Policy-makers and health finance planners need to consider the potential of fee-for-service family planning programs as a strategy for mobilizing revenue. While raising revenue is but one of many reasons for charging fees for services (see Figure 3.1), it is emphasized here due to our overall focus on potential alternatives to government and donor financing.
|
Figure 3.1. Rationale for Fee-for-service Family
Planning Programs
Revenues and Quality of Services
Increased revenue from user fees can:
- augment financing of recurrent inputs, such as contraceptives, and improve the quality and effectiveness of services;
- free up resources to expand the availability of services; and
- lessen dependence on donor funds.
Efficiency
Strategic pricing of services:
- can reduce excessive use of services and bring supply capacity in line with willingness and ability to pay; and
- can direct clients to lower-cost sources (for example, to pharmacies/shops for re-supply methods) while higher-level facilities provide clinical methods and methods to new acceptors.
Equity
User fees can improve equity:
- if higher prices are charged to those most able to pay, making it possible to channel family planning subsidies to the poor; and
- if exemptions are implemented for the most destitute.
Public-Private Collaboration
User fees for government services:
- foster greater competition between private and public providers and improve the efficiency of both; and
- divert demand to private providers, freeing up government resources to improve care for the poor.
|
The concern with charging or increasing fees for family planning services is that it will discourage service use. This chapter, therefore, addresses both the revenue-raising potential of fees and the conflict between raising revenue and ensuring access to services. Because information on the impact of fees on the use of family planning services is limited, we also refer to the literature on the health sector as a whole.
How Do Fees Affect the Demand for Family Planning Services?
An economic model of contraception and fertility
The economic model of household production dates back to Becker (1981) and has been applied to contraception and fertility (Rosenzweig and Schultz, 1985). According to the model, if the costs of children outweigh their benefits, then couples may choose to contracept to limit the number of children they bear. Similarly, if couples believe that spacing their children will bring them health or other benefits, they may contracept to achieve these benefits. The decision to use contraceptives will depend on the perceived costs of contraceptives relative to the benefits of using contraceptives. Costs will depend on the efficacy of the contraceptive, the cost of obtaining it (its price, as well as the costs of travel time, waiting time, and transport) and the cost of using it (including side effects, cultural or spousal sanctions, etc.) (Thomas and Maluccio, 1995).
Early studies of the impact of prices on contraceptive use had mixed results. Some studies suggested that prices have a minimal impact on demand; others suggested a dramatic impact. This variability is likely to reflect measurement differences as much as true differences in the price elasticity of demand in different settings (see Lewis, 1986; Janowitz and Gould, 1993, for a review of early studies).
In this chapter, we provide a review of more recent studies of the impact of prices on contraceptive use in sub-Saharan Africa. We first provide a framework to classify the studies by the type of methodology used. In this way, the methodological strengths and weaknesses of each study can be kept in mind when its findings are assessed.
A framework to examine the impact of fees on demand for services
Shaw (1995) categorized studies of health care demand into five types, as follows:
Type I studies use facility-based data to compare service use before and after fees are instituted without adjusting for client characteristics or the quality of care, and without using a control group. Studies of this type are likely to measure incorrectly the effect of prices on demand because they neglect the impact of quality improvements made possible by increased revenues (consumers may be willing to pay more for services if fee revenues are used to improve service quality). There also is no consideration of the possibility that consumers will simply find alternative sources of care when confronted with a price increase; instead, it is assumed that declines in service use at any given facility represent declines in overall use.
Type II and Type III studies are based on multivariate statistical analyses of cross-sectional household data. These studies control for some client characteristics and other variables that affect use. They are generally computer simulations of what individuals would do, rather than studies of what individuals actually do, in response to a price change. Type II studies use non-monetary costs to the client (such as distance and time costs) as proxies for prices, either because services are provided free of charge or because information on prices is not available. Shaw (1995) notes that such variables may not adequately reflect how price changes affect demand for services because: (a) service fees tend to represent a far smaller proportion of income than these other costs; and (b) the opportunity cost of time may be so low as to obviate the impact of time costs on service use. Type III studies, on the other hand, are based on actual price data. These studies separate the effects of the price of services per se from the effects of distance, time and other costs to the client.
Type IV studies also use cross-sectional data, but they improve on study types II and III by building in data on facility characteristics, thereby allowing consideration of the impact of quality on demand. Because these studies are cross-sectional, they do not actually measure behavior change. Nonetheless, they generate an important policy consideration: The impact of fees on demand may hinge on whether or not revenues are used to improve the quality of care.
