A critical interpretive synthesis of informal payments in maternal health care

Abstract

Informal payments for healthcare are widely acknowledged as undercutting health care access, but empirical research is somewhat limited. This article is a critical interpretive synthesis that summarizes the evidence base on the drivers and impact of informal payments in maternal health care and critically interrogates the paradigms that are used to describe informal payments. Studies and conceptual articles identified both proximate and systems drivers of informal payments. These include norms of gift giving, health workforce scarcity, inadequate health systems financing, the extent of formal user fees, structural adjustment and the marketization of health care, and patient willingness to pay for better care. Similarly, there are proximal and distal impacts, including on household finances, patient satisfaction and provider morale. Informal payments have been studied and addressed from a variety of different perspectives, including anti-corruption, ethnographic and other in-depth qualitative approaches and econometric modelling. Summarizing and discussing the advantages and disadvantages of these and other paradigms illustrates the value of an inter-disciplinary approach. The same tacit, hidden attributes that make informal payments hard to measure also make them hard to discuss and address. A multidisciplinary health systems approach that leverages and integrates positivist, interpretivist and constructivist tools of social science research can lead to better insight. With this, we can challenge ‘master narratives’ and meet universalistic, equity-oriented global health objectives.

Key Messages

• Informal payments can best be understood by taking a multidisciplinary approach.

• Extant research indicates that informal payments are caused by multiple contextual factors.

• Factors include resources, governance, norms, knowledge and beliefs.

• Informal payments can impact patient welfare, quality and health system functioning.

Introduction

At their most vulnerable moments, labouring women may be confronted with coercive, financially taxing demands for informal payments in order to receive the health care to which they are entitled (Afsana, 2004; Riewpaiboon et al., 2005; Tibandebage and Mackintosh, 2005; Lewis, 2007; Kruk et al., 2008; Mæstad and Mwisongo, 2011; Vian et al., 2012; 2015; Pieterse and Lodge, 2015). Following Gaal et al.’s (2006a) definition, we use the term ‘informal payments’ to describe a cash or in-kind payment made to health care providers for a service to which the patient is entitled and which is additional to any officially sanctioned contribution required. This includes payment for care or for material entitlements, such as medicines. Informal payments are a subset of the broader category of ‘out of pocket payments’, which, in addition to informal payments, includes formal user fees and any other costs incurred while seeking and obtaining health care, such as transport.

Despite significant financial and political investment in maternal health in the Millennium Development Goal era, informal payments continue to undercut ambitious plans to enhance access, utilization and quality of prenatal and delivery care. Moreover, demands for payments are often experienced by poor women as yet another moment when governmental employees prey on them, rather than providing succour and reaffirming the entitlements of citizenship (Diarra, 2012; Coffey, 2014; Dasgupta et al., 2015). At the same time, the frontline health workers demanding these payments may be struggling to fulfil their professional mandate in a health system characterized by inadequate infrastructure and inputs, little or even punitive supervision, and poor morale and trust (Tibandebage and Mackintosh, 2005; Aberese-Ako et al., 2014; Hahonou, 2015).

Quantitative and qualitative peer-reviewed and grey literature studies focused on informal payments or related health systems issues have found informal payments to be prevalent in health care (including but also beyond maternal care) in many low- and middle-income countries (LMICs) in Latin America, Asia, Africa and the former Soviet Union (fSU; Lewis, 2007; Gao et al., 2010; Mæstad and Mwisongo, 2011; Paredes-Solís et al., 2011; Brody et al., 2013; Arnold et al., 2014; Coffey, 2014; Abdallah et al., 2015; Vian et al., 2015; Bertone and Lagarde, 2016; Kankeu and Ventelou, 2016; Habibov and Cheung, 2017). Reported prevalence rates vary significantly; the studies cited above, e.g. vary from 20% to 70%.

Researchers and programme evaluators often identify informal payments as health system factors that make women less likely to deliver in a health facility (Dasgupta et al., 2015). Civil society groups and activists routinely decry their impact, and, in some countries, informal payments are regularly discussed in the print media (Gopakumar, 1998; Thampi, 2002; Chandra, 2010; Karmakar, 2015; Wojczewski et al., 2015; Mudur, 2016). Yet, given their primacy in the patient experience, some aspects of informal payments are comparatively under-addressed in research, policy and programmes. Research gaps include those relating to the experience of informal payments in certain regions, notably sub-Saharan Africa; as well as thematic and conceptual gaps such as how patients experience informal payments, and how informal payments can be understood within the complex ecology of health service facility-level dynamics (Kankeu and Ventelou, 2016). In brief, the ‘on the ground’ salience of informal payments to understanding both access to, and quality of, maternal health care is not matched by top-down attention and action.

This article is a critical interpretive synthesis (CIS) that summarizes the evidence base on the drivers, and impact of informal payments, and critically interrogates the paradigms that are used to describe informal payments. The intent is to provide a comprehensive synthesis of ‘what we know’ about informal payments; and then to step back, assess the theoretical bases of ‘what we know’ and make propositions regarding the strengths and weaknesses of how the phenomenon has been researched and understood. This research offers researchers, policymakers and donors a broad picture of research and theory, helping them to situate the more parsimonious studies of prevalence and drivers, and to identity and critically engage the assumptions in research and policy articles. Our key concern is maternal health. However, given the fact that there is relatively little research on informal payments within maternal health care specifically and that most frontline providers and communities draw conclusions about the health system based on their interactions with all types of health providers—not just maternal health providers—we often speak about informal payments and access to health care more broadly.

CIS facilitates broad-based, multidisciplinary exploration of topics of interest. In contrast to systematic reviews, CIS is inductive and iterative (Dixon-Woods et al., 2005; Heaton et al., 2012; Wilson et al., 2014), and it facilitates exploration of a heterogeneous body of literature (Moat et al., 2013). Beyond aggregating and/or synthesizing data, CIS also enables identification of new analytic constructs, synthesizing arguments, and questions (Flemming, 2010; Moat et al., 2013; Wilson et al., 2014; Ako-Arrey et al., 2016). It has successfully been used to explicate health systems questions in high-income countries (Dixon-Woods et al., 2006; Flemming, 2010; Entwistle et al., 2012), and on a limited basis, in reference to health systems in LMICs (McFerran et al., 2017).

