officials” and provides date, location, and attendees participating in meetings, as well as the topics discussed. Pacific Forum’s High-Level Visits database provides similar information with slightly less detail, dating back to 1974. I will use these data sources as a starting point for creating a timeline of interactions between the two countries.

Supplementing these sources, I intend to search news media sources (newspapers, etc.) to gain additional data for the analysis. Western papers like the New York Times and Washington Post might contain useful information about these meetings. I will also examine Chinese and North Korean official state media published in English, though these sources shall be considered with a skeptical eye toward reportage of the content of such meetings, should any be included.

Network Measures

After collecting the data, I will create a two-mode dataset for all known PRC-DPRK meetings since Kim Jong-un came to power in late 2011. This will show who attended a meeting, where it was at, what was discussed, and so forth.

Network centrality measures and subgroup analysis should reveal key relationships and possibly notable factions.

It is my hope that analysis of these measures will reveal valuable results helping to specify which persons or positions on both sides have been the most important during the current regime.

Opportunities for Further Analysis

Future researchers could create network diagrams/data sets for top DPRK leadership such as the ones Keller has generated on China.

Using SNA to identify network structures that can facilitate population-level behavior change in Pakistan

Background


Long-term development impact relies heavily on target populations adopting the intervention and continuing with it long after the intervention has been completed. Previous studies have shown that knowledge and behavior can spread through social networks and that certain types of social networks can enable population-level behavior change which can help prolong the impact of the intervention and have long-term benefits [1]. There are several sectors within the development arena that can benefit from the knowledge of how social networks function and precisely which individuals to target to ensure a better uptake of the intervention and its subsequent effectiveness. Specifically, in the context of Pakistan, where interventions focused on polio, latrine use, etc. are also shaped heavily by other cultural and religious contexts [2], information about how social networks can improve flow of information and encourage certain behaviors have the potential to be very useful for targeted interventions that can yield maximum benefit.


Hypothesis


I expect that there are certain individuals within a social network that yield influence within their community that encourages better uptake of some development interventions. In some cases it might be individuals who are good broadcasters as opposed to conventionally influential individuals such as village elders or tribal leaders. For instance, a village elder may be considered influential but the son of a village elder who knows people of the same age, has similar concerns related to health, nutrition etc may be able to yield more influence among peers and be a better broadcaster. Or it could be a government official trusted by the tribal leader who also brings with him/her the influence that comes with being associated with the government (or not). These angles of analysis can provide important information that can be used by aid agencies to carry out targeted interventions for sustainable impact in Pakistan.


Methodology


This SNA will explore if interventions targeted towards certain individuals within a community lead to behavior change among the population? Does uptake of the intervention by certain social contacts influence individuals to follow?


I propose that this sort of analysis is done as part of a baseline survey of the area of intervention. This SNA will rely on primary data collected through a survey that seeks to answer the following questions:


1. Which individuals do you go to for health/sanitation related advice? Name up to 5 people.

2. Why do you trust these people for advice? [Similar circumstances, they are more experienced, their advice is supported by their own actions (such as using the same sanitation techniques or getting their children vaccinated)]

3. What other mediums do you rely on to get information on health/sanitation related matters?



Apart from these questions, the survey will also collect information on age, gender, religion, occupation, education, number of children, levels of friendship (whether people visit the 5 named individuals or invite them to their homes, etc.). To see who the most influential individuals are, centrality measures including eigenvectors can be used as well as in-degree measures to see who people turn to the most for advice and who is close enough to the influential nodes to be a good broadcaster. The information collected in the survey can also help  check for homophily (whether people turn to those for advice who have similar traits) and betweenness.


Limitations


This SNA can benefit from other forms of qualitative analysis with more data collected over time, such as before, during and after the intervention. Additional qualitative analysis can include running regressions to check for the impact of certain characteristics while holding other variables constant and that will provide a more thorough analysis of how much influential individuals matter, given other factors.


I am not taking the second module. 

[1] Fowler, James H., and Nicholas A. Christakis. "Cooperative behavior cascades in human social networks." Proceedings of the National Academy of Sciences (2010): 200913149.

