Using Conditionality as a Solution to the Problem of Low Uptake of Essential Services Among Disadvantaged Communities: A Social Determinants View

Abstract

Conditional cash transfer schemes, which use cash to incentivize uptake of basic health and educational services, are well established among social planners inlow- and middle-income countries and are now taking hold in high-income countries.

We appraised these schemes within a social determinants framework and found some encouraging signs in their first decade of operation. Success, however, has been inconsistent, and it is unclear whether conditional cash transfer schemes can reliably secure meaningful improvements in participants’ health and nutritional status or educational attainment.

Conditional cash transfer schemes’ objectives will not be met unless they are transformed in 3 ways: transferring power as well as resources, emphasizing entitlements alongside conditionality, and avoiding the trap of incoherent or residualized policy.

Essential welfare services, such as preventive health care, often go underused by those who most need them; the consequences of this inverse relation between care and need are well known.¹ One response to the problem has been the emergence of conditional cash transfer schemes, an increasingly popular policy to improve the welfare of disadvantaged families. In essence, the schemes offer regular sums of cash if families comply with specified behaviors such as ensuring children's attendance at school and health checkups and mothers’ attendance at health and nutrition seminars. The policy arose in response to the deep economic recessions seen in Latin America during the 1980s. Municipalities in Brazil began experimenting with conditional welfare in 1995,² and Progresa, the first nationwide conditional cash transfer scheme, began in Mexico in 1997. Since then, schemes have been rapidly replicated across Latin America, Jamaica, Kenya, Macedonia, Pakistan, South Africa, and Turkey, as well as some higher-income settings.

An evaluation of New York City's Opportunity NYC, the first such scheme in a high-income setting, was published recently. Closely modeled on Progresa, this scheme offers families eligible for food stamps in the boroughs of the Bronx, Brooklyn, and Harlem cash transfers ranging from $20 to $600 if they comply with conditions such as attendance at parent–teacher conferences and annual medical checkups.

The core problem that conditional cash transfer schemes seek to solve is that of inadequate service uptake among disadvantaged households, even where services are believed to be accessible, appropriate, and of high quality. The deeper objectives are to break the intergenerational transmission of poverty by stimulating uptake of services that support children's early education, health, and nutrition. They are used to tackle social exclusion, both by contributing to the upward social mobility of future generations and by partially relieving the immediate poverty of beneficiary households.

Good syntheses of evidence showing whether conditional cash transfer schemes meet these aims are available.^3,4 Conditional cash transfer schemes have generally been successful in targeting families most in need,^5–8 and it appears that, with few exceptions, they are strongly progressive. For example, in a cluster-randomized evaluation, Mexico's Progresa was found to significantly reduce poverty. During its first 2 years of operation, the number of people in poverty declined by 17%, the poverty gap by36%, and the severity of poverty by 46%⁹; furthermore, the scheme's poorest infants consumed the most nutritional supplements and gained the most height (equivalent to about one quarter of their estimated height deficit).¹⁰ Similarly, randomized evaluations from Ecuador and Nicaragua show progressive trends across socioeconomic position for primary school enrollment and completion and health checkups. Families participating in Opportunity NYC received more than $6000 on average during its first 2 years and reported less poverty and hardship as a result, although parents receiving the highest transfers were more educated and more likely to be working full time, married, or in a legal domestic partnership than were other parents.¹¹

Effects on health outcomes, however, are less clear. Conditional cash transfer scheme households are known to consume 10% to 20% more goods, particularly meat, fruits, vegetables, and children's clothing and footwear. Importantly, the consumption of alcohol and tobacco appears to remain constant.^12,13 Service uptake increases, as anticipated by the imposition of conditions. In Colombia, for example, Familias en Acción was associated with significant increases in preventive health care checkups for children younger than 2 years (from 17.2% to 40.0%) and children aged 2 to 4 years (from 21.3% to 66.8%).¹⁴ Similar patterns were observed in Nicaragua,¹⁵ Jamaica,¹⁶ Honduras,¹⁷ and Mexico.