Given the current state of knowledge on informal payments, CIS is particularly apt. As a cross-cutting health systems and governance concern, informal payments have been described and addressed from a variety of different fields and paradigms. Respecting a ‘principle of pluralism’ reveals how different approaches can illuminate the problem as a whole (Greenhalgh et al., 2005). Synthesizing discussions across these approaches and putting them in dialogue with one another in light of the empirical evidence highlights the contributions of each approach. There are several non-systematic, reviews of informal payments that focus on particular geographic regions or that appear in the grey literature; as well as published reviews of related issues, such as how to define informal payments, and the abolition of formal user fees (Gaal et al., 2006a,b; Lewis, 2007; Vian, 2008; Cohen, 2012). There are two related systematic reviews—one on methods for assessing the burden of informal payments (Khodamoradi et al., 2018), and one on defining informal payments in health care (Cherecheş et al., 2013). Building on these studies, CIS facilitates exploration of the terminological and conceptual confusion that characterizes discussion of informal payments, and of the research on broader health systems concerns that are germane to informal payments, but that do not take informal payments as their central focus (Gaal et al., 2006a). In other words, we seek to go beyond the aggregation of insights that are contained in articles focused on informal payments, and to synthesize insights included in articles that explore informal payments as part of a broader health systems analysis. In brief, rather than understanding informal payments as a dependent variable, i.e. shaped by independent variables, we looked at how these payments are rooted in an overall cultural, social, political and economic system, and how this system iteratively interacts with informal payments.

Methods

The CIS is grounded in a health policy and systems research framework, and adopts the premise that health systems are core social institutions. A health system perspective entails exploring how practices at the frontlines are embedded within the larger system, including across levels of the health system and across health concerns (Gilson and Daire, 2011). As such, our search explicitly sought out insights from more positivist approaches to describing and prescribing, such as classical microeconomics and epidemiology, as well as significant research using interpretivist and constructivist approaches from the social sciences. The latter are particularly helpful for understanding ‘how health system actors understand and experience particular services or policies and what social and political processes, including power relations, influence them’ (Gilson et al., 2011).

This CIS was completed in a multistage process. The initial questions that guided our practice were:

• What empirical evidence do we have regarding the drivers and impacts of informal payments for (maternal) health in LMICs?

• What paradigms and approaches are used to assess and address informal payments, and what are the advantages and disadvantages of these?

Consistent with the CIS approach, these questions served as a compass, rather than an anchor, for the research; we followed relevant strands in the literature as they emerged, rather than establishing a priori areas of interest (Eakin and Mykhalovskiy, 2003; Dixon-Woods et al., 2005; Ako-Arrey et al., 2016). The steps we took included the following.

Our article selection process is illustrated in Figure 1. The first phase employed diversity sampling; we sought to get an overall view of the empirical research, social science theory and programme types focused on informal payments. Consistent with our cross-disciplinary interest and as per methodology for CIS, the criteria for article inclusion related to relevance and the likelihood that the article would contribute to theory development, not to study design or to a prima facie set of quality indicators, which would be difficult to apply to a heterogeneous collection of literature (Wilson et al., 2014; Ako-Arrey et al., 2016). We began with relevant articles we knew about, and searched ScienceDirect, PubMed, GoogleScholar and Google, using the terms ‘informal payments AND health’, ‘bribes AND health’, ‘out of pocket payments AND health’ and ‘corruption AND health’. We did not establish any limitations regarding when the article was published. All empirical studies from LMICs as well as from countries of Eastern Europe (EE) and the fSU were included for initial review. EE and fSU countries were included because several of them are middle income, and because there has been a significant amount of scholarship on informal payments in these countries. Finally, a few articles containing significant theoretical or conceptual discussion but with data from high-income countries were included. Of the 260 articles identified for initial review, we selected 59 for inclusion in our synthesis; we chose these 59 after reading the abstract and deciding whether or not they were indeed focused on informal payments and would thus aid theory development. We excluded, e.g. articles that were about out-of-pocket payments in general, but that did not acknowledge informal payments.

Fig. 1:

Article selection process.

We then read these 59 articles and identified new areas of research that we felt would further illuminate the proximate and distal drivers and impacts of informal payments (Dixon-Woods et al., 2006). We searched these terms in GoogleScholar and ScienceDirect. The new areas of research were patient satisfaction, disrespect and abuse, formal user fees, trust, and health service utilization. The specific search terms applied included: ‘patient satisfaction’, ‘disrespect AND abuse AND health’, ‘user fees AND health’, ‘trust AND health’, ‘health AND utilization’ and ‘health AND acceptability’. The objective of following these lines of enquiry was to place informal payments in a larger conceptual and policy context, rather than to review exhaustively the literature in each of these areas (Moat et al., 2013). Based on the abstract, the most relevant articles (usually 2–4) were chosen for each set of search terms. We determined relevance based on how much the article addressed informal payments. For example, did the article just mention the broad category of out-of-pocket payments as a driver of low patient satisfaction, or did it discuss informal payments specifically, and how patients experience or interpret these? None of these articles focused exclusively on informal payments, but they helped us to better understand the context of informal payments. A total of 19 articles for our synthesis were identified and added in this phase.

These 78 articles were then entered into an extraction tool that included fields for methods, drivers, impacts and key conceptual points, such as the author’s perspective on whether or not informal payments constituted corruption. There were no pre-set categories for the key conceptual points; we identified them inductively. We also synthesized data as we entered it into tool, and in so doing, started to draw conclusions. For example, our analysis of the extent to which institutionalization of formal user fees increased or decreased informal payments was entered into the ‘drivers’ field, irrespective of whether the study authors were examining formal fees as a driver per se. The drivers, impacts and key conceptual points then informed the structure of the article. The citation list of all but the most tangential articles was assessed for additional relevant articles that may have been missed (snowballing). This phase helped us to deepen our understanding of the contextual health systems issues, as we sought out theoretical and conceptual work underpinning some of the tensions that had emerged in our analysis, such as whether or not informal payments should be considered corruption and how they relate to structural adjustment. A further 27 articles were identified in this way and then also entered into the extraction tool and analysed in the same way as the initial set of 78 articles, bringing the total number to 95 articles. If these articles reinforced or contradicted conceptual points we identified earlier, we recorded this in the tool. If they raised new points, we added these as well. We also developed memos on topics that emerged and could not be adequately entered into the tool; some of these memos were used in early drafts of the article.

As the writing process was near completion, we did a final search (using our initial search terms) on GoogleScholar, PubMed and ScienceDirect for any new articles related to informal payments for health care in LMICs that may have appeared since our initial search. Ten new articles were identified in this way, and the findings were integrated into our extant draft. We integrated the articles by citing them if they confirmed or undercut existing points, or expanding arguments where they offered new insights. All told, our synthesis was informed by 115 articles, of which 100 are from the peer-reviewed literature and 15 of which are grey literature publications.

The first author conducted the substantive review and summaries, and discussed findings and interpretation with the second author.

Results and discussion

Types of informal payments

As per our definition of informal payments, we only considered a payment to be informal if it was for a service, i.e. part of the standard entitlement. In some countries, entitlements may include food and laundry in the health facility, support for transportation to and from the facility, and other such ancillary services. In the case of labour and delivery care specifically, patients report being asked to pay for drugs and other medical supplies; non-medical supplies; blood; laboratory tests; birth registration and other needed documents; to receive a facility delivery-related conditional cash transfer to which they were entitled; to see and hold their newborns; and for doctors, nurses and other providers and health facility staff to provide medical care (Killingsworth et al., 1999; Afsana, 2004; Tibandebage and Mackintosh, 2005; Lewis, 2007; Chuma et al., 2009; Human Rights Watch, 2009; Stringhini et al., 2009; Hunt, 2010; Stepurko et al., 2013; Tumlinson et al., 2013; Arnold et al., 2014; Dasgupta et al., 2015).

