Competitive and organisational constraints on innovation, investment and quality of care in a liberalised low-income health system: evidence from Tanzania
Maureen Mackintosh* andPaula Tibandebage**
* Professor of Economics
Department of Economics,
Faculty of Social Sciences,
The Open University,
Walton Hall,
Milton Keynes MK7 6AA,
UK
**Training co-ordinator
Research on Poverty Alleviation (REPOA),
P.O. Box 33223
Dar es Salaam,
Tanzania
Word length 8103
Acknowledgements
Research for this paper was funded by the UK Department for International Development (DFID) and by the Open University. We thank: participants in workshops in Dar es Salaam, London, Cambridge and Johannesburg; Lucy Gilson, Gavin Mooney, Samuel Wangwe and Marc Wuyts; also the following researchers and consultants on the project: A.D. Kiwara, P. Mujinja, P. Ngowi, G. Nyange, V. Mushi, J. Andrew and F. Meena. We are grateful to everyone in the fieldwork districts who gave their time to facilitate our research. The opinions expressed here are solely those of the authors, and do not represent the policies and practices of the DFID.
Competitive and organisational constraints on innovation, investment and quality of care in a liberalised low-income health system : Tanzanian evidence
Maureen Mackintosh andPaula Tibandebage
International health policy proposals increasingly emphasise health system strengthening and innovation. In a context of liberalised provision, the scope for innovative health system rebuilding depends on the viability, effectiveness and capabilities of the non-governmental providers. Yet the research literature examining the market behaviour of private health care firmsin developing countries, and the incentive structures and constraints they face, remains limited. We demonstrate here the extent of perverse health care market dynamics found in Tanzania in the late 1990s, in relation to patients’ need for reliable health care,and show thatthe financial and operating fragility of the firms constrainedinvestment and innovation.We aim to focus attention on the challenge for innovative approaches to poverty reduction represented by the current market and business structure of health care in low income countries and to discuss some policy implications.
INTRODUCTION
There is now international recognition, in the policy literature on poverty alleviation and the problems of attaining the Millennium Goals, of the importance of health system strengthening in developing countries and notably in Sub-Saharan Africa (Pearson 2004; Freedman et al 2005).Associated with this theme is an emerging emphasis on the relevance of innovation in health system organisationand behaviour to achieving health improvement (Janovsky and Peters 2006).Given the now widespread recognition of the scale of private funding and provision in health care in low and middle income countries, innovative health system rebuilding necessarily involves an important role for non-governmental providers (Bloom 2004; Bennett et al 2005).
The range of current policies towards private sector innovation in African health systems is large. It includes promotion of private corporate investment in African health systems, as in the 2006 International Finance Corporation initiative[1]; supportive and regulatory initiatives to change small providers’ behaviour, such as a Tanzanian Food and Drugs Authority initiative to train staff, accredit and locally monitor a network of rural drug shops;andnumerous small scale insurance initiatives. All such initiatives require robust information about market and business structure in order to be effective; howeverthe field research-based and analytical literature on the operation of the private health sector in developing countries remains thin (Bloom 2004, Bloom and Standing 2001; Mackintosh and Koivusalo 2005a, Leonard 2000; Söderlund et al 2003; Bennett et al 1997).
We do know however that in low income Africa as in a number of other low income countries, two key features of the current liberalised health care markets interact in ways that may be damaging. First, health care is dominated in all sectors by formal and informal fee-for service provision. Exceptions includethe removal of primary care fees in the public sector in Uganda and a national health insurance system designed to end ‘cash and carry’ in the Ghana Health Service. But in general, the poorera country, the larger the expected share of health funding represented by out-of-pocket spending (Mackintosh and Koivusalo 2005b). In Tanzania an estimated 45% of all health care spending is private expenditure, largely out of pocket[2].
Second, liberalisation of clinical provision has been associated with the expansion of informalised private provision on which large numbers of poor people depend. By ‘informalisation’, we mean a lack of enforcement of basic regulatory constraint including registration requirements; very poor clinical oversight and supervision; absence of quality assurance in provision and medicine sales; and at worst, a shift of health care into an informal sector of unlicensed, unstable and abusive services and drug sales (Bloom and Standing 2002; Asiimwe 2003; Mujinja et al 2003). The implications for care appear to be worst, as economic theory would lead us to expect, at the lowest-charging, lowest-income end of provision: employment of often under-skilled or almost unskilled staff, use of medicines of doubtful quality, over-charging and patterns of abuse (Tibandebage and Mackintosh 2002, 2005)[3].
This combination of charging system and provider behaviour has resulted in a crisis of access and quality of health care that was severe even before the recent wave of out-migration of health professionals from low income countries (Freedman et al 2005; Save the Children 2005;Mensah et al 2005). In this paper, we draw on case studies of the financial situation and business strategies of non-governmental health facilities in Tanzaniain response to market incentives to argue the following.
