Public health and development: Infrastructure, social norms, and health behaviours

5 Conceptual framework

This section explains a conceptual framework and applies it to sanitation. This framework can be extended to other health investments that low-income countries need.

public good

A good for which use by one person does not reduce its availability to others. Also known as: non-rival good. See also: non-excludable public good, artificially scarce good.

external effect

A positive or negative effect of a production, consumption, or other economic decision on another person or people that is not specified as a benefit or liability in a contract. It is called an external effect because the effect in question is outside the contract. Also known as: externality. See also: incomplete contract, market failure, external benefit, external cost.

free ride

Benefiting from the contributions of others to some cooperative project without contributing oneself.

Sanitation provision is difficult for two reasons. Firstly, a sanitation system is a public good. It is non-rivalrous and also non-excludable: the usage of sanitation by one person does not diminish the availability of the good to others, and no individual can be excluded from using it. Public goods tend to be underprovided because everyone prefers to benefit from the investments of others without paying anything themselves (the ‘free rider’ problem). In the case of sanitation, investment is often shared between the government and households. For example, the government finances treatment plants and the households pay for emptying septic tanks. Both governments and households would prefer the other party to allocate the resources needed to ensure investments are made. This means that for an intervention to work, policymakers must ensure an adequate level of investment from both parties. (In game theory, this type of interaction is a type of coordination game: either both parties invest or neither does.)

To learn more about public goods and external effects, read Sections 10.3 and 10.5–10.7 of The Economy 2.0: Microeconomics.

Since it is a public good, an investment in sanitation by any individual has positive external effects (externalities), meaning that the private benefits to the individual (better health) are smaller than the benefits to society as a whole (reduced risk of spreading infections). These external effects explain why individual willingness to pay for sanitation tends to be low, despite the large potential benefits to society.^1 2 For example, in a survey of households in rural Kenya, only 3% of households were willing to pay for a latrine at the market price. Many of the examples of health interventions in this Insight have positive external effects, and so will be underprovided. This is why governments are needed to run or oversee health interventions such as mass vaccination campaigns, and why initiatives such as Deworm the World advocate the provision of free deworming medications at the population level to achieve better health and economic outcomes.

social norm

An understanding that is common to most members of a society about what people should do in a given situation when their actions affect others.

equilibrium

A model outcome that is self-perpetuating. In this case, something of interest does not change unless an outside or external force is introduced that alters the model’s description of the situation.

game theory

A branch of mathematics that studies strategic interactions, meaning situations in which each actor knows that the benefits they receive depend on the actions taken by all. See also: game.

coordination game

A game in which there are two Nash equilibria, of which one may be Pareto superior to the other. Also known as: assurance game.

Aside from overcoming the underprovision of sanitation, a major challenge to improving individual health outcomes is encouraging behavioural change. Within a community, there are social norms about sanitation practices. Social norms are common understandings about how individuals should behave; they are maintained through time for two reasons. Firstly, individuals believe that most people in their community conform to these norms. Secondly, most people in the community believe that these norms should be followed. The former are called empirical expectations or beliefs about what we expect others to do. The latter are called normative expectations or beliefs about what others think people should do. So in the case of sanitation, having an OD-free community means creating a new social norm that finds OD unacceptable. (In game theory, this social interaction is another coordination game: if everyone in the community stops practising OD then everyone benefits, but if only some people stop practising OD then nobody benefits.)

To learn more about how social norms affect individual behaviour, read Section 4.8 of The Economy 2.0: Microeconomics. To learn more about coordination and how coordination problems can be analysed using games, read Section 4.13 of The Economy 2.0: Microeconomics.

The presence of social norms about sanitation means that individuals are unlikely to change their behaviour unless others in the community are also doing so. Sanitation interventions therefore have two possible outcomes (equilibria): limited behavioural change (OD is prevalent), or widespread behavioural change (OD-free). To observe health benefits, there needs to be enough coordination within the community to switch to OD-free sanitation practices. Studies on sanitation programmes in India, Indonesia, Mali, and Tanzania show that there is a critical threshold (minimum percentage of adopters within the community) that needs to be achieved for community-wide health benefits to be observed: changes in sanitation below 40% coverage have minimum impacts on health but above that, benefits start to accrue. Households may therefore need a big push in investment coming from the government to start investing themselves.³ An important question for policymakers is what can prompt better coordination across agents to move from the OD to the OD-free sanitation equilibria.