Nevertheless, although the Mexican,¹⁸ Nicaraguan, and Colombian¹⁹ schemes were associated with small but significant improvements in child growth, no effect was seen in Honduras or Brazil.²⁰ Mixed results have been seen in hemoglobin levels as well,^13,15 however, and the prevalence of both anemia and stunting remained high. Cognitive and educational test scores also have shown little if any improvement (around 0.2 standard deviations at best),^21–23 and the World Bank concluded of Mexico's flagship scheme that its effect on the intergenerational transmission of poverty may be but a “small dent.”³

A similar picture emerges in New York City, where school attendance increased but overall school outcomes did not improve for elementary or middle school students. In regards to health-related outcomes, although preventive health care behaviors increased as mandated by the scheme's conditions, there were no effects on self-rated health, psychosocial well-being, smoking, body mass index, or rates of receipt of medical treatment. A possibly regressive effect is apparent: better-educated participants reported a significant reduction in health needs unmet because of cost, compared with those with less education.¹¹ Although the follow-up time for Opportunity NYC is 3 years, there is perhaps no reason to expect that its longer-run effects will be any different from the Progresa-type programs that it sought to replicate.

Because of this uncertainty over outcomes, it may be necessary to consider a breadth of additional factors when judging the effectiveness and acceptability of conditional cash transfers. Specifically, conditional cash transfer schemes focus attention entirely on the way the individual or the household behaves and intervene at this level, without considering that there may be drivers of individual and household behavior that operate at, and thus require intervention at, the social level. Here we present a fresh analysis of conditional cash transfer schemes from a social determinants perspective, seeking to draw attention to the most deeply situated causes of health inequity, and highlighting the need for action to improve the conditions within which people live, tackle unfair distributions of power and resources, and quantify the drivers of health inequity and the effects of action.²⁴ We highlight a series of issues about which the literature onconditional cash transfer schemes, although copious, says little—namely, effects on empowerment, rights, service quality, social cohesion, interrelation with other welfare policy, and participants’ views on the schemes.

TRANSFER OF RESOURCES ALONE IS INSUFFICIENT

Conditional cash transfer schemes appear to transfer resources effectively, but do they do as well at transferring power? The distinction is critical, because social exclusion will persist unless imbalances in both are tackled.

Social exclusion is driven and perpetuated by resource dependency or power dependency. Both manifest a relationship in which one party is subordinate to the other. Resource dependency results from imbalances in money, materials, skills, or expertise and is occasionally necessary, as in the relationship between a parent and child. Clearly, this form of dependency does not imply social exclusion. Other forms of resource dependency, however, are exclusionary, whether preferable to the dominant party (as characterized by abusive relationships) or to the subordinate party (as characterized by freeloader relationships). Similarly, in power dependency, weaker parties often do not choose to be in a subordinate position, but power structures keep them there.

What is the interplay between conditionality and these 2 drivers of social exclusion? Regarding resource dependency, conditional cash transfer schemes demonstrate an effective transfer of resources to the most marginalized of society, mainly through much better targeting than earlier welfare programs managed. Furthermore, transfers are generous (up to 20% of household income). After reviewing several conditional cash transfer schemes, Coady concluded that 81% of benefits go to the poorest 40% of families.²⁵ Soares estimated the effect of conditional cash transfer schemes on the Gini coefficients of Mexico, Brazil, and Chile²⁶ and found that, on average, about 60% of resources flowed to the poorest 20% of the population. Between the mid-1990s and mid-2000s, the schemes accounted for 21% of inequality reduction in Brazil and Mexico and 15% in Chile. Such figures are surprisingly large considering that conditional cash transfer scheme transfers are tiny in comparison with mainstream social welfare programs. Crucially, a recent review of evidence from low-income countries concluded that conditional cash transfer schemes do not induce financial dependence or lassitude among recipient communities.²⁷

Much less clear is conditional cash transfer schemes’ effect on power dependency. Although conditional cash transfer schemes’ core narrative of coresponsibility and specific elements (such as nominating the female head of household as the recipient for cash transfers and a greater incentive for girls’ schooling compared with that of boys) were designed to transfer power, there is little evidence on whether conditional cash transfer schemes reduce or promote this driver of social exclusion.

