Introduction

Social norms often sustain gendered discrimination. At the same time, however, interventions that aim to shift those norms can become catalysts for change. The measurement of norms is critical in analysing both of these processes and is, therefore, attracting growing attention from researchers, policymakers and practitioners. Researchers have explored whether norms fuel specific practices and the strength of their hold using quantitative and qualitative methods, both individually and in combination. Policymakers and practitioners have, in turn, incorporated this information into the design of initiatives to drive change, and into the monitoring of such initiatives. 

This Guide focuses specifically on quantitative measurements, noting that this field is still nascent, and that far more work is needed to strengthen and streamline existing methods. It provides an overview of quantitative approaches to measure gendered social norms, highlighting the following.

• the value of measuring social norms
• the components of social norms that can be measured
• specific instruments to measure social norms
• broader data instruments that capture evidence on social norms 
• the need for diverse approaches
• useful resources. 

The value of quantitative measurement of norms

Quantitative approaches are best placed to illuminate – at scale – concepts and components (see next section) that can be measured, including expectations, sanctions, and individual attitudes and behaviours. Their emphasis on statistical representativeness makes it possible to draw conclusions on the prevalence and strength of norms within a given setting – be it a village, a country or across groups of countries – as well as any shifts over time. Quantitative approaches are, therefore, an indispensable part of monitoring and evaluation efforts because they provide standard, objective scales that can be compared across different places and time periods. 

Quantitative data are also well-placed to illustrate how norms are distributed within a population and whose opinions different groups of people find influential.    

Research by the non-governmental organisation (NGO) Tostan, for example, shows how the norms that underpin female genital cutting (FGC) varied across Mali. Their initial analysis was highly contradictory: 70% of respondents reported that others did not participate in this practice while 30% reported that everyone did. More granular analysis found that the practice was prevalent in some villages, but had almost vanished in others (Tostan 2014, cited in Heise and Cislaghi 2016: 23). In contrast, in Brazil, Pulerwitz and Barker (2008: 333) show considerable variation in support for gender-equitable norms among young men in the same locations: ‘in the same neighbourhoods, in the same households, in the same schools.’ 

Such information on who adheres to a given norm is of vital importance when designing strategies to improve gender equity. In the Mali example, where FGC was widespread, community-wide awareness programmes may be advisable to tackle the practice. In the Brazil example, with varied support for gender equity, practitioners might urge individual champions who hold more progressive views to persuade others.

The work of social norms theorists confirms that what people expect of themselves may not always align with what they expect of others, or even with their own behaviour. By asking systematic questions, quantitative measures can determine whether distinct elements of a social norm are in place – information that is useful in understanding its presence and strength and, therefore, in informing change.

Take the example of Saudi Arabia. As highlighted by Bursztyn et al. (2018) fewer than 15% of women in the country were in paid work in 2017, but legal changes – such as the lifting of the ban on women driving – had created a more favourable environment for female employment. An online and anonymous survey of young married men showed that the vast majority (87%) supported women working outside the home, while dramatically underestimating such support among other men, including their neighbours. An experiment aimed to ‘correct’ the beliefs of a random group of the men surveyed by informing them of the true level of support for female labour-force participation among their peers. These men were then more willing to forego a monetary incentive (an Amazon gift voucher) and instead to sign their wives up to a job-matching service. Four months later, their wives were more likely to have applied and interviewed for a job outside the home than the wives of other men who remained unaware of widespread male support for women’s employment. 

Despite the potential benefits of such quantitative approaches, the development of norms measures, and measurement tools, is still nascent. In a 2015 review of 173 publications on social norms, Mackie et al. (2015) find that just 14% discussed measurement methods. Similarly, a 2016 review of 42 studies by the Institute for Reproductive Health and Save the Children on interventions to improve adolescent and youth sexual and reproductive health by focusing on norm change, only four were explicit about the norms being measured. The authors conclude: ‘The need to measure normative change and the absence of such measures in reviewed documents indicate it is not well understood and/or not prioritized as an outcome’ (p. 3). 

Key measurable components of social norms

The quantitative measure of norms should not be the first step in a research process, as noted by Cislaghi and Heise (2017) and Stefanik and Hwang (2017). Rather, researchers are advised to start by exploring the presence of norms through exploratory and open-ended qualitative questions. Once there is evidence that an existing norm may be sustaining a given practice, researchers can use quantitative tools such as survey questions or vignettes to gain further insights.  

Based on social norms theory, efforts to establish the presence of norms generally try to assess whether its key components are in place (see figure below), though they may vary in the weight that they assign to them. Most measurement approaches draw on the work of Bicchieri (2006 and 2016), who contends that a norm is made up of two types of social expectations:

• empirical expectations, or behaviours considered typical within a group (also known as descriptive norms) and 
• normative expectations, or behaviours considered appropriate within a group (also known as injunctive norms). 

Measurable components may also include sanctions, or any consequences (whether positive or negative) that result from rejecting or conforming to social expectations, as well as an assessment of reference group(s), i.e. the people whose opinions and behaviours have the greatest influence in relation to a given norm. Finally, measures of individual attitudes and behaviours can show the extent to which individuals within a community endorse and/or act in accordance with a given norm. 

Terminology and definitions used in norms measures

Source: Stefanik and Hwang (2017) p. 9.