A model of sanitation

To learn more about the underlying mathematics of this model, read Section 2 of Britta Augsburg, Andrew Foster, Thomas Johnson, and Michael Lipscomb. 2024. ‘Evidence on Designing Sanitation Interventions’. Journal of Development Economics 171: 103316.

Suppose that a household has wealth, \(w\), to spend on two goods: sanitation goods, which benefit the household members by improving their health (reducing their expected costs from illness), and non-sanitation goods, which we will call ‘consumption’. The price of sanitation goods (relative to consumption goods) is \(p\). If the household spends \(y\) on consumption and buys \(s\) sanitation goods, its budget constraint is \(ps + y \leq w\). Households must choose how much of each good to spend while remaining within this budget constraint.

Sanitation provision is the joint responsibility of households and the government. A household that spends \(s\) on sanitation benefits from a total amount of sanitation goods \(T =s + g\), where \(g\) is provided by the government (and is the same for all households).

To learn more about constrained choice problems and the concepts discussed in this section, read Sections 3.2–3.5 of The Economy 2.0: Microeconomics. This problem is very similar to other constrained choice problems in The Economy units, except that one of the goods is partially provided by the government. This government provision makes the household’s endowment of sanitation greater than its private contribution.

constrained choice problem

This problem is about how we can do the best for ourselves, given our preferences and constraints, and when the things we value are scarce. See also: constrained optimization problem.

The household faces a constrained choice problem. Its utility depends on consumption, \(y\), and total sanitation, \(T = s + g\). Given the government’s contribution, \(g\), it will choose \(y\) and \(s\) to maximize its utility, subject to its budget constraint, \(ps + y \leq w\).

marginal rate of substitution (MRS)

The trade-off that a person is willing to make between two goods. At any point, this is the slope of the indifference curve. See also: marginal rate of transformation.

marginal rate of transformation (MRT)

The quantity of some good that must be sacrificed to acquire one additional unit of another good. At any point, it is the slope of the feasible frontier. See also: marginal rate of substitution.

The problem is illustrated in Figure 11. The household will choose \(s\) and \(y\) so that its marginal rate of substitution—the rate at which it is willing to give up consumption goods for sanitation—is equal to the marginal rate of transformation, which is the price of sanitation goods, \(p\).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-11

Figure 11 A household’s choice of sanitation and consumption goods.

When choosing \(s\), the household allows only for the benefit to itself—the effect on its own utility—and not for the external benefits to other members of society. If the government is concerned with the welfare of society as a whole, it would like each household to increase its choice of \(s\). The model allows us to examine what policies might be used.

Figure 12 shows the effects of four policies on a household’s choice of \(s\). In all panels, point A represents the household’s choice before the policy, and point B represents the household’s choice after the policy.

To learn more about income and substitution effects; read Section 3.7 of The Economy 2.0: Microeconomics.

crowding out

There are two quite distinct uses of the term. One is the observed negative effect when economic incentives displace people’s ethical or other-regarding motivations. In studies of individual behaviour, incentives may have a crowding out effect on social preferences. A second use of the term is to refer to the effect of an increase in government spending in reducing private spending, as would be expected for example in an economy working at full capacity utilization, or when a fiscal expansion is associated with a rise in the interest rate.

income effect

The effect that the additional income would have if there were no change in the price or opportunity cost.

substitution effect

The effect that is only due to changes in the price or opportunity cost, given the new level of utility.