Moreover, through our readings, we identified significant reliability and validity challenges to the measures researchers use. First, many studies relied on patient self-report. Yet, several studies showed that patients were often unable to distinguish between official user fees and informal payments, so self-reported survey information may not accurately capture informal fee prevalence as distinct from formal user fees (Killingsworth et al., 1999; Falkingham, 2004; Mamdani and Bangser, 2004; Gaal et al., 2006b; Lewis, 2006; 2007; Cherecheş et al., 2013). Often, patients pay a combination of both (Killingsworth et al., 1999; Afsana, 2004; Perkins et al., 2009). They may also be deliberately misled about what they are paying for, such as being told that they are paying for necessary drugs when they are not (Sharma et al., 2005). Second, the distinction between gift giving and informal payments can be blurry. In surveys, patients report giving both, with the most widely accepted distinction being whether the money was provided prior to or after care was received, with money given before care being understood as a payment and money given after care understood as a gift (Balabanova and McKee, 2002; Tatar et al., 2007; Cherecheş et al., 2013). However, there are reports of very forceful demands for informal payments being made after the provision of care, so this distinction between voluntary and involuntary and when the service is provided does not always hold (Afsana, 2004). Moreover, some report giving ‘gift assurance’ to improve the quality of care provided, suggesting that the gift is understood to be necessary in order to receive appropriate care (Ayanore et al., 2018).

Our review identified other measurement challenges related to prevalence in addition to inability to distinguish between informal payments and other types of financial outlays. Respondents—both patients and providers—may be reluctant to report engaging in practices that are not openly discussed and that may be associated with corruption (Vian, 2008; Lindkvist, 2013; Abdallah et al., 2015). Also, informal payments may be so normalized that respondents do not mention them when they are asked about payments for health care as part of a wide-ranging household survey. Indeed, household surveys generally reveal lower informal payment prevalence rates than small, dedicated surveys, where interviewees are asked multiple detailed questions about payments and their responses are probed (Balabanova and McKee, 2002; Lewis, 2007).

Measurement challenges also reflect deeper conceptual and definitional challenges. Informal payments may or may not be illegal. Even if they are illegal, they could be widespread and considered to be legitimate (Gaal et al., 2006a). In the same facilities, there can be many types of informal payments. They can vary in terms of who is making the payment, to whom, how much they are giving, when the payment is made, where it is made, and for what reason (Sharma et al., 2005; Gaal et al., 2006a). Payments may be made to the treating doctor, nurse or other medical professional; an administrator; pharmacist; janitorial or other facilities employee; or someone else. Many patients (and their families) make multiple payments to multiple people during an extended interaction with the system (Sharma et al., 2005; Jeffery and Jeffery, 2010; Mæstad and Mwisongo, 2011). They may consider some of these to be gifts and others to be coerced. Patient and provider interpretations of payments vary enormously as well; patients and providers may reportedly have different interpretations of the same interaction as well as of the phenomenon as a whole. Moreover, practices and interpretations are embedded in the larger health system; ‘each transaction is thus understood, not as a one off market event, but rather as shaped by information, expectations, levels of trust, norms of behavior and incentives, all of which evolve over time through market and other social interaction’ (Tibandebage and Mackintosh, 2005).

Of those studies that differentiate among different types of services, most find that informal payments may be particularly prevalent in the obstetric care setting. First, studies have concluded that payments are more likely to be made—and are higher—for inpatient care and/or for specialist care, either or both of which are usually entailed in delivery care (Killingsworth et al., 1999; McPake et al., 1999; Miller et al., 2000; Riewpaiboon et al., 2005; Lewis, 2007; Vian, 2008; Aarva et al., 2009; Perkins et al., 2009; Baji et al., 2012; Mokhtari and Ashtari, 2012; Joe, 2015; Vian et al., 2015). Though few studies examine payments in such granular detail, it appears that even as compared with other reasons for inpatient care, obstetric care may be more likely to incur informal payment (McPake et al., 1999; Falkingham, 2004; Riewpaiboon et al., 2005; Aarva et al., 2009; Mokhtari and Ashtari, 2012; Stepurko et al., 2013). For example, a study on payments for healthcare in Hungary found that those receiving inpatient care were more likely to make an informal payment than those receiving outpatient care, and, of those receiving inpatient care, patients receiving labour and delivery care were even more likely than those receiving other services (Baji et al., 2012). The dynamics of obstetric care delivery, in particular, may contribute to higher rates of informal payments. Women are often urgently in need of care and they and their families have insufficient time to negotiate, leaving them with little leverage. They are also concerned with the health of both the mother and the newborn (rather than just one person as in most interactions with the health system); some women are even asked to pay after delivery in order to see the newborn, often more for a boy (Sharma et al., 2005; McPake et al., 1999; Holmberg and Rothstein, 2011). Moreover, in some settings, obstetric care entails a long-term (6 to 10 months) relationship with the same obstetrician. Studies from Thailand and Ukraine found that women were willing to pay to achieve interpersonal trust and care, as they intend to rely on the same provider through the pregnancy and delivery; women paid to facilitate the relationship, and to ensure that the doctor they paid would indeed deliver their baby (Riewpaiboon et al., 2005; Stepurko et al., 2013). With the exception of the Thailand and the Ukraine examples, overall, the studies indicated that providers may have used the antenatal period to lay the groundwork for demanding significant informal payments at the time of delivery, but we did not see evidence that antenatal visits (as opposed to delivery care) themselves were more or less likely than other visits to entail an informal payment.

Drivers

Studies and conceptual articles identified both proximate drivers of informal payments, as well as distal systemic causes. The drivers below are common to many studies, though some are more common in specific types of studies. For example, human resource scarcity is most often examined in economic analyses. Bringing together these analyses from different traditions helps us to arrive at a richer and more profound understanding of informal payments, as well as of how they are understood (Gilson et al., 2011). We start with proximal causes and move on to systemic causes.

