Poverty and enrolment in Ghana’s National Health Insurance Scheme
Equity in coverage is at the heart of the NHIS’ policy. Yet our findings show that the poor are less covered in the scheme. The lower enrolment among the poor supports previous studies that enrolment in the NHIS increases with income [
11,
12]. In their study in a rural district of Southern Ghana, Sarpong and others found that 38 % of households surveyed were currently enrolled in the NHIS. Of this, 21 %, 43 % and 60 % were of poor, middle and high SES households respectively [
12]. Our results show that poverty affects enrolment but is less significant in membership renewal. Thus the poor when enrolled were more likely to remain in the scheme. This suggests that the poor value health insurance more than the rich. The question then is: why is the poor the least enrolled in the NHIS? Our results and evidence from earlier studies indicate that many healthcare interventions, targeted at the poor, do not often reach them as planned is overwhelming [
31‐
37].
When community members who were considered to be core poor by local standards were asked why they had not enrolled in the NHIS or had not renewed their membership, poverty, “No money to pay premium”, was most frequently mentioned. A community leader summarised the situation of the core poor as follows: “They are very poor individuals who have no stable source of income. They struggle to get one meal a day and cannot afford the cost of premium.”
A health provider also described the situation as follows:
The good thing about the NHIS is that the insured patients are coming to the hospital early with less complicated cases. But people who are very poor come to the hospital without insurance and can’t pay their bills so we refer them to the DHIS for exemption to enable us get our money.
Lack of commitment to achieve equity in health insurance coverage
Our results brought to the fore the underlying, less understood causes of the NHIS’s failure to achieve one of its primary objectives: protecting Ghana’s most vulnerable citizens (the core poor) against the disastrous consequences of ill health. Though policy makers’ clearly stated their commitment to bring about equity in healthcare access within five years of NHIS’ introduction through exemptions almost a decade into its operations, the core poor were the least enrolled. The exemption is not reaching them, thus contradicting policy makers’ claim that the NHIS would bring about equity in healthcare access within five years.
First, critical analysis of the criteria for exemption show that one of the conditions policy makers set for granting exemption to the core poor does not reflect local conditions of poverty. The following are the criteria stipulated in the National Health Insurance Regulations 2004, LI 1809:58 for identifying the core poor:
1.
A person shall not be classified as an indigent under a district scheme unless that person
a.
is unemployed and has no visible source of income;
b.
does not have a fixed place of residence according to standards determined by the DHIS;
c.
does not live with a person who is employed and who has a fixed place of residence;
d.
does not have any identifiable consistent support from another person.
2.
The conditions under sub-regulation (1) for ascertaining who is an indigent shall be incorporated in the registration form of a district scheme.
3.
A person assigned the duty by a district scheme of registering persons for the scheme, shall elicit the information required under the sub regulation (1) for the classification of indigents as part of the registration process.
4.
Every district scheme shall keep and publish a list of indigents in its area of operation and submit the list to the NHIA for validation [
38].
Though all these criteria are restrictive, the 1b criterion which DHISs referred to as homelessness, defined as lacking a roof and not having any one to provide care, to qualify the core poor for premium exemption [
39], disqualifies almost everybody. The reality is that homelessness is nearly non-existent in the districts. The core poor are described as persons afflicted with ‘
ohia buburoo’ (abject poverty) who do not have a stable source of income to enrol their household members and renew their membership regularly. In farming communities, a number of them were engaged as labourers by farmers to sow and harvest crops. They were also engaged by boat owners during the bumper harvest in fishing communities. Some did menial jobs and did not have a stable income while others were totally unemployed and usually lived on the occasional benevolence of family members or friends. The core poor normally live in family houses, with friends or in dilapidated houses. Apart from the cities, homelessness does not exist. I did not find a normal homeless person in the 15 communities I visited during my fieldwork. Everybody including the core poor had a home. Homelessness is mostly a characteristic of mentally disturbed people who roam cities and towns. In her study of community concepts of poverty in the CR of Ghana, Aryeetey and her colleagues also observed that homelessness is an inappropriate condition for granting exemption to the poor [
28,
40]. As has been observed, the effectiveness of SHISs is the ability to reduce genuine exclusion [
41]. We therefore question the motive for setting a criterion which disqualifies almost all potential beneficiaries as the basis for granting the core poor exemption.
