Do divergences between stated and actual willingness to pay signify the existence of bias in contingent valuation surveys?

Does improved risk information increase the value of cholera prevention? An analysis of stated vaccine demand in slum areas of urban Bangladesh

As the world's longest running pandemic, cholera poses a substantial public health burden in Bangladesh, where human vulnerability intersects with climatic variability. Barriers to safe water and sanitation place the health of millions of Bangladeshis in jeopardy - especially those who have highly constrained choices in preventing and responding to cholera. In this paper we investigate demand for cholera prevention among residents in the Mirpur and Karail slum areas of urban Dhaka. Using survey data from 2023 households in two slum areas, we analyze responses from a contingent valuation questionnaire that elicited willingness to pay (WTP) for cholera vaccines across household members and under varying disease risk scenarios, finding higher valuation for cholera prevention for children and under scenarios of greater epidemic risk. We estimate the average WTP for a cholera vaccine for a child ranges from TK 134-167 (US$ 1.58-1.96). Consistently, respondents with prior knowledge of the cholera vaccine reported lower WTP valuations, providing suggestive evidence of concerns about vaccine effectiveness and preferences for cholera treatment over prevention. We supplement the contingent valuation analysis with cost of illness estimates from both our household sample as well as from administrative hospital records of over 34,000 cholera patients. We estimate that a household incurs costs of TK 801-922 (US$ 9.43-10.50) per episode of cholera that requires medical treatment. Taken together, these findings indicate higher WTP for cholera treatment compared to prevention, but increased interest in prevention under early warning system scenarios of high disease risk.

Health shock and preference instability: assessing health-state dependency of willingness-to-pay for corrective eyeglasses

BACKGROUND

Differences in contingent valuation (CV) estimates for identical healthcare goods can cast considerable doubt on the true economic measures of consumer preferences. Hypothetical nature of CV methods can potentially depend on the salience, context and perceived relevance of the good or service under consideration. Thus, the high demand elasticity for healthcare goods warrants careful selection of study population as the contexts of valuation significantly changes after experiencing health shock.

METHODS

In this study, using triple-bounded dichotomous choice (TBDC) experiments, we test how negative health shock (namely, being diagnosed with refractive errors), can alter preference over a common health good (namely, corrective eyeglasses). We compared elicited WTP of diagnosed patients with a synthetically constructed comparable cohort without the same health shock, controlling for the possible self-selection using a number of matching techniques based on the observable socio-demographic characteristics.

RESULTS

The consumers diagnosed with vision problems exhibit a rightward shift in their demand curve compared to observationally identical consumers without such problems resulting in about 17% higher consumer surplus. The consumers without the health shock are willing to pay about BDT 762.4 [95% CI: BDT 709.9 - BDT 814.9] for corrective eyeglasses, which gets 15-30% higher for the matched with-health-shock consumers. Multivariable analyses suggest more educated and wealthier individuals are willing to pay respectively BDT 208 and BDT 119 more for corrective eyeglasses. We have tested the models for different matching protocols. Our results are fairly robust to alternate specifications and various matching techniques.

CONCLUSION

The preferences for healthcare goods, such as eyeglasses, can significantly depend upon the respondent being diagnosed with refractive errors. Our findings have implications for general cost-benefit analyses often relying on WTP, which can vary depending on the contexts. There are also increasing interests in cost recovery models, which require understanding the demand for healthcare goods and services. We find eliciting the demand needs to consider the health status of the population from which the respondents are sampled.

The criterion validity of willingness to pay methods: A systematic review and meta-analysis of the evidence

BACKGROUND

The contingent valuation (CV) method is used to estimate the willingness to pay (WTP) for services and products to inform cost benefit analyses (CBA). A long-standing criticism that stated WTP estimates may be poor indicators of actual WTP, calls into question their validity and the use of such estimates for welfare evaluation, especially in the health sector. Available evidence on the validity of CV studies so far is inconclusive. We systematically reviewed the literature to (1) synthesize the evidence on the criterion validity of WTP/willingness to accept (WTA), (2) undertake a meta-analysis, pooling evidence on the extent of variation between stated and actual WTP values and, (3) explore the reasons for the variation.