Norms of gift giving and reciprocity

Patients and providers sometimes assert that gifts or payments are consistent with cultural norms of gift giving and reciprocity, particularly in the context of the government health system, where patients are usually receiving care from someone of a higher social status (Gaal and McKee, 2004; Chiu et al., 2007; Mokhtari and Ashtari, 2012; Vian et al., 2012; Nekoeimoghadam et al., 2013; Cohen and Filc, 2017). However, ascertaining to what extent norms of gift giving play a role is challenging. Some patients may consider that they are giving a gift or a tip, while another patient may consider a very similar transaction not to be tipping. These differences in interpretation can be seen in large-scale surveys. For example, in response to a survey question in the 2008 round of the Vietnam Household Living Standards Survey about whether a government official receiving a ‘small gift or money after performing duties’ was corruption, 45% said yes, 37% said no and 18% were undecided (World Bank, 2010b). Researcher interpretations are layered on top of patient interpretations. We found a variety of researcher interpretations of the extent to which gratitude played a role; these differences may be due to real differences in the countries being studied, the researcher’s primary research questions and area of interest, and, the researcher’s personal feelings regarding payment for health care. Some researchers argue that ‘the concept of “gratitude payment” is no more than a convenient myth that has been used to make an unacceptable phenomenon acceptable’ (Gaal, 2006), an outlook, i.e. echoed in varying degrees in the broader health systems literature (Dasgupta et al., 2015) and in the human rights literature (Feinglass et al., 2016). At the same time, there are others who insist that norms of gift giving—such as in China—‘cannot be reduced to a modern western notion of corruption because the personalistic qualities of obligation, indebtedness, and reciprocity are just as important as transactions in material benefit’ (Yang, 1994, p. 108). There is variation in the extent to which researchers find support for the gift giving hypothesis, though it is fair to say that the notion of gift giving and tipping is evoked in nearly every global health article about informal payments. We did not find any articles or researchers who dismiss the entire phenomenon in a LMIC as patient-driven gift giving or tipping, and only one—from Iran—that concluded that expressing appreciation was the most important motive for making informal payments (Aboutorabi et al., 2016). Thus, among the research and analysis focused on LMICs, there is widespread agreement that, while there may be a cultural element, the economy of informal payments cannot be reduced to gift giving; there are other drivers at play.

Scarcity of providers

Scarcity of providers is often noted, though it is not extensively explored, as a cause of informal fee charges. Among other factors, scarcity is putatively caused by low salaries, maldistribution and inadequate opportunities for medical education and training (Chen et al., 2004; Rowe et al., 2005; Willis-Shattuck et al., 2008; Frenk et al., 2010). The assumption is that there are too many patients for the number of health providers, so the ‘market price’ of seeing a health care provider is increased; providers hold a monopoly on service provision (Gaal and McKee, 2004; Vian et al., 2012; Kaitelidou et al., 2013; Abdallah et al., 2015; Cohen and Filc, 2017). Informal payments thus fill a gate-keeping function by deterring some patients from seeking care at all, and/or by creating multiple tiers of wait time and quality according to ability to pay (Mæstad and Mwisongo, 2011; Abdallah et al., 2015). Some (though not all) patient survey evidence suggests that health providers of a higher professional status receive higher informal payments, buttressing a theory about there being a supply and demand-driven market clearing price (Bertone and Lagarde, 2016). On the other hand, it is also possible that providers purposely create scarcity—such as by artificially inflating wait times—in order to compel patients to make payments (Mæstad and Mwisongo, 2011).

Formal user fees

In global health policy circles, prevailing opinion has mostly turned against formal user fees as an appropriate way to fund health services (Robert and Ridde, 2013). Yet, some researchers and policymakers have proposed formal user fees as a way of decreasing informal payments, suggesting that there is a direct relationship between the two, with informal payments decreasing as formal fees increase, and vice-versa (Sharma et al., 2005). The prevailing hypothesis is that, if instituted well, formal payments introduce transparency and provide needed funding for the health facility. Formal fees may also exhaust patients’ willingness to pay, making it infeasible for providers to demand informal payments (James et al., 2006). However, empirical data from different countries are mixed. On the other hand, if formal fees are not channelled appropriately, informal payments may be introduced as a remedial measure to produce needed funding. A modelling study undertaken in Bangladesh reported that it was ‘difficult to determine whether official user fees crowd in or out unofficial fees at Bangladesh health facilities’, as informal payments and formal fees seemed to accompany one another (Killingsworth et al., 1999). We propose that, since patients are often unable to distinguish between formal and informal fees, it seems likely that in some contexts, particularly those with poor governance, formal fees actually create space for the charging of informal payments. If patients knew for sure that all care was mandated to be free, they may be less willing to make payments.

Consistent with this varying relationship between formal and informal fees, data are mixed on whether or how the institutionalization or abolition of formal user fees affects the likelihood of patients making informal payments. A policy review found that efforts to replace informal payments with formal payments and allowing health facilities to keep the revenues led to improved quality of care and reduced informal payments in Cambodia and the Kyrgyz Republic, suggesting that the fees were indeed being used as intended (Akashi et al., 2004; Lewis, 2007). However, a scoping study assessing 20 studies on the abolition of formal user fees found several examples of the commencement or the continuation of informal payments following the abolition of formal user fees (Ridde and Morestin, 2011). Moreover, our ability to draw conclusions is limited by the fact that there are few longitudinal studies showing how the institutionalization of formal user fees shapes informal payment prevalence in the long run (Witter et al., 2007). At the very least, there is widespread agreement that enshrining formal payments in policy is only a piece of the puzzle; transparent guidance, mechanisms to ensure appropriate use and a host of other governance factors shape the health systems impact of formal payments (James et al., 2006; Lewis, 2007). In fact, broader health system governance may be the most determinative factor. One study found that, when accompanied by adequate drug supply and financial transfers to the facility, the exemption of certain services from user fees in Ghana led to the disappearance of informal payments for these services. This suggests that ending formal fees can lead to decreases in informal fee payments in a context of strong health system governance (Aberese-Ako et al., 2014).

Inadequate health system financing

Other research looks at salaries and funding of the health system more broadly. In many contexts, funding is inadequate for the goods and services which consequently become the subject of informal payments. This can stem from absolute resource deprivation in the health system, as well as governance and human resource constraints undercutting timely and appropriate funds and supplies transfer from the central level to lower levels of the system. As a result, health providers claim that health facilities are not adequately resourced to provide the services they are mandated to provide, so patients must contribute (Falkingham, 2004; Gaal et al., 2006a; Chuma et al., 2009; Stringhini et al., 2009; Nimpagaritse and Bertone, 2011; Stepurko et al., 2013).

Findings regarding the relationship between provider salaries and informal payments are not consistent. Focus groups conducted among providers in Tanzania found that doctors and specialists commanded higher informal payments than nurses and assistants (Stringhini et al., 2009), a finding that was confirmed in another study in Tanzania (Mæstad and Mwisongo, 2011). A regression analysis of data reported by providers and patients regarding informal payments in Tanzania found that providers earning relatively lower salaries were somewhat more likely to receive informal payments than those receiving higher salaries (with the likelihood of receiving payments being a separate question from the amount of the payment) (Lindkvist, 2014). In keeping with these findings, in-depth interviews among lay people and providers in Togo revealed much higher willingness to excuse demands for informal payments when made by providers with low salaries (Kpanake et al., 2014). Here too, there could be measurement challenges, as providers redistribute payments among themselves (Mæstad and Mwisongo, 2011). Indeed, there may be a divergence between the amount different types of providers’ request, and the amount they ultimately receive. Additionally, as suggested by some researchers, a theory about salary relevance might be advanced by acknowledging that the notion of ‘adequate salary’ and minimum standard of living are economically and socially governed, such that the relationship between provider salary and informal payments is contextually specific, and thus not comparable or meaningful across contexts (Transparency International, 2006; Stringhini et al., 2009).

