Is It Good For Health Systems in Latin America?
We were in the mayor’s office in a small village outside Cochabamba, Bolivia to interview officials about the impact of the new Popular Participation Law that had granted municipalities new resources for social and economic programs like health, education, sports and municipal roads and civil works. The mayor was proud of his new proposal to build a hospital to replace the clinic that had only one doctor. I asked him why he thought he could afford a hospital in such a small village. With good transportation to the large hospitals in Cochabamba, it seemed to me no real need existed for a hospital in his village. He smiled and told me that he did not have to pay the salaries of the doctors they would be paid by the Ministry of Health, and his new resources would be enough to cover the costs of medicines and other supplies.
“But the Ministry has frozen hiring for the last four years, they surely will not want to staff a hospital where it is not needed,” I protested. He smiled even more broadly and pointed to the main road to Cochabamba and the cement factory on the other side. “I am a popular mayor, and if I lead my village will follow. We just have to occupy the road between the cement factory and the city where they need cement for all their new construction and soon the government will supply the doctors.” I could see that the process of decentralization was going to be hard to study with this kind of political interplay between the center and the local governments. The myths of years of ideology about the advantages of decentralization were going to be challenged.
And the myths were great. Decentralization has been touted as a means of improving the equity, efficiency, quality, and financial soundness of health systems. It is argued, especially by economists, that local people with local information can make much better decisions than distant bureaucrats. Local people can reflect local preferences better and make better management decisions. They know who works hard and can assign people and resources in more flexible ways without the rigid centrally determined rules.
On the other side, there have been fewer voices, mainly Health Ministry officials, who have feared decentralization as an invitation to chaos, disruption of effective priority programs, local patronage, and waste. While decentralization is being implemented in a growing number of countries, there are few studies that show whether the advocates or the detractors are right.
Perhaps too eager to try to fill this gap, I was able to design and implement a series of studies on decentralization in Chile, Bolivia and Colombia, as part of the Data for Decision Making Project of the Harvard School of Public Health. These studies, funded by the LAC Health Sector Reform Initiative of the United States Agency for International Development, selected countries with a significant period of actual implementation of decentralization policies. We sought to use evidence from their experiences to see if decentralization was making things as good as advocates said or as bad as my experience with the Bolivian mayor suggested.
The research was carried out with an innovative “decision-space” methodology, developed at Harvard School of Public Health, to determine the range of choice (from narrow to wide) allowed to local officials for functions such as financing, service provision, human resources, and governance. We found that the “decision-space” varied among countries as well as over time within countries. Countries tended to give wider choice initially, but to reduce the decision space over time. In Chile, for example, municipalities were initially allowed to determine salaries and to hire and fire staff. Eventually, however, many of the national civil service protections were restored, thus reducing the choice allowed municipalities.
In general, greater choice was allowed over contracting of private services and governance decisions, while the “decision space” for financial allocations tended to be moderate. Human resources, service provision, and targeting of priority programs usually remained centralized. This tended to limit local control over those functions most likely to affect the efficiency of health services.
In each country we selected a well-qualified research team and developed a national data base with a minimum of three years of data for municipalities to examine the impact of decentralization. The most important and reliable findings were related to changes in equity indicators at the municipal level.
In all three countries, we found that per capita health spending was increasing during the period of decentralization. In Chile and Colombia, although wealthier municipalities were spending more per capita than poorer municipalities, the gap between them was narrowing over time, resulting in more equitable allocations. In addition, per capita utilization of health services was increasing and the gap between wealthier and poorer municipalities was also declining.
Three important mechanisms seemed to be responsible for greater equity of allocations. In Chile, a horizontal equity fund called the Municipal Common Fund, reassigned up to 60% of the “own-source” revenues from the wealthier municipalities to the poorer municipalities using a formula based on population and municipal own-source income.
In Bolivia, the mechanism earmarks central government transfers to municipalities, requiring that 3.2% of these transfers be assigned to fund a priority benefits package for mothers and children. And in Colombia, a mechanism mandates that a minimum percentage of central government transfers be assigned to health in general by municipalities.
Since these formulas were largely population-based, they appear to have resulted in more equitable spending patterns and perhaps in protecting priority programs. In Chile, municipalities were only responsible for primary health care, so increases in municipal health funding did not go to hospital-based care. In Colombia, a proportion of one type of intergovernmental transfer was assigned to prevention and promotion, which resulted in a doubling of per capita expenditures on these programs and a narrowing of the gap between wealthy and poor municipalities.
These research findings suggest that neither the advocates nor the detractors of decentralization policies are one hundred percent right. In most cases, decentralization is neither likely to lead to radical improvement in a health system, nor to produce a disaster. However, forms of decentralization that include mechanisms to improve equity, like the Municipal Common Fund in Chile and the earmarking of central funds in Bolivia and Colombia, can definitely improve resource allocations and utilization.
The range of choice allowed to municipalities is quite limited for certain functions that might be needed to improve performance such as hiring and firing, payments to providers, and decisions about health service norms. It seems likely that experimenting with wider decision space, and appropriate incentives for guiding those choices might be worth evaluating for their impact on efficiency and quality. For instance, had the mayor at the beginning of this story been responsible for the salaries of the doctors, he might not have thought it wise to build an unnecessary and wasteful hospital.
Thomas Bossert is a Senior Political Scientist and Lecturer in the International Health Systems Group at the Harvard School of Public Health.
What brings you in to see me today?” This is the typical question asked of new patients during a psychiatric evaluation in a Boston clinic. When asked of a Latino patient, however, the answer…
At no time have I felt more vital than in serving at Partners In Health (PIH), and seldom more needed than in places like Haiti. In the Haitian people, as well as in every other country that PIH serves, we find an…
A waiting room in a charity clinic in rural Haiti. It is a humid afternoon, and huge drops of warm rain are starting to fall. A young woman is watching as her ten-year-old son, Dominique, clutches…