We show that quality of care was a major casualty of price-focused competition at the lower end of the health care market. Provision by skilled professionals tended to be squeezed out, as didthe scope within private and non-profit providers for cross-subsidy of public health activity from fee income. In association with health businesses’ very precarious financial status, these market pressures were severely undermining the scope for innovative provision and investment by committed staff and investors. Only facilities serving the small upper end of the market, or those receiving substantial subsidy and using it effectively, couldwithstand some of these pressures: commitment alone was not enough. Policies that aim for innovative practice, for effective collaboration and system integration, and for sustainable private investment that can promote inclusiveness, quality and effective response to crisiscan therefore only work if the worst of these perverse market dynamics can be contained and constrained. In this light, the interactions we sketch here between market structure and incentives and business behaviour –we label this interaction ‘market dynamics’–meritfurther policy-oriented research.
Methods
The evidence reported here derives from fieldwork in 1998-9 in Tanzania.The price data are 1998 and the financial data 1997. Two distinct urban health care markets were studied plus part of the rural hinterland of each: a district of Dar es Salaam (the largest city) and a contiguous district of predominantly rural Coast region; and Mbeya in the Southern Highlands, a smaller town with sufficient income to attract support some private providers,and the adjoining district of Mbeya Rural. Interviewees in 10 hospitals and 36 health centres and dispensaries included facilities’ staff and patients on exit from facilities;members of 108 households in the facilities’ catchment areas were also interviewed.
The facility interviews included discussion of pricingand business strategy and facility finances. As agreed with respondents, some of whom were extremely frank, we have somewhat anonymised the financial data by rounding and simplifying, and in some stated instances by combining more than one facility’s data to create ‘representative’ accounts. The case studies presented are of non-governmental facilities: private for-profit ( ‘private’ for simplicitly),and religious-owned and secular NGO(‘religious/NGO’ or‘non-profit’).
price competition and market structure in Tanzanian health care markets
Tanzania relies substantially on non-governmental health care, though less than some African countries. Just under 40% of registered dispensaries in Tanzania in 2001 were in theprivate and non-profitsectors (URT 2002). The 2001 Household Budget Survey suggested that about one in five visits to health facilities were to the non-governmental sectors in all wealth quintiles; furthermore the urban poor rely particularly heavily on non-governmental primary providers (Kida and Mackintosh 2005). The Tanzanian Demographic and Health Survey of 1996 showed that of children taken for treatment for two common childhood illnesses, over 50% on average were seen by non-governmental providers, with higher percentagesin lower quintiles[4].
Two key features of the local market structure in health care in Tanzania in the late 1990s shaped very strongly the observed business behaviour of non-governmental facilities. These features were market segmentation in pricing – that is, the existence of distinct higher and lower charging segments of the market – and strong price-based competition for patients in the lower income, lower charging market segment.
Dar es Salaam city has a high income segment of its population that supports some high-charging non-government, mostly non-profit hospitals. These charged very substantially more than the small Mbeya private hospitals. In primary care however the situation was reversed. In Dar es Salaam, which has limited government and faith-based primary provision, many very low income people rely on private primary care, and prices were correspondingly low. In Mbeya, non-governmental providers were charging on average more (Mackintosh and Tibandebage 2002).
In each local market, mean and median prices for a bundle of services supplied by all primary providers did not differ greatly between the private and non-profit providers, and this was also true of observed payments by patients on exit from facilities. However, the pattern of pricing within the private and non-profit sectors was very different. The private facilities were notably price-conscious, reporting that they observed and responded to the prices of competitors. The price data support this qualitative evidence: the spread of private facilities’ reported prices was narrow as compared to the much wider spread of prices reported at religious-owned and NGOfacilities(Tibandebage and Mackintosh 2002).(Figure 1 also includes government primary facilities imposing formal charges).
Figure 1 Dotplot: primary facilities’ reported prices: robust mean of reported prices for a standard set of services,Tanzanian shillings, 1998
Each local market displayed an emerging pattern of market segmentation (Figures 2 and 3). Each had a low-charging segment, including some large facilities all government and religious-owned. Each also had a more scattered higher charging group of private and non-profit facilities including several larger facilities. Figures 2 and 3 show, by region, scatter plots of median observed charges (on exit) against the robust mean[5] of list prices for a standard set of services. The size of each circle is proportional to the size of facility as measured by out-patient visits. The observed charges are generally consistent with list prices except for a few Mbeya observations where high chargesrelative to prices suggest informal charging in these non-profit facilities (Mackintosh and Tibandebage 2004).
Figure 2Dar es Salaam health centres and dispensaries, median charges plotted against robust mean of prices,Tanzanian shillings, 1998
Figure 3 Mbeya health centres and dispensaries, median charges plotted against robust mean of prices, Tanzanian shillings, 1998
Business strategies within market constraints in the Tanzanian health care markets
Market structure and incentives influence business growth and behaviour; business responses feed back into changing market structure, influencing opportunities and constraints. Using case studies, we examine in this section several key aspects of such cumulative path-dependent change in the relatively young non-governmental health care businesses in Tanzania in the late 1990s. Private clinical practice had been liberalised only in 1991; before that, only a few religious-owned independent facilities had been licensed. All the private dispensaries and health centres and two private hospitals studies had been established between 1993 and 1997; two Dar es Salaam private hospitalshad started earlier than 1991 under a religious ‘umbrella’.