In some contexts, informal user fees may comprise a significant portion of the operational funding for health facilities (Barber et al., 2004). Such fees are collected and spent at the discretion of facility management, rather than going entirely to individual providers. Facility management uses the money to fund goods and services that go directly to the patient as well as necessary supportive inputs, such as petrol (Falkingham, 2004; Diarra, 2012). Informal user fees might be considered to be a manifestation of what anthropologist Olivier de Sardan (2011) describes as ‘informal privatization’. They may be one of few means at frontline providers’ disposal to ‘make the system work’, and they may help to keep providers from leaving a poorly resourced health system to seek employment elsewhere (Gaal et al., 2006a; Olivier de Sardan, 2011; Diarra, 2012). Yet, these payments can also very easily ‘become a racket, benefiting only the providers to the detriment of the users’ (Olivier de Sardan, 2011). The boundary between necessity and racket is hard to delineate.

Structural adjustment, new public management and marketization

Structural adjustment programmes, the selective primary health care movement, the 1987 Bamako Initiative and its emphasis on cost recovery in health care, the institutionalization of so-called ‘New Public Management,’ and the associated focus on efficiency were part of a broader trend of decreasing the size of the public sector in the 1980s and 1990s (Tendler and Freedheim, 1994; Pfeiffer and Nichter, 2008; Janes and Corbett, 2009; Storeng and Béhague, 2014). The institutionalization of formal user fees and decreased state investment in the health sector were part of this trend. Health systems researchers explain that these and other changes often undercut citizen and provider trust in the system and in each other, laying the groundwork for more transactional relationships (Birungi, 1998; Gilson, 2003; Janes and Chuluundorj, 2004; Gaal et al., 2006a; Jeffery and Jeffery, 2010; Songstad et al., 2011; Spangler, 2011; Mokhtari and Ashtari, 2012; Sadruddin and Heung, 2015). For example, Birungi (1998) describes how, in Uganda, government disinvestment in health service inputs and in the health workforce pushed government health workers to adopt ‘survival strategies’, including initiating their own private sector activities and levying informal payments on patients. So, while absolute resource deprivation may be one cause, the concomitant transformation of the doctor–patient and government–citizen relationship to a provider–customer relationship may also be germane to understanding informal payments (Riewpaiboon et al., 2005; Spangler, 2011; Mokhtari and Ashtari, 2012).

Likely overlapping with the marketization dynamic, where studied, provider morale seems to relate to the likelihood that they ask patients to make payments. Studies of health providers in Ghana and Tanzania found that the providers who felt more abused by their supervisors and by the system and/or who lacked the basic inputs required to carry out their jobs were more likely to abuse patients, including pushing them to make informal payments (Tibandebage and Mackintosh, 2005; Aberese-Ako et al., 2014). Providers in Ghana explained that they were being asked to provide people-centred care while the health system employing them did not value them as professionals or people; there was a disconnect between their employment context and the performance expected of them (Aberese-Ako et al., 2014).

Paying for better care

By their own admission, many patients make informal payments in the public sector in the hopes that they will receive better care. They may be paying to ensure a continuous, interpersonal relationship with the provider; for more personalized care; for higher quality clinical care; for shorter wait times; and for more comfortable care (Vaithianathan, 2003; Gaal and McKee, 2004; Mamdani and Bangser, 2004; Riewpaiboon et al., 2005; Sepehri et al., 2005; Aarva et al., 2009; Stringhini et al., 2009; Vian et al., 2012; Lindkvist, 2013; Nekoeimoghadam et al., 2013; Stepurko et al., 2013; Karibayev et al., 2016; Baji et al., 2017; Najar et al., 2017). For example, Riewpaiboon et al. (2005) describe how Thai women select and pay an obstetrician to see them through their pregnancy and delivery, in the hopes that the money will transform an impersonal doctor–patient relationship to a trustful interpersonal relationship. The imbalance of power between patients and providers may be relevant too. Research in the USA and Hungary suggests that gifts from patients to doctors may be a way for patients to redress the status imbalance between them and their doctors by redeeming their status lost by being ill and dependent, or by imposing a non-professional identity on the physician (Drew et al., 1983; Gaal et al., 2006a).

To unpack the exercise of patient agency in the structure of governmental health systems, Gaal and McKee (2004) take economist Albert Hirschman’s theory of ‘voice and exit’ and devised the concept of ‘inxit’ to describe informal payments. In many contexts, patients lack an ‘exit’ option, as there are no alternative facilities (government or non-government) nearby. They may also have inadequate knowledge and opportunity to express ‘voice’ (e.g. dissatisfaction or demands), and they can be dissuaded from voicing their opinions by the significant social risk inherent in alienating the health providers on whom they depend. ‘Inxit’—or exercising choice within the same service by making informal payments—might be one of few means at patient disposal to influence the quality of health services, though, unlike voice, the positive results accrue only to the individual making the payment. In fact, patients making payments could result in poorer quality of care for those who do not make payments (Mæstad and Mwisongo, 2011).

But, even if patients willingly pay to obtain better care, what other options do they have? Similar to the blurry boundary between necessary financial support and a ‘racket’, the boundary between patient agency and obligation or coercion is nebulous and likely contextual. While patients with at least a minimal amount of disposable income may wish to make informal payments to reduce the opportunity cost of obtaining care (by cutting down on the time entailed) or to express their status and right to receive higher quality care, they may also feel that they have little choice, particularly in the obstetric context (Sepehri et al., 2005). If the expected quality of maternal and newborn care absent a payment is poor, or if they feel payment is required to receive any care at all, then from the patient perspective, informal payment is non-negotiable (Gaal et al., 2006a; Mæstad and Mwisongo, 2011).

Several other potential drivers and associated factors are raised in the literature, but they are mentioned rarely, making synthesis difficult. These include the relative size of the private sector, the strength and detail of law and policy relating to informal payments, norms around physicians asserting their own professional status by demanding fees, the entitlements knowledge of the patient making the payment, patient characteristics beyond income level (such as caste) and larger questions about modes of health systems financing (Stringhini et al., 2009; Mokhtari and Ashtari, 2012; Nekoeimoghadam et al., 2013; Arnold et al., 2014; Renfrew et al., 2014; Abdallah et al., 2015).

Finally, some of the drivers can also be impacts, and vice-versa. Low levels of interpersonal and institutional trust, e.g. can both drive informal payments and result from them (Gilson, 2003; Stringhini et al., 2009; Najar et al., 2017).