[2] Murakami, Hitoshi, Makoto Kobayashi, Masahiko Hachiya, Zahir S. Khan, Syed Q. Hassan, and Shinsaku Sakurada. "Refusal of oral polio vaccine in northwestern Pakistan: a qualitative and quantitative study." Vaccine 32, no. 12 (2014): 1382-1387.

Global Trade, Tariffs, and War

Background: The Trump Administration is attacking the idea of free trade: abandonment of TPP, the “renegotiation” of NAFTA, and a trade war with China. All of this has caused angst among allies and ire from adversaries leading the world into a state of chaos reminiscent of the 1930s. The economic policies of the United States during the interwar period such as the Smoot-Hawley tariffs and not forgiving war debts contributed to onset of the second world war. In addition to protectionist economic policies, rising nationalism and ethnic tension make the current period unfortunately like the interwar period.  

"When goods don’t cross borders, troops will” 

This project will examine the trade networks of the interwar period (1918-1938) and the current period (1998-2018). Mapping the similarities in trade flows may give us further insights into the causes of war and to see if the policies of the Trump administration should be a greater cause for worry. 

Questions: Can the flow of international trade or the lack of trade amongst nations be a predictor of the likely hood of war (WWII as the model)? How closely does the current period match that of the interwar period? What could this mean? 

Hypothesis: I believe that there is a strong link between the flow of trade and the relations among nations. I think that the analysis of the interwar period will show that trade flows decreased (as tariffs were imposed) coupled with a rise in nationalism (and racism) made WWII all but inevitable. Unfortunately, I also believe that we will see the same patterns in the current period with respect trade and nationalism. 

Data Collection: Data is widely available from the following sources: World Trade Organization (https://www.wto.org/english/res_e/statis_e/statis_e.htm), World Bank (https://wits.worldbank.org/CountryProfile/en/Country/WLD/Year/LTST/TradeFlow/EXPIMP/Partner/by-country), and the United Nations (immigration data - http://www.un.org/en/development/desa/population/migration/data/empirical2/migrationflows.shtml#). Attribute data is available from various government archives and academic publications.  MFN status among countries will also be an attribute. 

Methodology: Import and export levels between countries could use to measure the “tie strength” of trade between nations – exact levels need deeper research. Also given the concept of economic specialization, I am undecided if directionality matters -- I believe that directionality will not matter considering the natural imbalances between developed and developing nations. 

Attribute data will be overlaid on the network: political ideology, population demographic data, immigration data, public policy, defense spending/mobilization. This is perhaps the most important feature of the network analysis. What do the attributes tell us about the network? Does a "nationalistic" government lead to economic isolationism and increased trade barriers? Is there correlation between trade flows and immigration flows? 

Additionally, I would like to look at they basis for an economy such as commodity or service. Which type of economy is more likely to be the belligerent in a war?  

Faction analysis will be conducted and may lead to the identification of alliances. 

Centrality measures and egonets can be used to identify the role of nations in international politics. Mediators pre-conflict perhaps? 

Trade openness of an economy compared with GDP growth overlaid on a network will provide additional insights and may lead to further research.  

Limitations: As with all network analysis, conclusions cannot be drawn – only inferred. Additionally, as the international economy has developed greatly in the post war era, therefore we may not gain data or insights that are directly comparable.  Another issue that may arise is the availably of detailed data from the interwar period as most data available at the above sources covers the modern period (post-1990).  
Opportunity for Further Research: I think this type of network analysis would be a useful to apply to a multitude of historical conflicts to mine for trends. I have a hypothesis that economic factors are the best predictor of conflict compared to all other factors: religion, ideology, tribalism, ect. And I think trade network analysis can focus research to confirm my hypothesis. In addition to preventing war, this analysis may provide insights into economic growth nd development -- perhaps even more useful data for policy makers (additionally the data would be easier to come by as we would only need to go back a few decades).

Unfortunately, I will not be able to take the second module.

The Fletcher Mafia: Network or Urban Legend?

