Attitudes toward social assistance in the developing world

Voters’ views towards social assistance

Our project will seek to disentangle different potential drivers of voters’ views on social protection in Low- and Middle-Income Countries (LMIC).

We will survey individuals through market research companies in several LMICs about their views on different social protection policies that have been found to be effective in the recent economics literature, including conditional or unconditional cash, in-kind transfers, one-off asset transfers, and public employment programmes.

As a second step, we will conduct a randomised survey experiment to investigate whether views on the different programmes are affected by information treatments that (i) inform respondents of real-world evidence on their effectiveness and (ii) describe the economic mechanisms behind each programme.

We will also incorporate an AI-powered chatbot into the survey, drawing on recent findings in the survey literature (see Chopra and Haaland, 2023). This will allow respondents to have a natural conversation regarding each of the programmes.

This research will be the first to study social assistance programmes in low- and middle-income countries using large-scale survey experiments (see Stantcheva, 2023). It will contribute to a better understanding of how people reason about different social assistance programmes, helping policy makers design programmes that are in line with voters’ preferences.

Context and motivation

Progress towards eliminating global poverty has slowed considerably in recent years – the share of people living on less than $2.15 per day was the same in 2023 as in 2019 (World Bank, 2024).

As major donor countries cut back on international aid spending, funding for social assistance programs will increasingly need to be raised domestically. Policies must not only be effective in reducing poverty, but must also garner sufficient public support to be politically feasible. Understanding which factors influence voter support for – and opposition to – these different programmes will therefore be of importance to policymakers.

To shed light on these factors, we will implement a large-scale cross-country survey in multiple low- and middle-income countries. We intend to make use of both traditional closed-ended survey questions as well as text analysis of open-ended survey questions, which among other techniques could be analysed using Large Language Models. This will help us understand how participants reason and what their first order concerns are (Ferrario and Stantcheva, 2022).

We will also explore how views on and support for different social assistance programmes are correlated with participants’ level of trust in the government and their general views on fairness and redistribution.

Next steps

We aim to pilot our survey in Kenya, Senegal and Peru in late 2025. We anticipate that we will conduct the full scale survey in multiple countries in the first half of 2026.