The relevant questions are: If the poor are the primary target of the NHIS, why set a criterion that excludes them? Whose definition of the core poor should count: the one by policy makers or the one by the community? Analysis of these questions reveals insights regarding the motivation to establish a criterion that eliminates potential beneficiaries: lack of commitment. If policy makers were genuinely committed to exempting the core poor, they would have ensured that all the criteria reflect the conditions of the target group and guide collectors on how to register them. Why these actions were not taken is explained by looking at the political situation at the time that the NHIS was introduced and taking into account the financial implications of granting exemptions to all those who would qualify. Enrolling all the exempt categories would have depleted the country’s budget so they needed a definition that would drastically reduce the financial burden without changing the election slogans.
At the time the policy was introduced and even now the NHIS has a large exemption group (more than half of Ghana’s population). The 2003 Ghana Demographic Health Survey reported that 44 % of Ghana’s population was below 15 years [
42]. The 2000 Population Census showed that 5.3 % were above 64 years [
43]. The 2010 Population and Housing Census also showed that 30.3 % and 4.7 % of the population were under 15 years and 65 years and above respectively [
44]. These nation-wide figures roughly agree with our 2011 survey results which show that children (0-17 years) form 48.1 % of the population and 3.4 % was above 70 years. In the case of the core poor, the Ghana Living Standard Survey shows that about a third (28.5 %) of Ghana’s population lived below the poverty line [
26]. Considering the core poor alone, it means that the government would have to pay premiums for about 2.5 million people. This equals a total cost of about GH₵35 million (US$25 million) per annum, which would be a significant demand on the country’s budget. For a country already overstretched with unfulfilled needs in other sectors, such as education and roads, the money to cover all these exemptions was simply not available. The NHIA debt to health facilities is an indication of the huge financial burden of exemptions on the country’s limited financial resources. By the end of 2008, the NHIA owed health facilities about US$34 million [
36].
The NHIS was introduced nation-wide in March 2004 and elections were held in December of the same year. The homelessness criterion thus seemed a strategy to lessen the financial burden of enrolling all core poor while serving as propaganda to accumulate political capital for the pending election. There was nothing to lose if in practice the homelessness criterion eliminated almost all potential beneficiaries while appearing to be fulfilling the government’s moral obligation to the poor and also showed the international community they were committed to ensuring the poor have access to healthcare. Also, the NHIS policy-making process was characterised by political rhetoric [
19,
45]. What was important at that time was to win votes and not practicalities of implementing the exemption policy. If this had not been the case, and politicians were truly committed they could have ensured that all the criteria set reflect the reality of the core poor. Poverty needs to be defined by the community. This view starts from the assumption that opinion and community leaders understand local conditions of poverty and are in a better position to devise effective guidelines that could be used to identify them. A critical analysis of various methods of identifying poor households, concluded that the community criteria of classifying the poorest members correlated with mean testing and the proxy mean testing considered as the gold standard [
28]. One can hardly reach another conclusion than that the realities of implementing the exemption policy for the poor was intentionally disregarded. As pointed out: “A policy is only as good as its implementation arrangement.” [
46].
First, policy makers recognised that no matter how low the cost of premium, the core poor cannot pay so provided exemption to ensure their inclusion in the NHIS. But the exemption is not reaching them.
When DHIS staff and collectors were asked why they did not enrol the core poor, they demonstrated lack of commitment to the equity goal when discussing the issue.. Their first response was normally the problem of identification. A DHIS staff explained why they hardly exempt the core poor from paying premium as follows: “The main criterion we use is homelessness which disqualifies almost everybody. But we occasionally give exemptions when health providers refer patients who cannot pay their hospital bill to us.” Critical observation of their countenance and further discussions with them revealed that it is not the problem of identification but their answer was often a convenient and morally acceptable excuse. Thus just like previous exemption policies which were not successful, only a few core poor benefit from exemptions under the NHIS. The core poor are unable to claim the exemption they are entitled to. How is this possible?
Contrary to the general opinion that inadequate exemption for the core poor is mainly due to identification difficulties, we argue that the explanation runs deeper. The DHIISs could have used local indicators of poverty such as unemployment, no visible source of income and consistent support from another person which share commonalities with what is stated in the NHIS policy (1a and 1d). Ignoring these community indicators of poverty and using homelessness as the decisive criterion proved a convenient tool to exclude nearly all core poor without seeming unreasonable.