Type V studies are the gold standard for studying the effects of fees on demand. They are based on actual behavior change in response to the institution or an increase in fees for services, as well as quality enhancements, and they include a control group. Although very valuable, these studies are time-consuming and expensive.
Family planning demand studies in sub-Saharan Africa
A series of studies has been conducted in Africa to assess the impact of costs to clients (fees, travel, and time) and quality on contraceptive use. The results of these studies vis-à-vis the impact on demand of the fee or price component of costs (or a proxy thereof), as well as the impact of quality, are described below.
In Tanzania (Beegle, 1995), no information was available on prices, but information on facility characteristics was available from a Service Availability Module (SAM). Cross-sectional household data were available from the DHS. This study, therefore, is a combination of types II (no information on prices) and IV (information on quality). The price of contraception was proxied using information on distance to the nearest facility. As noted earlier, when no information on price is available -- or when the price is zero -- distance is often used as a proxy for price, based on the assumption that time and travel costs affect service use in the same manner as monetary prices. The author notes that this proxy has some drawbacks, since visits to the nearest family planning facility may be "shared" by other services (i.e., individuals may obtain a range of services in addition to family planning when they visit the facility). The study found no consistent relationship between price and demand.
Only two of a number of quality variables used -- availability of pills and injectables -- were found to have a positive impact on use.
A study carried out in Zimbabwe (Thomas and Maluccio, 1995) may be classified as type IV. Information on household variables was available from the DHS and quality measures were available from the Zimbabwe Situation Analysis. The authors used the price of a package of oral contraceptives as the price variable. Partly because there is no heterogeneity of prices in Zimbabwe (all clinics charge Z$0.20 per package), the study found no impact of price on demand. In addition, experiments using different specifications for distance to facilities (which captures part of the time cost) indicate that it does not have an impact on use. However, women in areas with a community-based distribution (CBD) program were significantly more likely to be using contraceptives, and this variable may, in part, be capturing the cost of time.
Only two of the quality variables for clinics had a significant impact on use (the number of needles in stock and the number of nurses on site). Quality of CBD workers, as measured by whether or not they had a bicycle or had taken a training course, also had an impact on contraceptive use.
In Ghana (Oliver, 1995), information was available from the Ghana Living Standards Measurement Survey as well as from a survey of the health and family planning facilities nearest to each group of households. This study is therefore classified as type IV. The family planning consultation fee and the price of spermicides were used to construct price variables. The effect of distance to the nearest health facility offering family planning was also assessed. The price of spermicides was found to have a negative impact on use, but only in private facilities. The consultation fee had no impact on use. The author concluded that prices have very little impact on the probability of contraceptive use. Distance to facilities proved to be a significant constraint to use.
In this study, quality was represented by variables measuring the availability of other maternal and child health (MCH) services, the availability of several methods, and the number of staff available to provide family planning services. The coefficients on the measures "are inconsistent and sometimes perverse...." The author suggests that this result may reflect the lack of variability in quality levels, the inability of the variables to measure appropriately important aspects of quality, or the fact that at low levels of contraceptive prevalence, quality is not an important determinant of use.
The study in Nigeria (Feyisetan and Ainsworth, 1996) may also be classified as type IV since information was available on both the price and quality of services. Data were from a DHS-type survey and a SAM-type facility survey. No relationship was found between contraceptive use and the level of outpatient or registration fees. The price of methods (at pharmacies only) was associated with lower use. Distance to the nearest facility offering family planning services did not affect use.
The quality variables appeared to have limited impact. The authors suggest that the facility survey was unable to capture all aspects of service quality, though a recent Population Council situation analysis (Mensch et al., 1994) also found no statistically significant effect of quality on contraceptive use in Nigeria.
In sum, these four studies suggest that prices have a relatively limited impact on demand for family planning services. In addition, quality appears to have a less significant impact on demand than one might expect.
Lessons from the health-care literature
There is a growing literature from the health sector on the impact of fees on demand for care. Unfortunately, even in this more developed body of work, much of the evidence on price elasticity of demand is "mixed, piecemeal, and of questionable validity" (Shaw, 1995). Relevant findings are discussed below.
1. Individuals may react to fees by finding new sources of care. In Swaziland, Yoder (1989) found that when prices at Ministry of Health facilities were raised by 300 percent to 400 percent, there was a 32 percent drop in use; however, 30 percent to 40 percent of prior Ministry of Health clients switched to the private sector.