I am not taking the second module this Fall semester but am looking at taking it in Fall 2019.

Background

This project stemmed from my own experiences during the application process for The Fletcher School. While I had constantly been hearing about how well connected the alumni network was but I could not manage to get hold of any information about Indian alumni, the sectors they’ve worked in across the years, the locations they have moved to, a mapping of their various career trajectories etc.

Purpose of the project and Research Questions

I will limit this project to the Indian alumni of the Master of International Business (MIB) program since inception i.e. include a decade’s worth of metadata for inter-temporal analysis. The purpose of this project is to understand the career trajectories of this alumni network, the locations and industries they end up working in, how they are connected to other members of the network and how it has changed over the years.

This information can be used to better direct efforts of the Office of Career Services and Alumni Relations to provide targeted information to alumni, prospective and current students, connect members of the community through programs which serve common purposes and interests.

The further scope of this project will be to expand it to other degree programs and also look at geographical areas from where we have less / dwindling student representation and engage alumni in those areas for information sharing and recruitment programs and initiatives.

Research Questions:

1.       What are the various subgroups that alumni networks are divided into?

Intent of this question is to understand if alumni are connected closely based on their location, industry, sector, employer, graduation year or designation (especially if it is at a senior position such as partner / director etc.)

2.       Have the cluster composition identified above changed over the years?

Intent of this question to understand the trends (if any) in terms of Fletcher alumni flow towards certain sectors / industries / employers over the years and if it can be leveraged to inform current students about opportunities in those areas.

Specifically it will be interesting to see if there are certain regions where several Indian alumni tend to relocate for work which would help international students further in their career search.

3.       Are there any specific employers who are connected to a significantly large number of Indian Fletcher alumni and / or help them transition between employers / sectors and / or industries?

Data Collection and SNA Methodology

While most of the metadata required for this analysis should already be available with the Office of Career Services and Alumni Relations, there may be a requirement to send a simple survey to all Indian alumni to get further information.

To begin with the attribute data required for this project will be:

1.       Location currently based out of (name of city)

2.       Industry (financial services, media etc.)

3.       Sector (private, public, non-profit)

4.       Designation (director, partner, associate etc.)

5.       Graduation Year

6.       Employer name (last 3, this data will not be valued i.e. there is no value given to recent employers vs. past employers. The main focus is to understand what kind of employers are a part of the network)

This analysis will primarily be a two mode network (alumni – industry, alumni – sector, alumni – employer). Using a two mode network to understand how alumni are connected through different organizations / locations at various points in time will provide better insight about which kinds of organizations to tap into to build alliances through research projects / student activities / seminars etc. via alumni engaged in those organizations and sectors. In this structure the nodes will be organizations and Fletcher alumni and the connections will be edges.

Network Measures:

1.       Network Description: it will be interesting to see in terms of location if the network over the years has been centralised (i.e. majority alumni have ended up staying in the United States for instance and within the US if its specifically in Boston / New York / California which may give an indication of what kind of industries and companies are open to hiring and retaining international students). Similarly if the alumni network has been centralised in a given sector (for instance the private sector primarily), but within the sub network is there small world-ness across allied industries (such as technology, impact investment, financial services).

2.       Network Type: It will be a two mode, binary, scale free network with hubs of connectivity which shape the way in which the whole network operates.

3.       Centrality Measures of nodes (organizations) for every given year will help understand if there are any specific organizations which hire and retain a high number of Indian Fletcher alumni (high degree, betweenness), perhaps investigate further on what the reason could be (for instance they could be large corporations which have visa sponsorship budgets or they’re developing business in emerging markets etc.)

4.       Factions and subgroups: Assess if alumni network is divided into factions which could be based on industry affiliation, graduation year etc. Within those subgroups assess if there are certain organizations or alumni which serve as brokers and help connect other members of the network (high betweenness and eigenvectors). Also see if there any visible trend across the years for these brokers / high connectors (for instance belong to the same industry in majority of the years under consideration)

5.       Two mode to one mode: For every year convert the two mode network to a one mode network to see who are the most well connected alumni in every batch (look at centrality measures) and use them to broadcast information across the network.