Impact

Informal payments can have multiple immediate and distal effects on households, communities and the health system. First, informal payments can form a significant part of a catastrophic out-of-pocket expenditure associated with an illness event, particularly in the event of labour and delivery complications (Tibandebage and Mackintosh, 2005; Jeffery and Jeffery, 2010; Perkins et al., 2009). Families may be forced to borrow money at high rates, solicit contributions from friends and family or sell productive assets (Kruk et al., 2008; Joe, 2015). The poorest are more likely to fall into this ‘poverty trap’ of debt and selling productive assets (Commission on Macroeconomics and Health, 2001; Kruk et al., 2008; Tambor et al., 2014; Joe, 2015). Moreover, as a generally flat fee levied on families regardless of their ability to pay, informal payments can be regressive, though whether or not the poorest are more or less likely to pay seems to vary among and even within countries (Killingsworth et al., 1999; Riewpaiboon et al., 2005; Kruk et al., 2008; Aarva et al., 2009; Hunt, 2010; Nekoeimoghadam et al., 2013). Two recent analyses of secondary data from many countries in sub-Saharan Africa determined that informal payments were generally concentrated among the poorest, undercutting the theory that scarcity and absolute resource deprivation in the health system are the primary drivers, and suggesting that the social status of certain patients may prevent providers from asking them to make payments (Justesen and Bjørnskov, 2014; Kankeu and Ventelou, 2016).

When patients anticipate having to pay, or have paid in the past, it can also erode trust and satisfaction with the health system. Outcomes include women bypassing facilities known to demand informal payments or avoiding facility-based delivery altogether (Birungi, 1998; McPake et al., 1999; Gilson, 2003; Mamdani and Bangser, 2004; Uslaner, 2004; Tibandebage and Mackintosh, 2005; Mrisho et al., 2007; Otis and Brett, 2008; Kruk et al., 2009; Hunt, 2010; Izugbara and Ngilangwa, 2010; Jeffery and Jeffery, 2010; Janevic et al., 2011; Mokhtari and Ashtari, 2012; Vian et al., 2012; Brody et al., 2013; Coffey, 2014; McMahon et al., 2014). The relationship between satisfaction and payments can be dynamic, with poor satisfaction both driving and resulting from informal payments (Tibandebage and Mackintosh, 2005).

It appears that many women experience requests for payments for maternity care as extremely coercive and disrespectful (Bowser and Hill, 2010; Jeffery and Jeffery, 2010; Bohren et al., 2014; 2015; Coffey, 2014; Freedman and Kruk, 2014). Egregious examples of coercion and disrespect include threatening statements such as women being told they will die if they do not pay, being asked repeatedly by different people working in the facility to make payments or risk negligence or worse, being denied pain relief during suturing unless a payment is made immediately, women being told they cannot see their newborn until they pay, and providers arguing with the family about a payment while the woman is in active labour (Afsana, 2004; Sharma et al., 2005; Ith et al., 2013; Coffey, 2014). In these cases, providers exploit women’s vulnerability and sense of urgency, leaving patients and families with little room to negotiate. Moreover, they can impinge significantly on a childbirth event, changing the dynamics to be about power and poverty, rather than welcoming a new baby. Those who are ultimately unable to pay (or suspected of such) may face ongoing disrespectful treatment, poorer quality of clinical care or outright denial of care (Izugbara and Ngilangwa, 2010; Coffey, 2014; McMahon et al., 2014).

Informal fees can also negatively affect provider morale and behaviour. Providers report that they feel forced into asking for payments as they otherwise would not have adequate salary or materials. Doing so, however, can make them feel like they are failing to fulfil their professional mandate; fearful of being caught; or, in some cases, that the balance of power has shifted in favour of the patient, who has essentially become a customer (Human Rights Watch, 2009; Nekoeimoghadam et al., 2013; Cohen and Filc, 2017; Najar et al., 2017). Moreover, the desire or the imperative to receive informal fees can drive providers and facilities to consider factors other than patient and population health in making clinical decisions. This may include pushing unnecessary clinical services that garner a higher payment, providers competing for patients who are perceived to be more lucrative; and health workers deliberately providing poor quality of care or exerting less effort until a patient pays, or, until another health worker who has already accepted an informal payment from this patient shares that payment (Gaal et al., 2006a; Vian, 2008; Stringhini et al., 2009; Mæstad and Mwisongo, 2011; Lindkvist, 2013; Cohen and Filc, 2017).

Lenses applied

The evidence and policy synthesized above were influenced by different conceptual approaches to informal payments and to health systems governance. We do not describe below well-known strengths and weaknesses of each approach, such as the cost of research, required time investment and ability to generate ‘thick descriptions’ or population-wide data. Rather, we summarize the analyses within each conceptual approach below and offer counterpoints to each. As we read the articles we noticed a few prevailing conceptual approaches, namely corruption, econometrics and qualitative research. We were able to class all of the empirical articles we read into these broad categories. That being said, a minority of articles have elements of more than one approach. The summaries of each approach below describe various strands and tensions within each conceptual approach, illustrate how different studies and policy documents may reflect particular conceptual approaches, and help us to critically assess the potential advantages and disadvantages of each approach. This lays the groundwork for subsequent discussion on addressing informal payments.

Informal payments as a form of corruption

Currently, one of the most widely used definitions of corruption is the ‘misuse of entrusted power for private gain’ (Mackey and Liang, 2012). Informal payment requests are frequently described as a type of corruption in the health sector (Lewis, 2007; Vian, 2008; Mackey and Liang, 2012). Researchers employing a traditional corruption lens based in classic economic theory posit that corruption stems from monopoly, discretion and lack of accountability (Klitgaard, 1988; Gebel, 2012). Service providers with a monopoly (in this case, the public sector) face little competition. Facing little to no credible threat of sanction for demanding payments (discretion), these providers make a choice to misuse their power for private gain (Mackey and Liang, 2012). The assumption is that the incentive structure in the health system does not prevent corruption (Bukovansky, 2006; Gebel, 2012). As explained by Lewis (2000), ‘informal payments…provide a means by which corrupt public servants can ensure or maximize their income, evade taxes, and effectively “beat the system” and consequently are a form of systemic corruption’.

However, empirical evidence from several countries suggests that this classic corruption paradigm does not describe all instances of informal payments, and that the blanket deployment of corruption discourse can risk undermining research and action. Genuine gift-giving and informal payments that are considered absolutely necessary to keep the facility operating or to deliver a service, such as when providers ask a patient to purchase drugs that are part of the entitlement but absent at site level, can hardly be described as corrupt. There is no private gain in these instances. Also important to consider is the much larger grey area of payments that patients or providers consider to be necessary but others judge to be unnecessary, including those with some gratitude component. Too, patients may wish to make payments in order to reduce wait times and assert their status as being above the most poor. Finally, some argue that corruption flows partly from marketization, and that the concepts of monopoly, discretion and accountability are insufficient to understand corruption; poor morale, insufficient funding and acceptance of health care as a transaction engender corrupt practices (Gebel, 2012).