2. Quality is more important than price. The results of health-care demand studies support the hypothesis that price plays a limited role in determining use of health care. However, quality may have a stronger impact on health-care use than it does on family planning use. For example, Lavy and Germain (1996) found that fee increases contributed little to the dramatic decline in health-facility use in Ghana between 1973 and 1987, particularly in comparison to the effects of distance and service quality. This type IV study used a nationally representative survey of the population and a comprehensive survey of facility characteristics. A study in Nigeria (World Bank, 1991) also found that the impact of quality on use was far greater than that of price, and that quality improvements could offset the impact of higher prices.
The role of quality in determining the demand for health care is confirmed by the results of type V studies. Litvack and Bodart (1993) conducted research based on a natural experiment in Cameroon, through which three of five facilities introduced a user fee and quality improvements, and two introduced fees but did not make improvements. They found that while fees do have some effect on demand, this effect is limited when quality improvements are made simultaneously.1
A similar study conducted under the Health Financing and Sustainability (HFS) project in Niger (Ellis and Chawla, 1994 in Wouters, 1994) assessed the impact of two types of payment systems, introduced in conjunction with quality improvements (improved drug supply), on revenues and demand. The positive effects of quality improvements again offset the demand-dampening effects of price.
3. Improving quality can be costly. Wouters et al. (1993) caution that the cost of implementing necessary quality improvements could outstrip potential fee revenues, given the lack of a service base in many settings. This is particularly true in sub-Saharan Africa. For example, the value of drugs that facilities consumed in one Niger district undergoing quality improvements was 2.5 times higher than the annual Ministry of Health drug budget. The authors suggest that even if the primary purpose of fees is to raise revenue to improve services -- or if quality improvements are necessary to prevent demand from falling in response to prices -- initial improvements will require financing beyond fee revenues. If, on the other hand, the purpose of fees is to cover existing costs, making quality improvements may make it impossible to achieve this objective.
4. Local fee retention can increase revenue collection. Governments have tended to require that facilities return any fees they collect to the central government for distribution in the manner most likely to enhance the public good. There are, however, benefits to local retention and use of fee revenues, including a greater incentive to collect fees and the potential improvement of service quality on site. In an analysis of Cameroon, Central African Republic, and Swaziland (McInnes, 1993), facilities that retained revenues performed substantially better than facilities that remitted revenues to the treasury. In Cameroon, facilities that retained revenues were able to cover drug costs, as well as increase use of health centers, with the poor benefiting more than the rich. In Swaziland, one of the reasons for the low percentage of costs recovered by public facilities was lack of incentive to collect fees. By comparison, Osuga and Nordberg (1993) found that the introduction of user charges for outpatient services at Ministry of Health facilities in Kenya resulted in a 20 percent to 40 percent decrease in overall health-care use after six months. This occurred despite the fact that 75 percent of the revenues collected were to be retained by facilities -- presumably to finance service improvements. However, it is not clear that fee revenues were used to improve quality.
5. Means testing cannot guarantee that the poor will be protected. Even if most people are able and willing to pay for services, there is evidence that some people cannot. For example, based on a review of the literature and research in Côte d'Ivoire and Peru, Gertler and van der Gaag (1990) conclude that user fees may cause large reductions in service utilization by individuals in lower income groups. Similarly, Leighton (1991) found that in Senegal, the ultra-poor were far more likely than the moderately poor to reduce their use of primary health care in response to rising prices.
Some form of means testing is needed to identify those who cannot pay and ensure that they have access to care. This can be complicated; in some cases, the expense of designing a system that functions well may negate any benefits that could be derived from imposing fees. The results of recent reviews of means testing in sub-Saharan Africa are summarized in Box 3.1. None of the literature on means testing is specific to family planning. Nonetheless, a general lesson can be learned: If fees for family planning are instituted or increased, effective mechanisms to ensure that the poor will continue to have access to services cannot be taken for granted.
|
Box 3.1. Means Testing in Sub-Saharan Africa
An extensive review of means testing in developing countries (Levine et al., 1992) found that "success measures are neither widely available nor uniformly applied." One assessment based on staff opinions found that only ten of 28 programs were successful. Only one of these was in sub-Saharan Africa. The conditions required for successful means testing, such as formal wage records, adequate information, and administrative infrastructure, tend not to exist in the region (Willis and Leighton, 1995).