Future Scope of Study

1.       Include alumni from other degree programs

2.       Focus on specific geographic areas

3.    Focus on certain sectors / industries where we don’t have enough stronghold and compare them with sectors / industries that we do to see what lessons can be cross applied in terms of alumni reach out?

Community Structure in Rural Kenya and its Influence on Mobile Money Use

Background

Across East Africa, individuals across every demographic are rapidly adopting “Mobile Money” payment services; paying bills, transferring funds, as well as withdrawing and depositing cash. Since the early 2000s, Uganda, Tanzania, Zambia and Kenya in particular have been experiencing wide adoption of these tools. In Kenya, Vodafone’s M-Pesa cultivates a network of independent agents who offer banking services to their local communities. These agents are then able to visit independent businesses contracted by Vodafone where they can exchange e-money for cash and balance their books. Fees are extracted based upon the size of each transaction, allowing agents, contractors, and Vodafone to profit from the exchange. Similar models are being adopted by companies across the region. Currently, peer-to-peer cash flows in Kenya account for a significant portion of the Kenyan economy: roughly 25.6 billion US dollars. Roughly 70% of these transactions are occurring via mobile phone.^[1]

In rural areas, mobile money is a medium for financial inclusion. Lack of infrastructure, regulation, and institutional support, which in this analysis are considered transaction costs^[2], inhibit and disincentivize participation in the formal economy. Mobile money allows these isolated communities to access cash more readily and save their earnings. In 2014, 76% of Kenyans above the age of 15 living in rural areas used mobile phones to make digital payments.^[3] However, only 8% of money transfers between individuals and firms occur over mobile phone. 

What are the sources of these trends? Why is mobile money and the financial services it provides being adopted so quickly and across such a wide section of the population? Why is mobile money so frequently used in Kenya for peer-to-peer transactions, but not for payments to firms? Why is mobile money outpacing use of traditional financial institutions?

For the last few weeks, Hitachi Center Fellow Dr. Jonathan Greenacre and I have been discussing these questions at length. From our meetings and preliminary research, we have found that while a number of multinational corporations, international organizations, and academics have conducted comprehensive population surveys at the national level, none have accounted for social ties within individual communities. We believe that a formal social network analysis of a specified Kenyan community group and their money use would provide insight into the questions listed above. This social network analysis will supplement Dr. Greenacre’s broader study of bottom-up mobile money regulation within Kenya.

Hypothesis

Following my conversations with Dr. Greenacre and Associate Director of Research and Doctoral Research Fellow for Innovation and Change, Ravi Shankar Chaturvedi, I hypothesize that mobile money popularity is in part due to the way it compliments traditional community structure. In addition to institutional mistrust and the transaction costs associated with the use of formal financial institutions, the financial services offered by companies like M-Pesa allow individuals to easily transfer money between friends and relatives; people they trust. Furthermore, agents providing banking services can be held accountable because they themselves are members of the community.

These agents will likely be central nodes within the network. High-density cliques and clusters may form within family groups, which may, in turn, be subsets of larger factions determined by social group. Agents will serve as primary brokers between these clusters and factions. This analysis will illustrate communal social ties, how they influence the use of mobile money services, and how those services are used.

Methodology

There are several relevant datasets available through the World Bank Global Financial Inclusion Database as well as the Harvard Dataverse. While the former offers primarily country-wide metadata, it is possible that Dr. Greenacre’s substantial connections within the World Bank may be able to yield individual household data. Alternatively, FSD Kenya’s survey of 1600 Kenyan households conducted by Dr. Tavneet Suri of MIT Sloan over a 5 year period looks more promising.^[4] Finally, as a contingency, Dr. Guy Stuart, Executive Director of Microfinance Opportunities (MFO) and Harvard Innovation Program Fellow, has shared data from his work with FSD Zambia with me.^[5] This includes individual household spending diaries from underserved Zambian families over several years. Whether Kenya or Zambia, the structure of this analysis will not be affected. MTN Mobile Money in Zambia uses the same model as M-Pesa in Kenya; the two formed a partnership to provide service to 19 African countries in 2015.^[6]