METHODS

Eight electronic databases were searched, along with citations and reference reviews. 50 papers detailing 159 comparisons were identified and reviewed using a standard proforma. Two reviewers each were involved in the paper selection, review and data extraction. Meta-analysis was conducted using random effects models for ratios of means and percentage differences separately. Meta-bias was investigated using funnel plots.

RESULTS

Hypothetical WTP was on average 3.2 times greater than actual WTP, with a range of 0.7-11.8 and 5.7 (0.0-13.6) for ratios of means and percentage differences respectively. However, key methodological differences between surveys of hypothetical and actual values were found. In the meta-analysis, high levels of heterogeneity existed. The overall effect size for mean summaries was 1.79 (1.56-2.04) and 2.37 (1.93-2.80) for percent summaries. Regression analyses identified mixed results on the influence of the different experimental protocols on the variation between stated and actual WTP values. Results indicating publication bias did not account for differences in study design.

CONCLUSIONS

The evidence on the criterion validity for CV studies is more mixed than authors are representing because substantial differences in study design between hypothetical and actual WTP/WTA surveys are not accounted for.

The impact of a mosquito net voucher subsidy programme on incremental ownership: The case of the Tanzania National Voucher Scheme

The subsidisation of mosquito nets has been widely used to increase ownership in countries where malaria represents a public health problem. However, an important question that has not been addressed empirically is how far net subsidy programmes increase ownership above the level that would have prevailed in the absence of the subsidy (i.e., incremental ownership). This study addresses that gap by investigating the impact of a large-scale mosquito net voucher subsidy--the Tanzania National Voucher Scheme (TNVS)--on short-term demand for unsubsidised commercial nets, estimating a household demand model with nationally representative household survey data. The results suggest that, despite the TNVS using a categorical targeting approach that did not discriminate by wealth, it still led to a large increase in incremental ownership of mosquito nets, with limited evidence of displacement of unsubsidised sales. Although no evidence is found of an additional TNVS voucher decreasing the number of unsubsidised sales in the same period, results indicate that an additional TNVS voucher reduced the probability of purchasing any unsubsidised net in the same period by 14%. The findings also highlight the critical role played by social learning or campaign messaging in increasing mosquito net ownership.

Willingness to pay for cataract surgery is much lower than actual costs in Zamfara state, northern Nigeria

PURPOSE

Direct medical and non-medical costs incurred by those undergoing subsidised cataract surgery at Gusau eye clinic, Zamfara state, were recently determined. The aim of this study was to assess the willingness to pay for cataract surgery among adults with severe visual impairment or blindness from cataract in rural Zamfara and to compare this to actual costs.

METHODS

In three rural villages served by Gusau eye clinic, key informants helped identify 80 adults with bilateral severe visual impairment or blindness (<6/60), with cataract being the cause in at least one eye. The median amount participants were willing to pay for cataract surgery was determined. The proportion willing to pay actual costs of the (i) subsidised surgical fee (US$18.5), (ii) average non-medical expenses (US$25.2), and (iii) average total expenses (US$51.2) at Gusau eye clinic were calculated. Where participants would seek funds for surgery was determined.

RESULTS

Among 80 participants (38% women), most (n = 73, 91%) were willing to pay something, ranging from <US$1 to US$186 (median US$18.5, interquartile range 6.2-31.1). Approximately half of the participants (n = 41) were willing to pay US$18.5 (78% men), one-third (n = 26) were willing to pay US$25.2 (77% men); and 11% (n = 9) were willing to pay US$51.2 (all men). Only six participants (8%) already had the money to pay; one quarter (n = 20) would need to sell possessions to raise the funds.

CONCLUSION

Willingness to pay for cataract surgery among adults with operable cataract in rural Zamfara state is far lower than current costs of undergoing surgery. People who were widowed-most of whom were women-were willing to pay least. Further financial support is required for cataract surgery to be universally accessible.