Only 1 study has looked at this issue. Adato et al. found that Progresa increased women's self-confidence, but it is unclear how the sample was selected.²⁸ Thus, little is known about whether conditional cash transfer schemes empower or disempower participants relative to baseline or to the rest of society. There is a real risk that conditional cash transfer schemes will do little to address social exclusion, principally because conditional cash transfer schemes transfer insufficient power. Conditional cash transfer schemes are liable to encourage a merely mechanistic engagement with public services and to fail to generate genuine demand or engagement in a way that empowers users. The deliberate bypassing of unresponsive local bureaucracies further heightens this risk. Although this bypassing is cited as a positive feature where local government is inefficient or weak, such centralization is not necessarily beneficial. Dysfunctional layers of local government should not be ignored, but reformed.²⁹ Furthermore, most conditional cash transfer schemes do not offer adequate means of responding to needs as voiced by beneficiaries themselves, meaning that they are clients rather than equal participants in solving the problem of chronic disadvantage. Taken together, these factors mean that conditional cash transfer schemes do little to address societal attitudes to poverty and may perpetuate social exclusion despite any resource transfer.

The Commission on Social Determinants of Health identifies hierarchies of power that undermine the control individuals and communities have over their lives as a fundamental cause of health inequity.²⁹ Conditional cash transfer schemes should avoid entrenching such hierarchies by giving beneficiaries a say in how schemes are designed, taking the opportunity to transform local government into effective bodies responsible for local welfare, and tempering conditionality with a stronger narrative around entitlements.

CONDITIONALITY MUST BE COMPLEMENTED BY OTHER APPROACHES

Conditionality has always provoked controversy. Some critics claim that conditional welfare is deplorable when essential to a family's livelihood. They state that conditions are drawn up by well-paid professionals with little understanding of the reality of poverty and can be demeaning, stigmatizing, or irrelevant as a result.³⁰ Others have suggested that conditions are only necessary because of a particular attitude to poverty, namely that it results from individual failure rather than lack of opportunity. Few evaluations have sought to quantify any additional benefit (or cost) of conditionality. Of those that exist, most derive from unintended mishaps in scheme implementation.^23,31–33 They find that the imposition of conditions generally improves service uptake compared with nonconditional cash transfers.

Only 1 experimental study seeking to measure the value of conditionality has been published to our knowledge. This finds that Malawian girls’ secondary school attendance was similar in conditional and unconditional transfer groups; among schoolgirls with a high propensity to drop out of school, however, the effect of the conditionality was significant.³⁴ Because of the ongoing debate on the effectiveness and desirability of conditionality, it is worth reflecting on other ways to tackle the problem of low service uptake.

Individuals will forego investments in their household's health and education if they perceive the net gain to be less than other options. Thus, efforts to increase services’ perceived net gain can be expected to increase their demand. This is possible through several means. Monetary costs, both direct (such as user fees) and indirect (such as travel costs to inconveniently located facilities) can be cut or subsidized. Nonmonetary costs, such as derogatory staff attitudes, must also be minimized or abolished. The perceived value of a service can be augmented by increasing both its true value—that is, its service quality—and its perceived value. The latter is achievable through socially and culturally tailored information that corrects misperceptions about the short- or long-term benefits of taking up health and education services. At the same time, such information should make clear users’ entitlement to effective, acceptable, and accessible services as an established and enforceable right. Such information, once given, is permanent and self-propagating and can drive a continuous process of quality improvement.

We call a program that encompasses all these mechanisms an “entitlements approach.” Its defining feature is the transfer of both power and, potentially, resources. Both types of dependency are thus addressed. The risk in a conditionality approach is that these other mechanisms to stimulate demand are forgotten or handled with only a symbolic effort. This becomes clear in the literature, in which issues such as service quality are only cursorily dealt with. For example, even though many conditional cash transfer schemes state that parallel investments in health and education were made alongside schemes, the literature barely discusses the quality of services before conditional cash transfer schemes or the effect of extra investment. Some sites report services struggling to cope with increased demand and the quality of care deteriorating.³⁰ This relative imbalance in reporting can reinforce the mindset that escaping poverty is entirely the responsibility of the poor. Additionally, conditionality does not provide any obvious mechanism to respond to user feedback or any other means for quality improvement. Indeed, nothing is known about beneficiaries’ views on conditionality itself. Considering the academic and political controversy surrounding conditionality, this represents a remarkable deficit in the literature.