The simplest way to represent tie-strength between community members is through the absence or presence of a mobile transfer between individuals in the past year. This will yield a binary directed dataset, however, it may prove too broad. Depending on the available data, I may use frequency of transaction as an indicator of tie strength, differentiating transaction frequency by week, month, and year. Alternatively, significance of transaction may be used as a measure of tie strength: goods purchase vs. lending emergency funds. If neither is a valid measure of tie strength, I may employ alternative tie strength measurement models.^[7] [8]

This network may be complemented or combined with 2-mode attribute datasets to further depict community spending patterns, namely transactions with businesses and financial institutions. These datasets will be calculated for two separate years, which will vary depending on the chosen dataset (e.g. Dr. Suri’s surveys: 2011 & 2014). This will show changes in mobile money use over time.

Limitations

The most pressing limitation on this analysis is the large size of the datasets I am using. First, a key question I will need to answer is how I will narrow down my analysis to a select number of households within a given community, ensuring that the chosen subset is representative of the larger whole. Second, I will need to widdle down the existing attribute data, choosing only the most relevant attributes to my analysis. 

I am taking the second module of the course.

---

^[1] https://sites.tufts.edu/ibgc/files/2016/04/IFMM-Case-Study.pdf Supplemented by private conversations with Mr. Chaturvedi

^[2] Posited by Dr. Jonathan Greenacre, Hitachi Center Faculty Fellow

^[3] http://databank.worldbank.org/data/reports.aspx?source=1228#

^[4] https://dataverse.harvard.edu/dataverse/mobilemoney Dr. Greenacre has contacted her for additional information on the study

^[5] http://www.fsdzambia.org/fsdz-data/

^[6] https://www.nation.co.ke/business/eMoney/Safaricom-MTN-partnership-Africa/1625712-2694198-yc82wo/index.html

^[7] http://eliassi.org/papers/gupte-websci12.pdf

^[8] https://www.jstor.org/stable/2579058?seq=17#metadata_info_tab_contents

Integrating Social Network Analysis and Geospatial Analysis to Predict Latrine Ownership in India

1. Background

Sanitation is one of the key issues in developing countries. In 2015, 2.3 billion people lacked even basic sanitation services, and 892 million people worldwide still practiced open defecation (WHO and UNICEF, 2017). Lack of access to adequate sanitation leads to contraction of diseases such as diarrhea through ingestion of pathogens from untreated feces. Those diseases are estimated to cause 280,000 deaths every year (Prüss‐Ustün et al., 2014).

There are several factors which affect the household’s ownership of latrine (toilet). Previous studies mainly focused on the effects of socio-economic conditions, financial conditions, and dissemination of information on the health benefits of a latrine. However, less literature focuses on peer effects among households, which means how latrine ownership of a certain household will affect the latrine ownership among neighboring households. Thus, in this research project, I focus on clarifying how social interactions among households affect the decision of latrine ownership.

Economists have mainly conducted researches on this topic and suggested that peer effects can be both positive and negative depending on the target areas and experimental designs. (Pattanayak, 2009; Guiteras et al., 2015; Yishay et al., 2016). On the other hand, Shakya et al. (2015) is the first paper which used social network analysis in examining the peer effects of latrine ownership. They showed that individuals are more likely to own latrines if their social contacts own latrines. This relationship among households is stronger among those of the same cast, same education, and those with stronger ties. In addition, individuals who are more central in the network are more likely to own latrines.

However, the analysis in Shakya et al. (2015) can be further deepened in the following ways. First, they only focused on the social ties between households and didn’t consider the distance among households. By incorporating geospatial aspect in social network analysis, I attempt to predict the interaction among households more precisely. Second, they haven’t considered how leadership affects the social contagion of latrine ownership. I will look at whether leaders in the village are more effective than other people at influencing people to own latrine.