External Validity of Contingent Valuation: Comparing Hypothetical and Actual Payments

Whilst contingent valuation is increasingly used in economics to value benefits, questions remain concerning its external validity that is do hypothetical responses match actual responses? We present results from the first within sample field test. Whilst Hypothetical No is always an Actual No, Hypothetical Yes exceed Actual Yes responses. A constant rate of response reversals across bids/prices could suggest theoretically consistent option value responses. Certainty calibrations (verbal and numerical response scales) minimise hypothetical-actual discrepancies offering a useful solution. Helping respondents resolve uncertainty may reduce the discrepancy between hypothetical and actual payments and thus lead to more accurate policy recommendations. Copyright © 2016 John Wiley & Sons, Ltd.

Comparing the validity of the payment card and structured haggling willingness to pay methods: The case of a diabetes prevention program in rural Kenya

BACKGROUND

The objective of this study was to compare the theoretical validity of two willingness-to-pay (WTP) methods, the commonly used payment card (PC) and the recently developed structured haggling (SH), for estimating the potential benefits of a diabetes prevention program in rural Kenya.

METHODS

A convenience sample of adult residents from a rural county in Kenya (Kiambu), with no history of diabetes, was randomly assigned to one of two WTP methods, PC or SH, using structured face-to-face interviews from December 2011 to February 2012.

RESULTS

A total of 376 respondents completed the interviews using PC (n = 185) or SH (n = 191). More than 95% of respondents were willing to pay something for program access. The study showed that both methods were feasible in rural Kenya. SH resulted in a higher annual mean WTP than PC, Ksh504.05 (US$7.25) versus Ksh619.95 (US$5.90), respectively (p < 0.01). Based on theory, it was hypothesized that certain predisposing factors would result in greater WTP. Greater socio-economic status (measured using income proxies) resulted in greater unconditional WTP for both the PC and SH groups (t-tests and bivariate correlations) and conditional WTP (GLM models). GLM for PC showed being male, employed and having distant relatives with diabetes were significant predictors for WTP, while for SH being educated, employed and owning a vehicle were significant predictors.

CONCLUSION

Both PC and SH showed theoretical validity in rural Kenya. However, the use of SH over PC in rural Kenya may be the better choice given that SH more closely mirrors marketplace transactions in this setting and the use of SH resulted in more significant variables in the GLM models.

Predictors of long-lasting insecticide-treated bed net ownership and utilization: evidence from community-based cross-sectional comparative study, Southwest Ethiopia

Background

Malaria is the notorious impediment of public health and economic development. Long-lasting insecticide-treated bed nets/insecticide-treated bed nets (LLINs/ITNs) are among major intervention strategies to avert the impact the disease. However, effectiveness of LLINs/ITNs depends on, inter alia, possessing sufficient number, proper utilization and timely replacement of nets. Thus, the World Health Organization (WHO) recommends surveys to evaluate possession and proper use of LLINs/ITNs by households.

Methods

A cross-sectional comparative household survey was conducted during peak malaria transmission season using interviewer-introduced questionnaires in southwest Ethiopia. A study site was selected from villages around a man-made lake, Gilgel-Gibe (GG) and a control site, with similar geographic and socio-economic features but far away from the lake, was identified. A total of 2,373 households from randomly selected cluster of households were included into the study and heads/spouses of the households responded to interviews. Binary and multinomial logistic regressions were used to identify predictors of LLIN ownership and utilization.

Results

LLIN/ITN ownership among the study populations was 56.6%, while 43.4% of households did not own a net. A higher proportion of households in GG reported owning at least one LLITN/ITN compared to control village (OR =2. 2, P <0.001) and more households in GG reported having only one LLITN/ITN in contrast to households in the control village (OR = 2.1, P <0.001). The mean number of LLINs/ITNs owned was 1.6 for GG residents and 1.8 for control village with a mean difference of -0.26 (95% CI = - 0.34, -0.19). The age of household heads, household relative wealth index (RWI), distance to nearest health service and accessibility to transportation showed a significant association with ownership of LLINs/ITNs. The probability of owning two or more LLINs/ITNs was positively associated with age of household head. Marital status of household heads, RWI, distance to nearest health service, accessibility to transport, residence and household size showed a significant association with utilization of LLINs/ITNs.