A private dispensary or health centre had by law to be owned by a medical doctor, though this was not always in fact the case. The private hospitals studied were all owned by one or more doctors. This private individual ownership had implications for investment and innovation analysed below. The religious and other non-profit facilities varied greatly in effective ownership and control. Some religious facilities were long established and were experiencing falling activity as a result of competition from new non-profit and private providers.
The private facilities tended to be small:the median number of outpatient visits per year to the privateprimary facilities was 3,358, and only one had over 10,000 out-patients. No private hospital studied had over 2000 inpatients. Religious and other NGO facilities were on average larger, but only one health centre had over 20,000 outpatient visits in the year; two religious-owned hospitals had between 5000 and 7000. Government facilities were generally larger and government hospitals very busy especially in Dar es Salaam.
The emptying middle: the consequences of price competition and bad debt at a Dar es Salaamprivate health centre
One of the most problematic market dynamics observed was the severe pressure on credible facilities in the ‘middle’ of the market. These were facilities seeking to provide quite a wide range of services, using skilled staff, to a local population with incomes well above the lowest but well below the very expensive non-profit hospitals and specialist private clinicsin Dar es Salaam.
The market pressures tending to ‘squeeze out’ of the Dar es Salaam market such reputable providers are illustrated by a case study based on a single private facility[6]. The health centre wasrun by full time medical personnel, not as was more usual by doctors ‘moonlighting’ from the government sector (Tibandebage et al 2001). Operating in a medium density residential area,it served a mixture of firms’ employees, paid for by employers, andpaying individuals.
The health centre provided curative and preventative care, took short-stay in-patients, and randental and eye clinics. The ownership was, like many Dar es Salaam facilities, joint between a businessman and a medical doctor, the businessman having put up funds from the profits of other businesses. Without such as arrangement, it wasalmost impossible for medical doctors to raise capital. The two owners estimated that they had put Tshs 30 million ($US 43,000 at the then rate of exchange)of capital investment into the business in less than five years; their building wasrented.
Recurrent running costs were substantial (Table 1). Rounded expenditure is recorded before depreciation and before owners’ remuneration. The staffing bill includes three clinical officers, nursing staff and a laboratory assistant; the medical officer-in-charge, one of the owners,worked there full time. This was therefore an attempt to build up a substantial business.
Table 1: Activity and recurrent expenditure estimates, main headings, for a private health centre, Dar es Salaam, 1997
Activity / NumbersOPD numbers / 30,000
In-patient numbers / 1,000
Total activity = OPD + (inpatients x3) / 33,000
Expenditure / Tanzanian Shillings (Tshs) / $US equivalent
Total recurrent expenditure:
of which: / 110,000,000 / 157,143
Salaries and wages, including NPF*, allowances, and staff medical treatment / 26,000,000 / 37,143
Drugs and other medical supplies / 65,000,000 / 92,857
Total expenditure/ OPD patient / 3,667 / 5
Salaries and wages/ total activity / 788 / 1
Drugs and supplies bill/ total activity / 1,970 / 3
Recurrent expenditure/ total activity / 3,333 / 5
* National Provident Fund contributions for staff
This facility faced a number of threats to its viability. To cover recurrent expenditure at these patient numbers it had to charge above median observed Dar es Salaamcharges; profitability required chargesof about US$6 per outpatient. In a country where an estimated 57% of the population live below the international dollar-a-day poverty line (URT 2005: 114), these charges are not easily affordable for each illness episode. One of the owners stated that theyhad been losing patients to competing lower-charging dispensaries, andnow had perhaps 60% of the activity of their early years.
All recurrent income was from fees and charges. Most patients (about 75%) were paid for by their employers, who were billed monthly; charges were somewhat higher for employersthan for individual patients, reflecting the fact that employees were treated on credit. The owners stated that while the facility was viable in 1998 in terms of the balance between invoiced income and costs, it had severe cash flow problems because of bad debts. Many company clients did not pay their bills on time, and the National Insurance Corporation was also a bad payer to the health centre for patients from companies entered with it.
The owners employed a number of strategies to keep recurrent costs down. The doctor owner ran the facility.Staff were wellqualified but numbers werelow compared with a Dar es Salaamreligious dispensary studied. From the owners' point of view the salary bill was an overhead cost that had to be recouped from sale of services and medicines. Income sharing with part time service providers wasalso used to keep down salary costs: for example, an assistant dental officer worked at the facility using his own equipment and shared the income 65/35 with the owners.
Facilities such as this faced market pressures to move towards hospital status and away from primary care because of competition from low charging dispensaries and drug shops. This competition reduced income from the sale of medicines – a key element of profit -by forcing down prices. Furthermore, the doctor owner saw the health centre as at a market disadvantage because it was paying higher salaries to qualified personnel while some other local facilities employed Red Cross or nursing assistants. Given the poor economic situation and lack of money in individual patients’ hands, and given that patients generally did not distinguish between the qualifications of staff in different facilities, individuals tended to go to a cheapest place first, arriving at the health centre only when seriously ill.