The interpretation of the results should be done with some caution due to the limitations of the study. The results might be affected to some extent by the small sample size. Therefore, we focus our conclusions on strong main relations that are not sensitive to the model specification. Estimations based on bigger samples might show more detailed variations of acceptability, ability and willingness to pay for physician services across population groups.
Another limitation relates to the methodology. Contingent valuation is known to be subject to hypothetical bias . That is respondents might not behave in the real world in the same way as they stated in a hypothetical experiment. However, some empirical studies have shown that open-ended CV (such as that used in our study) produces effect sizes that are rather comparable to real world WTP values (e.g. ). Nevertheless, we are not inclined to interpret the mean WTP as an indication of the possible service fee because this requires detailed analysis of demand behavior under different payment regimes. Our results should be interpreted in terms of the mere existence of the potential for patient copayments and the main value drivers for the patients. They may also serve as an indication of the societal benefits obtained through consumption of services of a given quality.
In the contingent valuation task respondents were presented with the scenario of an official fee. Therefore, their WTP statements might be affected by their attitudes towards formal and informal payment practices. Not all Ukrainians are positive about paying formally . Formal charges are not part of the personal communication between patient and physician. Thus, they do not necessarily add to the coverage of personnel cost (i.e. physician’s income) and do not assure better quality (i.e. quality and access desired by the patient). Besides, they may be charged on top of the informal charges causing a double burden for the patient. Moreover, Ukrainians are well aware of the fact that the official salary rate in the health care sector (1 555 UAH in December 2009 or around 135 Euro) is one of the lowest compared to other sectors of the Ukrainian economy . Patients in many cases may perceive informal payments as an act of solidarity and a necessary supplement to the miserable official salary of physicians . Taking these perceptions into account we might expect that on average the true WTP level of the respondents is higher than the ones stated in the presented contingent valuation task due to lack of trust in official financing channels.
Our results demonstrate that official patient charges have potential in Ukraine. Even when faced with less attractive characteristics, Ukrainians express a rather substantial level of WTP, although less than a quarter of them are willing to pay. However, for physician services with improved quality/access characteristics, the share of those willing to pay is more than 70% with an average WTP of 44.8 UAH for a visit to GP (3.9 Euro) and 51.9 UAH for a medical specialist (4.5 Euro). There are no reliable estimates of the cost of health care services in Ukraine due to the existence of the public funding system where facilities are financed on a line-item budget principle regardless of the number of services provided. These stated WTP levels, however, are rather substantial in comparison to the average monthly salary in the health care sector. Taking into account that primary care specialists are among the low-income medical workers, the stated WTP on average appear comparable or even higher than the official personnel costs.
The introduction of co-payments in the public health sector may have various effects both in terms of consumption patterns and the official cost of the services. The effects on consumption should be subject of demand modeling studies. From the system and provider’s perspective, co-payments generate additional funds that could be redistributed to achieve different goals. Our results suggest that Ukrainians place high economic value on quality and access improvements. Thus, patient charges can only be implemented together with effective investment policies targeted at improving quality and access. The probability to object to pay for these services is mostly explained by low quality/access characteristics. Additionally, the likelihood of the ability to pay and the level of positive WTP are positively and strongly related to the quality/ access profile. Combined with the evidence that Ukrainians in general are not satisfied with the quality of care they receive  these findings underline the necessity of quality/access improvements in health care. A rough and conservative estimation (the difference of the mean WTP in Table 3, objection answers excluded) suggests that the social benefit gained from simple improvements in the state of medical equipment, maintenance of the physician office, and reduction of waiting time is 16.5 UAH per visit to a GP and 23.0 UAH per visit to a medical specialist (in December 2009 prices). This can be regarded as an indication of the investment potential, although more robust estimates based on larger samples may provide more precise indicators. In Ukraine, increasing quality and access can not only be realized through investments in training, capital, and organizational changes, but is also tightly related to personnel remuneration. Failure to satisfy physicians’ needs may provoke both resistance to official charges and double charges: formal on top of informal.
Our results demonstrate that among the zero WTP answers, protesters (the objection motive) are not driven by economic or social barriers. The negative relation with the share of nonworking members in the household only supports this idea: it indicates that the more members depend on one’s alimony, the more responsibility one has for else’s health and life. This might increase the value of health care service and, consequently, decrease the likelihood of objection to pay despite the (in)ability to pay. Moreover, reporting objection to pay does not necessarily lead to similar behavior in real life as it is not related to the payment experience in the year before the survey (i.e. chances of paying in real life are similar for those who object to pay and for those who do not object). Thus, both from a methodological and a policy perspective it is rational not to account for the preferences of pure ‘protesters’.
As for the ability/inability to pay and the level of WTP, objective socio-economic barriers, such as age and economic status, apply. The relations with household income also support the theoretical validity of the construction of the models. It is notable however that inability to pay is related to the perceived income level, while the level of WTP is related to monetary income. This shows that inability is a perceptional issue depending on the evaluation of one’s own income level, while the level of the WTP is defined by real monetary budgetary constraints. This suggests that different mechanisms underlie the two stages of the decision about the willingness to pay for the physician services and this should be accounted for in the WTP modeling.
The substantial share of population that is unable to pay for physician services (at least 11.5% for the medical specialist with attractive characteristics) is concerning. This is in line with the extensive discussions in the literature (e.g. [24, 25]) that patient charges should be implemented together with exemption criteria related to age and income. To relate co-payment levels to the level of income might also help to reduce financial barriers to access although this is difficult to achieve in practice. A successful example is Bulgaria where patient charges are anchored on the minimum income in the country  although this does not eliminate barriers to access.
It is also worth mentioning that we observe a slight preference among Ukrainians for a direct referral to medical specialist and bypassing a GP. This is expressed through the higher likelihood of reporting inability to pay and the lower WTP level for the latter. However, this preference is practically non-existent for services with less favorable characteristics although a bit more explicated for services with more attractive quality/access characteristics. In the latter case, people are willing to sacrifice only around 7 – 8 UAH to bypass primary contact. This indicates that price signals (such as higher charges for specialists’ services without a referral) might still be necessary to discourage bypass practices if official patient charges are introduced. However, the variation of this preference across population groups should be studied in more detail in order to design an effective threshold for discouraging direct visits to a specialist without a referral.
To our knowledge the WTP for physician services in Ukraine has never been studied before. Thus, testing the external validity using other studies may only be done in relative terms. In our study WTP estimates range from 0.9% to 1.9% of household income when protest answers are excluded and from 1.6% to 2.2% when only positive values are considered. This is rather consistent for example with the results from Spain where WTP for physician/out-patient services represent 2% of household income .