2. Research Questions and Hypothesis

This research intends to answer how latrine ownership of certain households is affected by the status of latrine ownership of social contacts/neighbors. Specifically, I will focus on the following research questions.

1) Does the analysis of using social ties and distance among households reach the same conclusions as the analysis of using only social ties?

2) Does this combined method enable us to identify the significant social effects which were found to be insignificant in Shakya et al. (2015)? (eg. social effects in the network of “Go to temple with”)

3) Are leaders more competent in influencing others’ decisions in owning latrine?

As for the first and second questions, I expect that the distance is another factor in predicting how latrine ownership spreads across households. By combining distance with social ties, I can better predict the social effects on latrine ownership.

For the third question, I hypothesize that leaders have generally more power in influencing other decisions. But if leaders are further away from households, then the power of influence to those households decreases.

3. Data and Methods

I plan to use the social network data which were collected in the Banerjee et al. (2013). This dataset is mostly publicly available in Harvard Dataverse but GPS data of each household is not public. I will apply for IRB review and then receive additional GPS data from the authors of Banerjee et al. (2013).

Data includes network data on 16,579 individuals coming from 6,811 households in 75 villages. There are 12 types of network data, covering all individuals. The dataset also includes attribute data of all individuals. For example, you can know whether each individual owns latrine or not, whether each has electricity or not, and whether he or she is a leader or not. The dataset also contains information on caste, religion, language, and roof-type, etc. of each individual.

In the analysis, I first construct a dataset of each distance between households using GPS data. I integrate the value of network ties with distance by certain formulas to create new indices. Then, I will run a regression to see the social effects on latrine ownership. The regression sets dummy variable of whether a certain individual (ego) owns latrine or not as dependent variable. For independent variables, I include interaction terms of similar dummy variable of whether social contact (alter) own latrine or not, multiplied by new indices. I also include caste, electricity, roof-type, religion, language as control variables. After running a regression, I will check whether the coefficient of the independent variable is statistically and economically significant. Additionally, I will run this type of regression repeatedly using each type of network data to capture the heterogeneity in the results.

For the third research question, I will run this regression with samples including leaders, and run same regression with samples including non-leaders. I will then compare the results of the two regressions.

4. Limitation

Since the data used in this analysis are cross-sectional, we cannot infer causality of social effects of toilet ownership from this result, even if we can conclude the association among households from the regression. We will need time series data or new experimental design to rigorously infer causality.

5. References

[1] Guiteras, R., Levinsohn, J., and Mobarak, A. M. (2015). Encouraging sanitation investment in the developing world: a cluster-randomized trial. Science, 348(6237), 903-906.

[2] Pattanayak, S. K., and Pfaff, A. (2009). Behavior, environment, and health in developing countries: evaluation and valuation. Annu. Rev. Resour. Econ., 1(1), 183-217.

[3] Prüss-Ustün, A., Bartram, J., Clasen, T., Colford, J. M., Cumming, O.,  Curtis, V., Bonjour, S.,Dangour, AD., De France, J., Fewtrell, L., Freeman, M. C., Gordon, B., Hunter, P. R.,

[4] Johnston, R. B., Mathers, C., Mäusezahl, D., Medlicott, K., Neira, M., Stocks, M., Wolf, J., and Cairncross, S. (2014). Burden of disease from inadequate water, sanitation and hygiene in low‐and middle‐income settings: a retrospective analysis of data from 145 countries. Tropical Medicine & International Health, 19(8), 894-905.

[5] Shakya, Holly B., Nicholas A. Christakis, and James H. Fowler. "Social network predictors of latrine ownership." Social science & medicine 125 (2015): 129-138.

[6] WHO and UNICEF (2017). Progress on drinking water, sanitation and hygiene: 2017 update and SDG baselines.

[7] Yishay, A. B., Fraker, A., Guiteras, R., Palloni, G., Shah, N. B., Shirrell, S., and Wang, P. (2016). Microcredit and willingness to pay for environmental quality: Evidence from a randomized-controlled trial of finance for sanitation in rural Cambodia. Journal of Environmental Economics and Management.

I am taking the second module of the course.