Means testing systems in Africa tend to be decentralized and informal. Reviews have found that only two countries (Zimbabwe and Ethiopia) have official income ceilings for exemptions: one (Lesotho) has specific landholding and livestock ownership criteria; one (Malawi) has landholding criteria only; 12 claim to provide exemptions but do not specify the criteria used; and eight are based on ad hoc local policy (Nolan and Turbat, 1993; Russell and Gilson, 1995 in Leighton and Diop, 1995). On closer examination, many of these criteria appear to be ad hoc, vague, or inequitable. The programs have considerable "leakage" of exemptions to the non-poor, either intentionally or otherwise, and their coverage of the poor is inadequate.
In Lesotho, for example, while 30,000 health workers and their young children are exempt, the poor must be officially certified as "paupers" to receive an exemption; only 200 paupers are certified in the country as a whole (Nolan and Turbat, 1993). In Senegal, local health committees can exempt certain individuals, including "certain kinds of poor people" and various government workers (Vogel, 1988 in Levine et al., 1992). Malawi, on the other hand, has developed an exemption system through which the "core poor," a category being defined through a detailed examination of the landholding structure, receive exemptions; about 19 percent of the population will qualify (Ferster et al., 1991 in Shaw, 1995 ). In Ethiopia, peasant associations are authorized to determine eligibility based on income, and little leakage occurs (World Bank, 1987 in Levine et al., 1992).
In an extensive review of informal means testing systems in Burkina Faso, Niger, and Senegal, Leighton and Diop (1995) found that the proportion of poor clients receiving exemptions is higher than the proportion of non-poor clients receiving exemptions. This suggests that there is some attempt to identify and give waivers to the poor. Public facilities tended to provide about 25 percent of all exemptions to the poor, and 75 percent to the non-poor (in Senegal, the commensurate figures are 12 percent and 88 percent, respectively). |
Key Issues in Family Planning Pricing
Based on the lessons learned from the health-care finance and family planning literature, a number of issues must be addressed in considering the potential for fees as a means to raise revenue for family planning.
1. To what extent are findings from the health-care literature applicable to family planning? It is important to note that consumers may not react the same way to family planning price increases as they do to increases in prices for curative health care. They may be more ready to accept increases in the prices of curative health services without diminishing use if they see immediate benefits and if they are the direct and sole beneficiaries of those benefits. Consumer reaction to price increases for family planning is likely to be more like that for preventive health care (Desai, 1997). In the case of family planning, however, it is not the probability of a future illness that is being reduced, but that of a birth. The consequences of not taking action to prevent a birth are very different than those of not taking action to prevent future illness; as such, even reactions to price increases for preventive care may be very different than for family planning.
2. Will large increases in fees dissuade use? While the results reported above generally indicate that price does not have an important impact on demand for family planning services, this may be because contraceptive prices are very low. In Ghana (Oliver, 1995), the higher private sector price was found to have a significant impact on demand. A similar finding was reported for Nigerian pharmacies, where prices are likely to be higher than in public and NGO facilities. These findings also show why price is rarely cited in DHS surveys as a reason for not using or not intending to use contraception (Westoff and Bankole, 1995). If prices are low or nonexistent, then affordability would not be an important deterrent to use. We may conclude that a modest increase in fees from a low level is unlikely to dissuade use; however, large price increases to near-commercial levels might have a more significant impact on demand.2
3. Do family planning clients respond to price increases by seeking alternative sources of services? Shifts of demand to other sources of care may occur when fees are charged for family planning services. Ciszewski and Harvey (1995) point out that an increase in the prices of pills and condoms provided through the social marketing company in Bangladesh led to decreases in sales; however, while the company's market share decreased overall, contraceptive use rose (Janowitz and Bratt, 1996). The original study had failed to account for source substitution. Most studies of the impact of contraceptive price increases fail to take this phenomenon into account.
4. Will the impact of prices be different for long-acting and re-supply methods? Most of the family planning pricing studies cited above used information on the actual price paid for services. However, earlier work conducted in other regions (for example, Thailand, Indonesia, and Jamaica as reported in Akin and Schwartz (1988), Ashakul (1990), and Jensen et al. (1993)) used a price variable that averaged protection over the period of use. While an individual generally pays for a method at the time of service, some methods last a long time or are permanent. How do women view these prices? Is their reaction to price conditioned by the fact that they may pay more now (for longer-term methods) but will avoid paying later? Will the high up-front costs of long-term methods be a barrier to cost recovery? More research should be conducted to answer these questions.