Conclusion

Attention needs to be given to the poor, distant and inaccessible households in the efforts of malaria intervention programmes, such as free distribution of LLINs/ITNs. Well-tailored information, education and communication is needed to address the problem of non-users.

A systematic review and meta-analysis of willingness-to-pay values: the case of malaria control interventions

The increasing use of willingness to pay (WTP) to value the benefits of malaria control interventions offers a unique opportunity to explore the possibility of estimating a transferable indicator of mean WTP as well as studying differences across studies. As regression estimates from individual WTP studies are often assumed to transfer across populations it also provides an opportunity to question this practice. Using a qualitative review and meta analytic methods, this article determines what has been studied and how, provides a summary mean WTP by type of intervention, considers how and why WTP estimates vary and advises on future reporting of WTP studies. WTP has been elicited mostly for insecticide-treated nets, followed by drugs for treatment. Mean WTP, including zeros, is US$2.79 for insecticide-treated nets, US$6.65 for treatment and US$2.60 for other preventive services. Controlling for a limited number of sample and design effects, results can be transferred to different countries using the value function. The main concerns are the need to account for a broader range of explanators that are study specific and the ability to transfer results into malaria contexts beyond those represented by the data. Future studies need to improve the reporting of WTP.

Benefits of preventing a death associated with child maltreatment: evidence from willingness-to-pay survey data

Although assessing the costs of an intervention to prevent child maltreatment is straightforward, placing a monetary value on benefits is challenging. Respondents participating in a statewide random-digit-dialed survey were asked how much they would be willing to pay to prevent a death caused by child maltreatment. Our results suggested that society may value preventing a death from child maltreatment at $15 million. If a child maltreatment intervention is effective enough to save even 1 life, then in many cases, its benefits will outweigh its costs.

Voluntary counseling and testing (VCT) for HIV/AIDS: a study of the knowledge, awareness and willingness to pay for VCT among students in tertiary institutions in Enugu State Nigeria

OBJECTIVE

To examine the level of awareness of youths to voluntary counseling and testing and willingness to pay for this service, and to explore reasons for underutilization of this service.

METHODS

A cross sectional study was carried out among undergraduate students of two tertiary institutions in Enugu Nigeria using pre-tested interviewer administered questionnaire. Information was collected from 250 respondents per institution. Analysis was done using SPSS computer software package.

RESULTS

Most of the respondents (64%) have heard about VCT and 70.6% of the students obtained their information from the mass media (P<0.05) while a minority (3.8%) heard from families. 76.6% of respondents believe VCT can provide useful information on HIV/AIDS and VCT is obtainable mainly in teaching hospitals (78.5%) and to a lesser extent in government hospitals (9.8%) and NGOs (8.8%), while being almost non existent in private hospitals (2.9%). 81% of the respondents did not attend VCT while only 19% attended. The reasons for non attendance were that majority of the students (45.7%) were unaware of the services (P<0.05), indifferent to VCT (20.0%), (12.8%) felt it was costly and (13.3%) were afraid of discovering their HIV status. About 50% of the respondents were willing to pay for VCT and the mean willingness to pay was $3.2 (N370). Out of those willing to pay, 46% of them are willing to pay ($2.6) N300 while 34% and 20% are willing to pay $3.4 (N400) and $4.3 (N500), respectively (P<0.05). Among those not willing to pay, 67.6% of them think it should be free (P<0.05). Males and people with higher knowledge of VCT stated higher WTP values than females and those with less knowledge of VCT. Log OLS also showed that a higher level in the University was positively related to WTP.