• Money transfers (increases in household wealth \(w\)) shift the budget constraint outwards, without changing its slope. Household spending on sanitation increases (the income effect), as shown in panel (a) of Figure 12, where \(s_2 > s_1\).
• In-kind transfers for sanitation services (increases in \(g\)): As shown in panel (b) of Figure 12, when \(g\) increases, the household’s budget constraint again shifts to the right, and the total amount of sanitation, \(T = s + g\), rises (\(T_2 > T_1\)). However, because the household is now able to rely more on government-provided sanitation, it chooses a lower level of \(s\) than before (\(s_2 < s_1\)): government spending also crowds out household investment. In effect, the household uses some of the transfer to increase its consumption of other goods.
• Information interventions: Information campaigns are widely used in low-income countries. They aim to increase households’ perceived benefits of sanitation, or (equivalently) lower the perceived costs associated with preventing sanitation-related diseases. If successful, they will increase households’ willingness to give up consumption goods for sanitation. When the MRS increases, the indifference curves become steeper (in panel (c), \(IC_2\) is steeper than \(IC_1\) at any given level of \(T\)), so the household chooses a higher level of \(s\).
• Changing the price of sanitation via subsidies (decreases in \(p\)): If the price of sanitation faced by households falls, the budget constraint pivots outwards, becoming flatter (panel (d) of Figure 12). This increases the choice of \(s\) (through both an income effect and a substitution effect).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-12

Figure 12 The effects of policies that aim to increase a household’s choice of \(s\).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-12a

Money transfers (increases in household wealth w)

The budget constraint shifts outward in a parallel manner. Due to the income effect, household spending on sanitation (the horizontal distance between \(g_1\) and their preferred choice of \(T\)) increases from \(s_1\) to \(s_2\).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-12b

In-kind transfers for sanitation services (increases in g)

The budget constraint also shifts outward in a parallel manner. Total spending \(T\) increases by \(\Delta T\), but household spending on sanitation decreases: it is crowded out by the increased government spending (\(\Delta g\)).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-12c

Information interventions

Information interventions can increase households’ marginal rate of substitution (their willingness to give up consumption goods for sanitation). Household indifference curves become steeper at every given value of \(T\), so the household prefers to consume a higher level of \(s\).

Fullscreen

https://books.core-econ.org/insights/public-health-and-development/05-conceptual-framework.html#figure-12d

Changing the price of sanitation via subsidies (decreases in p)

The budget constraint pivots outwards from the endowment point (\(g_1, w_1\)), becoming flatter. Both the income and substitution effect result in a higher choice of \(s\).

However, such interventions may not work in such a straightforward way in practice. Information campaigns do not always bring about the desired behavioural changes because of adoption thresholds. Individuals may be aware of the negative health effects of poor sanitation, but they are also aware of the potential coordination problems, so information might not lead to increased investment unless a significant part of the community also invests. For example, communities in rural Mali are aware of the consequences of poor sanitation, but since sanitation requires collective action, they fail to coordinate building latrines in a way that will protect the community from health risks. And studies show that while lowering the price of sanitation goods can increase demand for sanitation in the short term, these effects may not be sustained, especially when the households must continually invest in maintaining the sanitation equipment—for example, having to empty the latrines or septic tanks. In practice, a mix of strategies or tailored approaches are needed to improve sanitation behaviours.

The next two examples show how our conceptual framework can be used to explain higher demand for sanitation either resulting from infrastructure investments or behavioural change campaigns.

Example 1: Increased government spending on sanitation infrastructure in India

India is the country with the highest OD rates in the world. There, OD causes around 9% of total infant mortality, or 6.5 deaths per 1,000 infants per year.^4 5 India’s size, poor sanitation coverage, and the widespread practices of OD present the following challenges:

• Urban WASH: Investment in infrastructure does not encompass rapid city growth due to rural-to-urban migration. As a result, some urban areas, such as informal settlements, may not yet have sanitation infrastructure.
• Rural WASH: OD in the rural areas is very high and adoption of better sanitation practices tend to be slower than in the cities. There are two main reasons for this: the cost of building toilets (most rural households think toilets are too expensive),⁶ and social and cultural norms related to OD.