The classic corruption label may seem inappropriate to some providers and patients. The moral culpability and illegality it implies may be overly harsh, particularly in a context where informal payments are pervasive and considered to be legitimate (Vian et al., 2015). For these reasons, some researchers advocate understanding corruption as a collective action problem; the individuals most engaged in delivering care at the frontlines may be the least able to effect change (Burns et al., 2013; Persson et al., 2013). Corruption continues unabated because individuals face strong pressures to continue; patients seek to obtain better care and providers face professional pressure to demand and share payments, just as their peers do. The opportunity cost for an individual being non-corrupt is quite high, unless everyone else becomes non-corrupt too (Persson et al., 2013).

However, our reading in other disciplines suggests that this collective action approach cannot explain the entire ecology of informal payments. For example, it fails to consider the social norms implicit in the interactions between patients and providers, including gift giving, as well as all of the health system challenges. For example, even if everyone at a particular health facility were to spontaneously agree to stop demanding informal payments, this does not mean that drugs would immediately become available (Menochal et al., 2015). Moreover, many providers and patients may prefer that the system continue as is, so they do not think there is a collective action problem. Patients with more resources may prefer a two-tier system of quality that benefits those who can pay, and some providers may prefer a system that benefits them directly (Vian et al., 2012; Walton and Jones, 2017).

Walton (2015) describes an ‘institutional decay’ understanding of corruption. We propose some researchers might consider designing studies that allow such emic understandings to emerge, rather than imposing a priori assumptions about corruption. These emic perspectives might yield more apt policy responses. The decay hypothesis is consistent with the proposition that we should focus on the ‘system in which professionals are working, rather than the persons themselves’ when it comes to understanding corruption (Ferrinho et al., 2004). Further, it improves upon ‘thin conceptions of institutions as incentive structures’ to look at political and normative underpinnings (Brown and Cloke, 2004; Bukovansky, 2006). Walton (2015) developed hypothetical scenarios reflecting different understandings of corruption and found that those matching the ‘institutional decay’ approach resonated strongly with survey respondents in Papua New Guinea; they considered the notion of moral atrophy of institutions to be especially harmful, and stated that it aptly described their experiences with the state. Though the low level of state penetration in Papua New Guinea is unusual, the broad notion of institutional decay is resonant with our synthesis of the systems drivers of informal payments.

The institutional decay understanding goes beyond a decoding of individual motivation and incentives to assess historical, social and institutional norms and modes of operating. These modes can be described in different ways. From the provider perspective, Olivier de Sardan (2008) describes practical norms, which are contrary to official norms, but widespread and embedded in civil servant practice. These norms are generally implicit and consistent over time (Olivier de Sardan, 2008; Olivier de Sardan and Ridde, 2015). Such norms may be particularly acceptable to patients who understand (under-resourced or poorly remunerated) providers’ plight (Kpanake et al., 2014).

Research assessing the link between trust and corruption has found that higher levels of trust in public institutions—as opposed to just interpersonal trust—are associated with decreased corruption (Sööt and Rootalu, 2012). This suggests that whole systems concepts such as institutional decay are ripe for exploration; it seems that what people think about the government or ‘the system’ relates to their experiences of corruption. Focusing solely on regulating individual encounters might be less effective absent efforts to reform the institution. Such approaches are not yet widespread in the literature, though it is increasingly visible. For example, Vian (2008), a long-time researcher on corruption in the health sector, proposes that Olivier de Sardan’s notion of practical norms is an important area for future study. Similarly, an analysis of the discourse within Transparency International (TI)—a global leader on anti-corruption discourse—stated that TI increasingly acknowledges the relevance of a more holistic, ethics-based approach, but that this approach is far from entrenched in their practice (Gebel, 2012).

Ethnographic and in-depth qualitative research

Ethnographic research has shed light on the patient and provider experience of informal payments, the local institutional context, and the wider social and political structures that influence the local institutional context. Moreover, given some of the measurement challenges described earlier, observational and in-depth interview techniques are particularly suited to drawing out the implicit, hidden nature of informal payments. Anthropology ‘has a long and rich tradition for studying hidden practices and illegal or semi-legal exchanges’ (Nuijten and Anders, 2007, p. 4).

Using surveys to understand individual motivations related to informal payments may be particularly ineffective in contexts where respondents associate lists of closed-ended questions with governmental data collection, and thus fail to provide honest responses (Sessener, 2001). In contrast, one-on-one in-depth interviews and observations may allow researchers to ascertain what informal payments mean from the actors’ own point of view (Sessener, 2001), and how informal payments are related to a ‘configuration of broader practices’ that illuminate the relations between patients and the health sector and relationships within the health sector itself (Blundo and Olivier de Sardan, 2006, p. 87). While ethnographic approaches do not communicate the scale of informal payments, understanding the meanings attached to informal payments is essential to establishing if they do in fact have an impact, i.e. on balance negative in a given context, and if so, how they might be changed. For example, Spangler (2011) recounts the statement of a Tanzanian woman: ‘You don’t have the power to refuse. What will happen when your child gets malaria? Or the next time you go to deliver. No, No. This you cannot refuse’. The inability to refuse and fear of future contact with the health system may not be easily discerned in a survey, yet these factors are essential to understanding the larger impact on trust and citizenship informal payments can have. Similarly, learning through a health system-based ethnography that informal payments are shared among several providers or that informal payments lower provider morale may be key to ascertaining how informal payments may be disrupted (Pfeiffer and Nichter, 2008; Aberese-Ako et al., 2014; Hoag and Hull, 2017). The fact that providers may feel their professional role is compromised by these payments is an important ‘hook’ for efforts to reduce informal payments.

Micro-economic

Economists have used a willingness-to-pay framework or econometric modelling to understand some of the immediate causes and impacts of informal payments. The concept of scarcity and much of the theory and data on the relationship of formal user fees and provider salaries to informal payments come from this tradition (Baji et al., 2012).

To unearth the prevalence, drivers and impact of fees, researchers have conducted original surveys, analysed large household survey data sets and proposed econometric models of factors associated with informal payments. Though this approach can make a focus on systems and complexity difficult, and are limited by measurement challenges, these studies provide the most complete data on the frequency and geographic scope of informal payments, including on the higher rates of payment by obstetric patients (Mokhtari and Ashtari, 2012). Moreover, they have played a key role in elucidating economic impacts at the household and facility levels, such as when and where informal payments are regressive, how informal payments and other out-of-pocket payments can have a catastrophic effect on household economic stability, and how the existence of payments creates a two-tiered (or even a multi-tiered) system of quality (Killingsworth et al., 1999; Hunt, 2010; Abdallah et al., 2015; Joe, 2015).

Synthesizing data from these and other paradigms illustrates the value of an inter-disciplinary approach. Each lens has particular added value and weaknesses. These attributes in turn affect the solutions proposed.

