Mapping social capital, economic vulnerability, and collective action in times of COVID 19- Part 1

By: Cristhian Parra, Claudia Montanía, Gustavo Setrini, Mónica Ríos, Marcos Martínez*

1 de Diciembre de 2021

In Paraguay, traditions of mutual aid play an important role when it comes to mitigating vulnerabilities. The COVID-19 pandemic has been no exception, our country has seen the emergence of countless mutual aid initiatives, from raffles and “ollas populares” (i.e., food drives, community soup kitchens, etc.) to the creation of digital spaces for articulating requests for aid, offers of assistance, and other citizen initiatives. 

How do vulnerable individuals coordinate collective action in this context, and what real impact do interpersonal networks have on reducing vulnerability? 

To answer these questions, the #AccLabPy launched a learning loop to study the role of mutual aid initiatives in Paraguay during the first year of the pandemic. We use the concept of social capital to analyze the role of interpersonal and trust networks in the articulation of collective action. In this blog we share what we learned when we started to map social capital in Paraguay.

Social capital: an ally during times of crisis

Social capital refers to the structures within a community that facilitate coordination and cooperation for mutual aid, such as interpersonal networks and social norms of trust and reciprocity (Putnam, 2000). Social capital’s effects on communities’ capacity to respond to and recover from disasters has been widely studied (Hurlbert, Haines & Beggs, 2000Shoji, Takafuji & Harada, 2020Smiley, Howell & Elliott, 2018). 

Crises often generate collective action problems. Responding to the need for food, medical assistance, and other needs, requires people to coordinate their activities and work together (Nakagawa & Shaw, 2004). During times like these, norms like reciprocity and trust, combined with the breadth and nature of interpersonal networks, can enable communities to respond better, faster, and more equitably.

In order to better understand the role of social capital, it is useful to differentiate among three types of social relations. 

Figure 1: Types of social capital according to the structure of interpersonal networks, adapted from Adams, 2020.

Bonding social capital describes relations among people who share a common identity and demographic characteristics. These include family members, but also members of one's own religious, racial and ethnic, class and other types of groups. These groups often share attitudes, information, and resources (Adler & Kwon 2002;  Mouw 2006).

Bridging social capital refers to relations among members of groups that do not share common identities (Paxton, 2002), with weaker social ties (Granoveter 1973). They are often forged through membership in organizations and associations. 

Linking social capital encompasses relations among common citizens and individuals in positions of institutional authority, often emerging out of formal and informal avenues (Szreter & Woolcock, 2004). This latter form is a vertical relation, while the previous two comprise horizontal relations.

Designing tools for analyzing social capital in Paraguay

A context-sensitive understanding of social capital dynamics and their relationship with vulnerability could be useful when managing crisis responses or even recovery and development initiatives. 

However, in Paraguay we lack data on social capital at the district, department or national level, and the concept itself is underexplored. 

Therefore, as a discovery strategy, we adopted an approach previously used in Paraguay (Rodriguez 2014Investigación para el Desarrollo 2015). We generated social capital indices using available data to infer levels of social capital. For example, by consulting the Permanent Household Survey (EPH), we can use education or age indicators to express bonding social capital in terms of homogeneity of a given indicator in a given district.

To bolster this work, we collaborated with Northeastern University Professor Daniel Aldrich's research team. We connected with them through the UNDP’s Global 
Network of Accelerator Labs
. In partnership with Prof. Aldrich’s team, we built a demo index for each of the three types of social capital, another representing vulnerability and finally, one representing community aid demand. 

In order to create the bonding social capital index, we used demographic and employment variables.  For the bridging social capital index, we used variables 
for the number of civil society organizations and initiatives, based on data mapped on the platform, Wendá**. For the linking social capital index, we chose variables related to levels of public investment and political-electoral behavior. 

Although the data sources presented different levels of geographical 
disaggregation, for the indices we created a standardized grid with 20 sq km units.  Each unit is given a rating of 0 to 1 for each index, indicating the abundance of social capital or the magnitude of demand for help in the area. The code for calculating 
index scores
 is free and open access. 

Figure 2. Indicators used for the creation of the vulnerability and aid demand indices in the proof of concept

In order to create the indices that measure vulnerability and aid demand, we followed the same procedure. For the social vulnerability index, we used demographic and household variables from the same data sources. In order to study the level of demand for aid, we analyzed the number of requests in AyudaPy***, a open source digital platform built and maintained by paraguayan developers during the pandemic, for posting and responding to requests for aid.

During the development of these indices, we invited a group of national and international experts to participate validate and discuss these tools. 

Figure 3. Indicators used for the creation of the vulnerability and aid demand indices in the proof-of-concept

Flashing red: mapping social capital density against levels of need in Paraguay

By comparing the values of these indices, we created a heat map that identifies 
hot spots, where the gap between the level of social capital and aid demand is 
higher, and cold spots, where the gap is lower. 

Figure 4: Hot and cold spot maps indicating the gaps between aid demand and available social capital within a given territory. 

Source: Data from Household surveys, AyudaPY*** and Wendá**


Figure 5: Other maps we explored to visualize (a) gaps between aid demand and social capital at the national level and in the capital, Asunción, (b) aid demand throughout the country, (c) charitable NGOs, (d) volunteer organizations, y (e) other citizen initiatives.

Source: Data from Household Surveys, AyudaPY*** y Wendá**

Estimating these gaps served as proof-of-concept for a method to identify regions that may require support in responding to aid demand. However, this approach has its limitations.

In sparsely populated areas, it's not possible to adequately measure the gaps. Moreover, the levels of geographic disaggregation of many of the secondary data sources do not permit an analysis that is localized enough to guide public intervention. Finally, it’s not possible to precisely distinguish the multiple causal
relationships among the different indicators and indices.

With better and more precise data and with algorithms that improve the estimates based on more direct measurements of social capital, the map could be a useful tool for guiding short-, medium-, or long-term responses to crisis. 

For this reason, the following phase of this learning loop focused on designing and carrying out a survey of social capital, economic vulnerability, and collective action to produce data representative at national and regional level. 

This survey was reviewed and approved by the same panel of experts mentioned 
earlier and sought to overcome the limitations we encountered during the discovery phase of this project. The survey provides a tool for exploration that will generate a more precise and valid database for the study of social capital in Paraguay. We are still working on the analysis of this survey, and its details and learnings are due for the next blog of this series. 

Lea este blog en español aquí.

References:

Marcos Martinez Sugastti studied economics at Columbia University (New York). Currently, he is pursuing a doctorate in agricultural economics at the University of California Davis. He has conducted public policy research and analysis within academia and for the Paraguayan government. 

** Wenda is a digital platform that we launched in collaboration with the Estrategia
 Nacional de Innovación (National Innovation Strategy) and various civil society organizations that maps and connects of citizen-led initiatives launched during the pandemic. 

*** AyudaPY is an open source digital platform developed by Marcelo Elizeche and other contributors. Its data and content are distributed under creative commons license CC BY-NC-SA 4.0. Source code is free under AGPLv3