Policies to ensure that conditionality is balanced with an equal emphasis on entitlements would include publication of service users’ rights backed by a legal framework of enforceability, a shift away from the assumption of adequate service quality toward an agenda of continuous quality improvement, and scrutiny and reorganization of services to remove hidden costs such as stigma and discrimination. Baltimore, Maryland's award-winning Citistat program provides an example of this approach in action. The scheme gives residents real-time data on issues such as crime and environmental quality and holds officials to account for results in public forums.

UNIVERSAL APPROACH MAY BE PREFERABLE

Targeting a service so that only certain groups (typically the poor or otherwise marginalized) receive help prevents the capture of benefits by the better-off. Targeted services, however, may entrench negative social attitudes about the nature and causes of poverty and lead to a set of interventions used only by the poor. Because these communities typically have little political leverage, such services are at risk for being low quality, incoherent, and disconnected from the opportunities and safety nets available in mainstream society—in short, “residualized.”³⁵ By contrast, capturing the interest of the better-off by widening or universalizing eligibility may secure a high-quality and sustainable intervention. This is the approach frequently used in European welfare states, where progressive universalism is recognized to contribute to social cohesion and tackle the gradient of unequal life chances throughout society.

In terms of policy, therefore, even where targeting is preferred, the potential for universalism should also be considered. It is easy to see how this would apply to conditional cash transfer schemes: certain elements could be extended to higher income groups in ways that do not pay them for doing what they would reliably do anyway. Childhood vaccination is a good example.

The related issue is that of an incoherent policy that does not recognize the multiple dimensions of chronic poverty (which is more than a lack of income)³⁶ or offer households genuine opportunities to graduate from social assistance to decent employment and training, with appropriate risk protection. A broad suite of social protection policies is required that addresses all the disadvantages associated with chronic poverty in a coordinated and inclusive manner. Such a suite would include insurance mechanisms to share risk and pool resources, emergency relief, additional investment in human capital (such as literacy and numeracy skills) beyond that currently offered by conditional cash transfer schemes, and efforts to tackle stigma. At the same time, wider macroeconomic policy must foster distribution within national growth that benefits the poor, decent labor standards, and action to secure basic rights and sociopolitical stability.^37,38

Only with action across all these fronts can effective and sustained progress to lift families out of chronic poverty be made. Although this point is self-evident, it is worth restating, because the huge international interest generated by conditional cash transfer schemes can exaggerate their overall significance within the wider portfolio of social policy.³⁹ In particular, conditional cash transfer schemes are likely to be weak instruments to overcome the fundamentally inegalitarian nature of most welfare systems, which tend to favor richer quantiles by being organized through formal employment and which have a preference for prestigious secondary or tertiary services at the expense of basic services more relevant to the poor.⁴⁰

To date, we know little about how conditional cash transfer schemes affect social cohesion. Although there is some evidence that conditional cash transfer schemes can redress intrahousehold power imbalances, there is conflicting evidence regarding intracommunity cohesion. A quantitative study in Colombia found evidence of a positive effect on social capital⁴¹; however, qualitative work in Mexico noted an increase in community tensions as a result of perceived unfairness in who was excluded from conditional cash transfer schemes eligibility.⁴² Similarly, there has been little examination of how conditional cash transfer schemes interrelate with other welfare and social protection policies. Further research is needed both from a policy analysis perspective and from the perspective of users that charts their experience as they progress within or across schemes and succeed—or fail—in graduating from welfare assistance to economic independence.

To conclude, conditional cash transfer schemes aim to tackle social exclusion by incentivizing the most marginalized in society to take up essential health and education services. In their first decade of operation, they have proven effective in stimulating service uptake, but the effect ontheir deeper aim of promoting social mobility is much less clear. An analysis and set of policy recommendations attuned to the deepest causes of ill health and health inequity can assist these schemes to meet their aim. Practically implementing the principles of transferring power, emphasizing entitlements, and universalizing elements of the conditional cash transfer schemes offer will vary from setting to setting, depending on local contexts. Further research is also needed to address some clear deficits in the literature, such as service quality and beneficiaries’ views on conditionality.