CONCLUSION

The high knowledge of VCT does not reflect on the attendance at VCT clinics. Respondents seem ignorant about where the services can be obtained and they believe VCT should be free or adequately subsidized. The cost of VCT is much higher than the mean WTP and governments should take this into consideration when subsidies are being considered. More VCT centers should be created and widely publicized in various communities.

Hypothetical versus real preferences: results from an opportunistic field experiment

Over recent years there has been renewed interest in cost-benefit analysis (CBA) in health care but the 'hypothetical bias' concern (i.e. the belief that WTP values overstate real preferences) is a remaining anxiety. This paper reports new empirical data comparing hypothetical and real preferences in a health care context, using the clinical setting of patient self-management (PSM) of anticoagulation (warfarin) therapy. The data offer considerable support for the use of WTP and CBAs in a self-management health care context; the hypothetical bias hypothesis is not supported by our data. The generalisability of these results to other health care settings needs to be explored.

A game of two halves? Incentive incompatibility, starting point bias and the bidding game contingent valuation method

The bidding game (BG) method of contingent valuation is one way to increase the precision of willingness to pay (WTP) estimates relative to the single dichotomous choice approach. However, there is evidence that the method may lead to incentive incompatible responses and be associated with starting point bias. While previous studies in health using BGs test for starting point bias, none have also investigated incentive incompatibility. Using a sample of respondents resident in Burkina Faso, West Africa, this paper examines whether the BG method is associated with both incentive incompatibility and starting point bias. We find evidence for both effects. However, average WTP values remained largely unaffected after accounting for both factors in multivariate analyses. The results suggest that the BG method is an acceptable technique in settings where prices for goods are flexible.

Willingness to pay for a hearing aid: comparing the payment scale and open-ended question

RATIONALE & OBJECTIVES

Different question formats elicit different willingness-to-pay (WTP) results, but there is no consensus on which method elicits the most valid WTP. In spite of the methodological controversies, WTP is a potentially valuable tool in health economics to value health services. Our general objective was to provide additional evidence on the validity of two WTP elicitation formats: the open-ended question and the payment scale.

METHODS

We elicited WTP for a hearing aid among hearing aid users (n = 108), using both a payment scale and an open-ended question. We compared the results from both formats. We tested criterion validity by comparing both formats with the actual out-of-pocket payment. Construct validity was tested by examining whether WTP was consistent with positive income elasticity.

RESULTS

The WTP results elicited with the payment scale and open-ended question were not statistically significantly different. Both formats showed good criterion validity, although the open-ended question showed a stronger association with the actual out-of-pocket payment. The open-ended format showed better construct validity, as it was influenced by family income.

CONCLUSION

The results of the present study showed that the open-ended question was more valid than the payment scale question. We, therefore, recommend that in future WTP studies on hearing aids the open-ended question is used to directly elicit WTP values. The same recommendation may apply to other studies where respondents are familiar with costs or payments for the intervention under evaluation.

Willingness and ability to pay for artemisinin-based combination therapy in rural Tanzania

BACKGROUND

The aim of this study was to analyse willingness to pay (WTP) and ability to pay (ATP) for ACT for children below five years of age in a rural setting in Tanzania before the introduction of artemisinin-based combination therapy (ACT) as first-line treatment for uncomplicated malaria. Socio-economic factors associated with WTP and expectations on anti-malaria drugs, including ACT, were also explored.

METHODS

Structured interviews and focus group discussions were held with mothers, household heads, health-care workers and village leaders in Ishozi, Gera and Ishunju wards in north-west Tanzania in 2004. Contingent valuation method (CVM) was used with "take-it-or-leave-it" as the eliciting method, expressed as WTP for a full course of ACT for a child and households' opportunity cost of ACT was used to assess ATP. The study included descriptive analyses with multivariate adjustment for potential confounding factors.