Also, it was observed that DHISs’ often undertake revenue generation activities but virtually do nothing to identify the core poor for exemption. Throughout my fieldwork I did not see the DHISs organising any activity to identify the core poor for premium exemptions. The following comment by a DHIS staff illustrates their attitude towards exemptions for the core poor: We need money so if we go to communities and tell them about exemption for the core poor, how do we get revenue to meet some of our expenditure?
Additionally, collectors who were expected to recommend the core poor to the DHISs to be certified as qualified for exemption usually do not disseminate information about exemptions. They focus mainly on premium collection. One of them said: “I am not paid for enrolling those who do not pay premium. So if I continue registering them how do I get money for the work I’m doing.” These comments and similar others show clearly that collectors’ attitude may have to do with the fact that they were not paid for registering the ‘exempt group’. This practice defeats the purpose of the NHIS as a safety net, which is expected to provide the poor with access to healthcare when ill and not when they are unable to pay for their healthcare. One can thus conclude that DHISs’ staff’s decision not to pursue NHIS’ equity agenda was often based on the financial implications of exemptions for their offices. The premium contributes significantly to their internally generated fund (IGF) which they need to meet some of their recurrent expenditure. The IGF is also used to judge the performance of DHISs. These created disincentives to exempt people especially the core poor whose endorsement solely depends on the staff.
Further, many community members told me they were not even aware of the exemption policy. This supports an earlier study on exemption across 18 communities in the Northern, Upper East and Upper West Regions of Ghana also found that 61 % of their respondents did not know about exemptions for the poor [
37]. A study found that exemptions for indigents (core poor) under Ghana’s NHIS were 1 % in 2008 [
36]. In their study on exemptions of Community Health Fund in Tanzania observed that the managers often refuse request for exemptions because they felt it reduces their revenue [
34].
These observations clearly indicate lack of interest from the DHIS staff and collectors to pursue NHIS’s equity agenda. Thus NHIS’ purpose as a safety net, which is expected to ensure equity in healthcare access, is not achieving its objective. Thus macro-level policies targeted at the poor often fail to achieve their objectives because of implementers’ attitudes.
We draw on Lipsky’s concept of street-level bureaucrats’ to explain the gap between the exemption policy and implementation. In his analysis of frontline public service workers’ behaviour in the United States of America [
47]. He observes that they generally display a high margin of discretion and their actions effectively become public policy rather than the objectives of the documents developed at the policy level. Similarly, in this study DHIS staff and collectors can be described as stress-level bureaucrats who used their discretion and decide which aspects of the NHIS policy needs to be pursued: revenue generation or exemption. As has been pointed out by earlier researchers, there should be trade-offs to achieve both goals [
48]. We found that DHISs’ staff and collectors, who are to pursue both goals, have gradually shifted their focus in favour of revenue generation. They were more concerned about increasing their IGF to enhance their image and not to vigorously look for people to exempt since giving exemptions means they lose revenue.
The argument that policy makers and implementers are not committed to pursuing the NHIS’ equity agenda is strengthened when one considers the fact that the non-applicability of the homelessness criterion has been discussed in the public domain since the NHIS was introduced, yet no solution is in sight. This is not to say that nothing is being done to improve the applicability of the exemption policy, but the moral urgency required of both past and present governments seems to be lacking.
The amended NHI Law, Act 852 section 29 (d) states that ‘a person classified by the Minister responsible for Social Welfare as an indigent (core poor), qualifies for exemption and are enrolled in the NHIS [
49]. With this, the NHIA depends on the Livelihood Empowerment against Poverty (LEAP)
5 Programme to identify the core poor. The LEAP started in March 2008 by the Government of Ghana as a social cash transfer programme to vulnerable households across the country. It is still in its trial phase and has reached only 35,000 individuals [
50]. Eligibility is based on poverty and having a household member in at least one of these three demographic categories: single parent with orphan or vulnerable child, elderly poor, or person with extreme disability and unable to work. Initial selection of households is done through a community-based process and verified centrally with a proxy means test. It must be noted that none of the core poor covered by the project in intervention communities in March 2011, benefited from the LEAP programme. This suggest that despite the improvement in the exemption criteria, it still does not adequately resolve the problem of excluding the poor from the scheme. Thus the LEAP though laudable in its current form seem not to be the panacea to the exclusion of majority of the poor in the NHIS.
These reflections clearly show that even if the implementation arena is littered with other barriers, policy makers’ professed goal of ensuring equity in healthcare access is not being given the urgency it requires to ensure all core poor are included in the NHIS.