Culture and economic development

Editor's note: This episode of VoxDevTalks is also available on Spotify, Apple Podcasts, and YouTube.

In this episode of VoxDevTalks, Tim Phillips is joined by Natalie Bau, Sarah Lowes, and Eduardo Montero to discuss their chapter for the Handbook of Cultural Economics. Together, they explore how culture shapes development policy – and how policy in turn reshapes culture. The conversation digs into evidence, examples, and the emerging frontier of research that blends economics with anthropology, sociology, and history.

The discussion is driven by a core idea: if culture is ignored, even the best-designed development interventions can fail, backfire, or produce unintended consequences.

What economists mean by culture

One of the first challenges involves defining culture.

“By culture, we generally mean the values and beliefs that are socially transmitted across generations.” Lowes

Examples include kinship systems, marriage customs, and trust in others. Economists, Bau notes, also view culture as shaping “the rules of the game”, whether through social sanctions, heuristics for decision-making, or inherited beliefs and attitudes.

This broad view allows economists to recognise culture’s impact across many aspects of life and development outcomes.

Measuring culture with rigour

Historically, economists considered culture ‘soft’ or ‘fuzzy’, something impossible to measure.

“Economists like rigorous things that we can measure. In the last few decades, economists have gotten much better about measuring culture and thinking about how to get micro data on it.” Montero

This new wave of research uses multiple methods:

• Surveys asking questions such as “Do you trust others?”
• Lab and field experiments testing behaviour in controlled scenarios
• Ethnographic data collected by anthropologists

Economists now borrow from other disciplines, blending qualitative and quantitative data to integrate cultural variables into econometric models.

Why culture matters for policy outcomes

The authors stress that policies often succeed or fail depending on whether they account for cultural knowledge. To show this, they analysed World Bank projects and their evaluations.

“If individuals happen to be from the place where the project is being implemented, the projects get higher outcome ratings.” Lowes

In other words, when project leaders share cultural and local knowledge with the communities they serve, outcomes improve. This suggests a strong link between familiarity with cultural context and project effectiveness.

However, the picture is more complicated. Organisations deliberately avoid always assigning local leaders to projects in their home countries. This helps reduce risks of nepotism, corruption, or capture by local elites.

There are also benefits to rotating leaders across countries, such as learning from diverse contexts and avoiding over-reliance on personal networks. However, one drawback of frequent rotation is that officials are often reassigned just as they start to understand the local culture.

So, cultural familiarity can make policies more effective, but institutions must balance that against governance concerns and the value of cross-country experience.

When well-meaning policies backfire

The conversation then moves to striking examples of unintended consequences:

• Pension plans in Indonesia and Ghana: Bau’s research found that pensions reduced children’s incentives to care for elderly parents, weakening traditional kinship arrangements.
• Lock boxes in Bolivia: An RCT aimed at encouraging savings unexpectedly led to increased alcohol consumption.
• Colonial health campaigns in Africa: Montero and Lowes show how forced medical campaigns created lasting mistrust in healthcare.

“We find places that had more of the history of these colonial campaigns are more likely to refuse free blood tests and have fewer vaccinations today.” Montero

The challenge of scaling successful interventions

Scaling is a perennial challenge. Interventions that succeed in one setting may fail in another. For example:

• Lock boxes worked well in other contexts but failed in Bolivia.
• Unconditional cash transfers, often praised for reducing poverty, faced rejection in Kenya. 

These examples highlight the danger of assuming universal applicability without considering local norms.

Towards culturally conscious design

The authors emphasise the importance of embedding culture into policy design from the start.

“What you don’t want to do is you run the project and it fails, and then you kind of ex post say, ‘Oh, was that because of culture?’ Then it’s very difficult at that stage to really investigate that.” Bau

Their recommendations include:

• Collaborating with anthropologists and sociologists early in the process
• Using focus groups and qualitative data to inform design
• Collecting data on cultural practices systematically
• Embedding cultural variables in large surveys such as the Demographic and Health Surveys

As Bau stresses, the aim is not to track every cultural nuance but to identify those that change incentives in predictable ways.

Better data and deeper collaboration

The episode closes by looking ahead. As Lowes summarises: “There’s a lot we don’t know.” The field needs richer, time-varying data on cultural practices, not just static snapshots from colonial-era ethnographies.

Equally important is interdisciplinary collaboration. Conversations with anthropologists, for example, can spark valuable partnerships that improve both the design and interpretation of research.

Culture as the critical missing piece

This episode underscores that culture is not a residual factor but a central driver of development outcomes. Ignoring cultural dynamics risks undermining even the best-designed policies, while engaging with them can unlock new effectiveness.

As Bau, Lowes, and Montero make clear, the task ahead is to build better data, foster interdisciplinary collaboration, and design policies that are culturally conscious from the ground up. 

“We want to understand why that intervention interacts with those characteristics and whether that’s likely to be externally valid to other settings.” Bau