5. Do women react differently to price changes than men? It is argued that women may be more affected by fees than men, given unequal gender relations and women's lesser say in decisions regarding household resource allocation. However, little research has been conducted on this issue. Some interesting findings have emerged from the Population Council's project in Navrongo, Ghana, but these have not yet been published (Phillips, 1997, personal communication). Phillips does suggest that women's fear of negotiation with their partners may deter them from requesting funds to purchase contraceptives. There is some evidence from the health sector on this issue. For example, a study in Senegal (Leighton, 1991) found that changes in primary health-care utilization in response to rising prices differed by gender, with women more affected than men.
6. Will young adults react differently to price changes than older men or women? There is very little data on the differential impact of fees on different age groups. A recent publication of the FOCUS Project, which reviewed social marketing projects directed at young people, noted that price was a potential barrier to service use among this group. However, none of the projects or research reviewed examined this issue (Israel and Nagano, 1997). Another recent study found that sexually active adolescents in Cameroon were more likely to use condoms they had paid for than those they had received free (Meekers, 1997). However, there was no attempt to determine the role prices play in influencing demand for condoms. If prices have a strong deterrent effect on adolescents' willingness to purchase condoms, free distribution may be preferable even if these condoms are more likely to be wasted.
7. Will quality improvements mitigate the impact of price increases? The family planning literature is only now beginning to consider the impact of quality on demand and the potential interaction of quality with price in determining demand. If fee revenues are used to improve service quality, then an increase in fees may have a less significant impact on demand. The studies cited above, however, found that quality had a limited and inconsistent impact on use in the case of family planning services. It may be that the quality variables used in these studies were inadequate.3
An alternative interpretation is that quality plays a different role in determining demand for family planning than it does in determining demand for health care. Results from the literature on willingness to pay do support the lack of a strong quality effect in family planning services. In Burkina Faso, for example, Sow (1994) found that while nearly all households were willing to pay for improvements in health-facility equipment and maintenance and for drugs to treat a range of common ailments, far fewer were willing to pay for contraceptives. Households were willing to pay amounts representing 5 percent to10 percent of their total expenditures to improve the quality of health care, but less than .05 percent for improvements in the quality of family planning services.4
8. Are time costs a barrier to family planning use and can proxies, such as time, measure the impact of price changes? The effects of distance and travel time on the use of family planning services are often used as proxies for the impact of prices. As such, it is worth examining how accurate these proxies are. While distance and travel time have been found to have an important impact on the use of curative health care, they do not have a consistent impact on the use of family planning services. This is because distance and travel time have different implications in the case of family planning. For example, the distance and travel time costs associated with obtaining contraceptives may be shared if family planning services are located near other amenities used by the individual (e.g., the market) as was noted by Beegle (1995). In addition, the impact of distance is likely to depend on the type of method a woman chooses. Women who choose long-acting or permanent methods may only need to travel once, whereas those who choose re-supply methods may have to travel often to continue method use. Finally, individuals other than the user (e.g., the husband) can purchase some methods. Because of these issues, inferences about the impact of "price" changes on demand and revenues based on distance and travel time proxies must be made with caution.
9. Can means testing ensure equity and access? There is evidence that the poor -- and particularly the very poor -- will be more affected by price increases than the wealthier. However, the health-care literature appears to indicate that successful means testing -- which should ensure that those who are entitled to exemptions receive them and that those who are not, do not -- can only be ensured with great difficulty and at very high cost. It is possible that efforts to ensure equity will be so costly that they will obviate potential fee revenues altogether.
10. How much revenue can fees be expected to generate? Recent data suggest that private financing, mostly by individuals, comprises about 43 percent of total health expenditure in sub-Saharan Africa -- a far more significant proportion than is generally assumed (see Table 3.1, Hanson and Berman, 1994). However, for the reasons discussed above, potential family planning fee revenues are likely to be less substantial. Family planning programs need to consider this possibility carefully in determining how much they can rely on fees as a revenue source. This is particularly true if quality enhancements -- which will require a significant proportion of revenue -- are key to ensuring that prices do not deter use. If the objective of fees is to improve program sustainability by finding sources of domestic revenue to substitute for donor funding, quality enhancements may not be possible or may need to be limited in scope.
Table 3.1. Private Health Expenditure as a Percentage of Total Health Expenditure, 1990
|
Country |
Percentage |
| Benin |
22 |
| Burkina Faso |
15 |
| Burundi |
25 |
| Cameroon |
43 |
| Chad |
16 |
| Cote d'Ivoire |
27 |
| Ethiopia * (1987) |
64 |
| Ghana * |
56 |
| Kenya |
38 |
| Madagascar |
50 |
| Mali * |
61 |
| Mozambique |
25 |
| Nigeria |
45 |
| Rwanda |
31 |
| Senegal * (1989) |
51 |
| South Africa * |
48 |
| Tanzania * (1991) |
62 |
| Uganda |
53 |
| Zaire * |
80 |
| Sub-Saharan Africa |
43 |
* The author replaced World Bank Development Report estimates with Data for Decision Making estimates, for reasons of accuracy. All figures are for 1990, unless noted otherwise.