The Swachh Bharat Mission Clean India Campaign was initiated in 2014 with the objective of making India free of open defecation by 2019. It mobilized nationwide participation and was one of the largest behavioural change campaigns worldwide. By 2019, 22 billion USD was invested, and more than 100 million household toilets were constructed.⁷ Despite such efforts, the consistent use of toilets is not yet universal. The campaign has not been enough to eradicate OD or translate into health improvements. Our conceptual framework can explain why toilet construction alone will not eradicate OD: this behaviour is related to social norms which do not necessarily think of OD as being harmful or frowned upon.

To learn more about social norm theory, read Cristina Bicchieri. 2017. Norms in the Wild: How to Diagnose, Measure and Change Social Norms. Oxford University Press.

Individual behaviour is influenced by individuals’ social expectations about what others are doing. For a new social norm to emerge, the new behaviour must be observable and apparent to others. Even when behaviour at the individual level has changed, it might not be apparent to other individuals in the community. So in the case of sanitation, even when toilets are present, people may still resort to OD if they think that others in the community find the behaviour acceptable. A successful sanitation campaign should change an individual’s social expectations about toilet usage, encouraging the adoption of new behaviours.

To investigate why the Clean India Campaign did not eradicate OD, researchers conducted a large-scale survey in Bihar and Tamil Nadu, two Indian states.⁵ The former has a poorer and less-educated population than the latter. Among the total population, 50% of men and 30% of women report practising OD in each state. The researchers also found that people who had toilets were using them, and toilet ownership was highly correlated with household income. This result suggests that financial support to lower-income households may help improve adoption. Social networks and spatial proximity also affected toilet usage and adoption. The authors found that normative expectations about others in the community (if they approve the use of a toilet instead of OD) had no influence and usage was predicted by what they thought those physically close to them were doing (known as empirical expectations). Empirical expectations are likely to be influential in this context due to social pressure, peer monitoring, and fear of social sanctions if caught practising OD.

Example 2: A behavioural change flagship programme in Bangladesh

Community-Led Total Sanitation (CLTS) is a flagship sanitation programme created by Robert Chambers. It seeks to induce people to use latrines instead of practising OD, and was first implemented in Bangladesh.⁸ This programme increases awareness of better sanitation practices and makes cheap solutions (such as pit latrines) easily available, which in turn increases demand for safe sanitation. The programme is considered successful (in a community) when all members commit to eliminating OD and achieve this goal. Another distinctive feature is that OD-free communities earn a certificate that can be shown to government or NGO authorities, stating that nobody practices OD in their community. This certificate can incentivize good sanitation practices through public recognition and establishing the norm that communities should be OD-free.

Decisions over sanitation can be made based on incomplete information on the programme. Individuals may ignore the health benefits, other benefits such as privacy and comfort, or the cost of construction. Latrines are also considered ‘experience goods’: in low-coverage areas, individuals may not be familiar with some attributes of its use. To learn more about the factors that affect sanitation decisions, read:

María Laura Alzúa, Habiba Djebbari, and Amanda Pickering. 2020. ‘A Community-Based Program Promotes Sanitation’. Economic Development and Cultural Change 68(2): pp. 357–90.

Pascaline Dupas. 2014. ‘Short-Run Subsidies and Long-Run Adoption of New Health Products: Evidence from a Field Experiment’. Econometrica 82(1): pp. 197–228.

Our conceptual framework can predict the possible outcomes of this programme. For an individual, demand for sanitation (using a latrine) depends on its perceived benefits and costs (money, time, health outcomes), but is also influenced by what other members in the community are doing. There are positive external effects (externalities), because the health benefits of using and building a latrine (a cleaner surrounding environment) extend beyond the household paying for the latrine. So, an equilibrium of poor sanitation might arise if community members’ efforts are substitutes (one member building or using a latrine makes it less likely for other members to do so), leading to the underprovision of latrines. On the contrary, if the incentive to adopt latrines increases with the number of people using latrines (for example, via peer pressure or shared infrastructure), an equilibrium of good sanitation might arise. A sanitation programme such as CLTS can help overcome coordination failures because it increases the social pressure to conform to the practice of using latrines.