What to do?

Proposed ways of reducing the harm of informal payments have run the gamut from general civil service reform to narrow efforts to change the ‘incentives’ providers and patients face to health system improvements, such as reducing stock outs. A discussion of all of these proposed interventions is beyond the scope of this article, but a brief summary elucidates how theoretical orientation, policy pragmatism and expediency shape some of the solutions proposed.

Global health experts and economists often suggest addressing the putative proximate determinants of provider incentives, such as raising their salaries, allowing private sector moonlighting, instituting formal fees and stronger sanctions for demands for informal payments, and stating provision of bonuses based on the number of patients served (Gaal and McKee, 2004; Lewis, 2007). There are multiple examples of these policies being implemented in isolation, or without attention to professional norms and larger issues such as trust (Cherecheş et al., 2011; Lê, 2013; Stepurko et al., 2013). There are also examples of reforms being implemented partially, such that achieving the intended impact is unlikely (Witter et al., 2007; Aberese-Ako et al. 2014). Even if new policies are implemented with full fidelity, they may be unable to change the broader dynamics, such that informal payments persist (Lewis, 2007). Some suggest ‘working with the grain’ by acknowledging practical norms and attempting to shift them; a group of anthropologists working in West Africa described a successful effort led by a ‘reformer midwife’ to cut the average daily health care worker income from informal fees in half (Olivier de Sardan et al., 2017).

Many researchers and advocates explore citizen ability to contest informal payments. They recommend individual and collective expression of voice, or dedicated monitoring efforts as a means to claim rights. The implicit assumption is that increased citizen and policymaker knowledge, collective action and opportunities to dialogue with local providers and officials will lead to greater accountability regarding informal payments (George et al., 2005; Vian, 2008; Vian et al., 2012; Schatz, 2013; Pieterse and Lodge, 2015; Molina et al., 2016). There are some examples of local level health facility responsiveness, including regarding informal payments, stemming from community monitoring and other social accountability efforts (Dasgupta et al., 2015; Molina et al., 2016).

However, there are several caveats in the empirical literature. First, individuals and communities need to be aware of their rights in order to claim them (Mamdani and Bangser, 2004; Chuma et al., 2009; Dasgupta, 2011; Spangler, 2011; Mokhtari and Ashtari, 2012). They also need to feel safe claiming them; as described, the poor, particularly women, may be reluctant to alienate providers at the only health facility in their area, particularly because they and their families will rely on them in the future (Spangler, 2011; George and Branchini, 2017).

Second, to address some of the institutional decay at issue, frontline monitoring efforts might need to move beyond the most sensational examples of health provider abuse to challenge the underlying system wide failures (George et al., 2005). This is harder to do with scattered efforts at community monitoring. An integrated, scaled up accountability effort that addresses multiple levels and agencies of the government, communities and institutional capacity may be needed (Fox, 2015; Halloran, 2015). Building alliances between providers and community members on shared priorities—such as lack of adequate drugs and supplies—hold potential as part of a larger strategic approach (Fox, 2015).

Others have proposed ways of addressing broader factors that can both be drivers and impacts of informal payments, such as levels of institutional and interpersonal trust. This might be accomplished through enhanced quality accreditation; changing the cost and reimbursement structure in hospitals; greater engagement of professional associations and training bodies; and greater attention to health system governance (Riewpaiboon et al., 2005; Piroozi et al., 2017).

Conclusion

CIS entails moving beyond aggregation and breaking new ground in synthesis. We have accomplished this by interrogating a heterogeneous literature, in a way that has not yet been done in discussions of informal payments. The synthesis of lenses that have been used to study informal payments further shows how these paradigms inform empirical work on fee prevalence, and we suggest ways in which approaches from outside the traditional global health literature can productively be applied to unpacking and addressing informal payments.

Informal payments are a manifestation of health system dysfunction. Their most negative effects are on those who are the most disempowered in under-resourced and poorly governed health systems—frontline providers and their patients. Recent conceptual work asserts that disrespect and abuse in maternity care should be defined by both patients’ subjective experiences and provider intent (Freedman et al., 2014). This approach can be applied to informal payments. As learning from several disciplines shows, the harm in informal payments is located in subjective patient experience of coercion, disrespect, fear or economic damage as well as provider intent to take advantage of patients and provider sentiments that the health system does not give them the resources required to realize their professional mandate.

Payments may allow some patients with adequate capital to bypass the most egregious manifestations of health system dysfunction, but they do nothing to mitigate that dysfunction. In fact, informal payments may feed dysfunction by perpetrating clientelism and corruption in the allocation of postings to health care workers (Schaaf and Freedman, 2015). Thus, harm goes beyond the individual. Informal payments can undercut trust beyond those people implicated in any given encounter, and contribute to health services being provided and received as a commodity, rather than an entitlement. This has implications for community willingness and capacity to access services, the quality of communication between patients and providers within the service, and community trust in the government.

Our chosen definition of informal payments includes all payments that are beyond entitlement; some of these payments may neither hurt patients nor stem from provider avarice. While this definition is conceptually clear-cut, it is empirically difficult to assess, challenging research and policy related to informal payments. There are advantages and disadvantages to the various approaches in which any definition is embedded. Like many health systems issues, it appears that different lenses each tell only part of the story. The appropriateness of an approach depends partly on contextual factors and the questions we seek to answer. For example, analysis of provider incentives might be more apt in settings where corruption is not endemic. The ultimate objective of any research should be to tell as much of the story about the practice and its meanings as possible, without getting lost in a hall of postmodern mirrors that offers few possible solutions.

The everyday relevance of informal payments to both maternity care providers and patients globally is not reflected in the research base, which privileges EE, fSU and analysis of proximate and individualistic determinants. Microeconomic analyses of these proximate determinants may lead to overly narrow solutions, but even here, we have little long-term data or fully implemented programmes on which to judge the efficacy of solutions. In any case, the demonstrable importance of trust, provider morale, institutional determinants of corruption and the social construction of rights revealed in qualitative analyses suggest that a multidisciplinary health systems approach that leverages and integrates positivist, interpretivist and constructivist lenses of social science research can lead to better insight and policy critiques. Among other questions, the boundary between informal payments as palliative mechanism and exploitation, the power and equity determinants and outcomes of payments, and the interplay between local and global (translocal) constructions of corruption and informal payments merit exploration. With this, we can challenge inadequate ‘master narratives’ and strive to meet universalistic, equity-oriented global health objectives.

Funding

Lynn Freedman provided critical intellectual input and thought partnership, and Lindsay Stark and Jonathan Fox provided helpful comments. Amy Manning and Francesca Heinz provided invaluable editorial and research assistance. Financial support for the conduct of the research and preparation of the article was provided to the Averting Maternal Death and Disability Program by the John D. and Catherine T. MacArthur Foundation. The Foundation played no role in study design; data collection, analysis or interpretation; or the writing of the article.

Conflict of interest statement. None declared.