Can behavioural economics improve the impact of development programs? Reflections on the World Development Report 2015

By Mariza Cooray-Dias
30 March 2015

The latest World Development Report [pdf] provides evidence that insights from psychology and economics can be utilised to improve policy decisions. Within the context of development programs, our understanding of incentives and behaviour are usually reflected within the logic of the program. Insights from behavioural economics provide substantial evidence that a number of assumptions within these frameworks, in relation to target groups including the poor, are, in fact, wrong.

The application of insights from economics and psychology, the core disciplines of behavioural economics, account for the behavioural cues formed in our individual identity and biases. Behavioural insights have seemingly impacted public policy worldwide including in Australia. Using experimental evidence, the World Development Report examines the use of behavioural cues, biases and messages to further improve outcomes within international aid and development programs.

Thinking automatically, socially and with mental models

The three elements in Part 1 of the report explore the assumptions we make during our process of thinking. The report:

  1. provides examples of when we default to thinking automatically, unlike when we use the application of science to consider all possible assumptions in a study;
  2. explores our association with social norms and networks; and
  3. takes stock of when we respond to experiences using our own constructs, shaped by our environment, instruction and training, which are known as mental models.

As suggested in the following diagram, a more behavioural model of thinking (on the right) accounts for when we process information that is more salient to us or when we choose to make decisions that serve us in the present, although we are aware they will not necessarily benefit us in the future (the classic example of Walter Mischel’s ‘marshmallow test’ comes to mind).

A more behavioral model of decision making expands the standard economic model

A more behavioral model of decision making expands the standard economic modelCorrectly framing the message of an intervention is a key insight highlighted throughout the report. For instance, a randomised experiment addressing the risk of HIV infection among teenage girls had divergent campaigns to test their efficacy. One information campaign provided ‘risk avoidance’ (abstinence) education. The other focused on ‘risk reduction’ (condom use). The latter campaign was more effective because specific facts including the age of riskier partners were explained as part of the campaign. The ‘risk reduction’ campaign was more successful because it targeted the environment, instruction and training (mental models) of the teenage girls involved (Dupas, 2011 [pdf]).

Using behavioural insights in design

Within the context of program design, recognising that individuals use mental models can advance the way we make assumptions about target groups and communities. A key insight provided by the report is that investing resources and time in the definition and diagnosis of problems at the beginning of a program led to better-designed interventions that fulfilled their intended outcomes.

What does this mean for the future of program design? The report recommends that the design of an intervention be trialled, adopted and tested prior to its release within a controlled environment. As part of the design process, the behavioural bottlenecks of an intervention need to be addressed and diagnosed, even before being piloted. For example, institutions frequently use targeted interventions to prevent or promote certain habits and behaviour, although they are unsure of the messages and how they will be perceived by target audiences. In one experiment, inadvertently reminding individuals of their gender identity negatively affected their performance.

The process of program design should be dynamic, allowing for an understanding of how interventions are perceived, how they work and the kinds of behavioural responses elicited by target groups. An investigative approach allowing for experimentation and prototypes prior to carrying out the pilot stage of an intervention is a crucial step in improving the impact of development programs.

Mariza Cooray-Dias is an Economist interested in program and policy evaluation. She has worked on a number of projects while at Abt JTA.

About the author/s

Mariza Cooray-Dias
Mariza Cooray-Dias is an Economist working on program and policy evaluation in the Pacific.

Page 1 of 1