Source: Derived by Hanson and Berman, 1994,
from estimates prepared for the World Bank Development Report, 1993.
11. Potential fee revenues are limited by the high cost of implementing collection systems and by inadequacies in these systems. The costs of establishing, administering and operating fee collection systems may be high. In addition, gross inefficiencies in collection systems limit the potential of private financing. Deficiencies in these systems, which are difficult and costly to overcome, should be taken into account when the potential of fees is assessed. The structure and poor administration of exemption systems -- which often fail to protect the poor while providing substantial numbers of non-poor individuals with fee waivers -- also limit potential fee revenues.
POLICY ACTIONS
1. Government and NGO programs should consider introducing limited fees for family planning services. Without some level of cost recovery, programs will be unable to expand services to meet high and growing levels of demand for services, make urgently needed quality improvements, or expand their service base to include elements of the ICPD Programme of Action. Indeed, the inadequate, existing service base will be increasingly under threat unless some cost recovery is instituted.
2. Means testing cannot be relied upon as a strategy to ensure that the poor and other vulnerable groups have access to services in a fee-charging system. Available evidence suggests that means testing, as currently implemented, is extremely costly and fails to protect many who need protection, while providing exemptions for those who are able to pay. The imperfect, short-term solution may be to charge fees at a very low level, and to use simple targeting strategies that do not require means testing, such as targeting by geographic area, facility type, etc. This implies accepting both limited cost recovery and mistargeted exemptions as inevitable until more is known about ways to institute effective means testing in the circumstances that prevail in the region, at reasonable cost. At that time, higher fee levels, and more effective, specific means testing systems, can be instituted.
3. At least some fee revenues should be retained by the service delivery outlets that collect them and, to the extent possible, used to improve service quality. This will enhance revenue collection and may mitigate the demand-dampening effects of fees. Quality improvements are also important in their own right.
4. Given the concern that high fees may deter contraceptive use and given the poor performance of means testing, greater attention should be given to encouraging the growth of the commercial sector and cutting costs in subsidized programs. These strategies are discussed in the following two chapters. Programs should center their resource mobilization efforts on strategies such as encouraging public sector users with the ability to pay to use commercial sector services, and creating an environment conducive to commercial sector growth.
Footnotes
- Clinics were not randomly assigned to experimental or control groups. It is possible that the clinics making the changes were in some way different from the ones that did not, and that part of the measured impact has to do with initial clinic characteristics. It is important that facilities be randomly selected into experimental or control groups. (FHI, INOPAL and the POLICY Project carried out a study in Ecuador in which clinics are randomly assigned to different price groups.)
- When determining the potential impact of an increase in fees, it is important to note that official prices may underestimate the amount clients currently pay. There is evidence from both the health and family planning sectors that clients often pay more for services than officially stipulated. Even where services are officially provided free of charge, people sometimes are charged for them at the point of service. In Ghana, the price charged for contraceptives was between 200 percent and 638 percent higher than the official price, depending on the method (Kress, 1995). The Population Council Situation Analysis in Senegal found similar price variability among facilities.
- In studies that use availability of contraceptives as a measure of quality, the impact of quality on demand may appear to be limited due to endogeneity. In a type IV study of Kenya, for example, Mwabu et al. (1993) focus on drug availability as a quality measure. They point out that lower quality facilities may lack drugs due to supply problems, but higher quality facilities may lack drugs at any given moment precisely because services are in such high demand. In econometric analyses of demand for family planning services, it may be that higher quality outlets lack contraceptives because of high demand, thereby obscuring the role of quality.
- In determining the impact of price changes on contraceptive use, it might appear that the most straightforward approach is to ask people how their behavior would change if prices increased; however, it is not clear that responses to these types of questions are reliable. Until results from willingness to pay studies are validated, we must rely on empirical results, derived either from cross-sectional household studies in which inferences are made about behavior or from field-based studies (with or without adequate controls) that assess the impact of actual price changes. POLICY/FHI/INOPAL collaboration in Ecuador is addressing the issue of validity of respondent results on willingness to pay.