RESULTS

Among 265 mothers and household heads, 244 (92%, CI = 88%-95%) were willing to pay Tanzanian Shillings (TSh) 500 (US$ 0.46) for a child's dose of ACT, but only 55% (49%-61%) were willing to pay more than TSh 500. Mothers were more often willing to pay than male household heads (adjusted odds ratio = 2.1, CI = 1.2-3.6). Socio-economic status had no significant effect on WTP. The median annual non-subsidized ACT cost for clinical malaria episodes in an average household was calculated as US$ 6.0, which would represent 0.9% of the average total consumption expenditures as estimated from official data in 2001. The cost of non-subsidized ACT represented 7.0% of reported total annual expenditure on food and 33.0% of total annual expenditure on health care."Rapid effect," "no adverse effect" and "inexpensive" were the most desired features of an anti-malarial drug.

CONCLUSION

WTP for ACT in this study was less than its real cost and a subsidy is, therefore, needed to enable its equitable affordability. The decision taken in Tanzania to subsidize Coartem fully at governmental health care facilities and at a consumer price of TSh 300-500 (US$ 0.28-0.46) at special designated shops through the programme of Accredited Drug Dispensing Outlets (ADDOs) appears to be well founded.

Demand assessment and price-elasticity estimation of quality-improved primary health care in Palestine: a contribution from the contingent valuation method

This paper proposes a new methodology to assess demand and price-elasticity for health care, based on patients' stated willingness to pay (WTP) values for certain aspects of health care quality improvements. A conceptual analysis of how respondents consider contingent valuation (CV) questions allowed us to specify a probability density function of stated WTP values, and consequently, to model a demand function for quality-improved health care, using a parametric survival approach. The model was empirically estimated using a CV study intended to assess patients' values for improving the quality of primary health care (PHC) services in Palestine. A random sample of 499 individuals was interviewed following medical consultation in four PHC centers. Quality was assessed using a multi-attribute approach; and respondents valued seven specific quality improvements using a decomposed valuation scenario and a payment card elicitation technique. Our results suggest an inelastic demand at low user fees levels, and when the price-increase is accompanied with substantial quality-improvements. Nevertheless, demand becomes more and more elastic if user fees continue to rise. On the other hand, patients' reactions to price-increase turn out to depend on their level of income. Our results can be used to design successful health care financing strategies that include a consideration of patients' preferences and financial capacities.

Willingness to pay for cataract surgery in two regions of Tanzania

BACKGROUND

Knowing what rural populations are willing to pay for cataract surgery is essential if improvements in cost recovery in eye care service provision programmes are to take place. The authors sought to learn about willingness to pay for cataract surgery in two separate regions of Tanzania.

METHODS

Patients desiring cataract surgery were interviewed in Kilimanjaro Region and Iringa Region of Tanzania to learn how much they and their families were willing to pay for surgery and how "wealthy" (using ownership of several household objects as a proxy for wealth) the household was.

RESULTS

60 cataract patients in Kilimanjaro and 49 in Iringa were interviewed. "Wealth" was significantly associated with willingness to pay in each region. The average expressed willingness to pay was 2457Tsh (SD 4534) or approximately $US2.30, which is far below the actual cost of providing the service.

CONCLUSION

There were significant differences in the expressed willingness to pay between Iringa and Kilimanjaro patients, which may reflect differences in the services provided in the regions. Willingness to pay may increase as the population gains familiarity and trust in the service. It may also be increased by ensuring that pricing is uniform and clearly advertised throughout regions and by educating health workers and counselling patients about the real costs of providing high quality surgery. Offering "free" services to all may result in lower expressed willingness to pay.

Willingness-to-Pay and Demand Curves: A Comparison of Results Obtained Using Different Elicitation Formats

Health economists use "willingness-to-pay" to assess the prospective value of novel interventions. The technique remains controversial, not least with respect to the formats under which values are elicited. The paper analyses the results of a series of studies of the same intervention valued by the same population, in which different elicitation formats were employed. The findings support the hypothesis that data collected using different formats give rise to different demand curves, from which different inferences about demand elasticity, profitability and consumer surplus will be derived. Judgements as to the relative merits of rival interventions depend crucially upon whichever format has been used to evaluate each intervention.