The decision by the government to convert contract staff to PNS had three main effects. First, it increased the total number of permanent civil servants in the health sector in these three districts by 43% - as a result the proportion of healthcare staff who are PNS increased from 53% to 65% while the proportion on contract decreased from 39% to 28%. Second, because the district governments then hired additional staff on local contracts, the total number of public sector healthcare providers increased by 16%. Third, there has been a significant increase in the public sector salary costs, partly due to the increase in total number of public sector staff (including the new hires on local contracts to, at least partially, replace those converted to PNS) and partly due to the increased commitments of the government to the benefits, including pensions, for the PNS staff.
There were considerable differences between the professions - overall the number of PNS nurses increased by 51%, midwives 35% and doctors 27%. There are also considerable differences between the districts.
These are very significant changes in the health sector within the space of a year. The results presented here for three districts, the outcome of a national policy, are likely to be indicative of what happened across the nation. They have important implications for the sector as a whole in terms of funding, decentralization, health facilities, sector performance, and the direction in which the sector is heading.
Clearly any increase in the number of staff also increases the salary bill, already running at 40% of all public expenditure for health at the district level in these districts . Increasing the proportion of healthcare providers who are PNS has the effect of increasing the cost of salaries and benefits (including their own health care) for civil service staff, the item that, under Indonesia's decentralization, has first call on the general allocation fund from the central government. Thus, in the absence of a matching increase in the general allocation fund, those available for operations expenses are decreased by the extent to which salaries increase. Because the decision to increase the numbers of PNS and the use of central transfers to meet the additional salary costs are made by the central government, the funds over which the district has discretion are also decreased, an action that continues the central claw back of control over funds supposedly, under Indonesia's decentralization , now under the control of the district. At the same time, the districts have replaced many of the staff converted to PNS with additional staff on local contracts. Even though it was not possible to repeat the detailed assessment of use of public funds made for 2006 , we know that an increase in PNS numbers decreases the funds over which the district has control. At the same time this central decision relieves district authorities of the need to increase productivity and rationalize their staffing patterns and levels - why worry when the central government will continue to pay. Districts had been creating some flexibility in their hiring patterns through the use of contract staff. Although it appears that flexibility is now reduced as the central government has prohibited districts from hiring contract staff, in reality it seems that is not the case for despite this 'ban', districts are still hiring staff on contract, apparently without incurring sanctions from the central government. The overall effect is that the public sector salary bill for the health sector has increased.
The public portion of the Indonesian health system has low levels of productivity [P Heywood, NP Harahap. Health centre productivity in West Java Province, Indonesia. Unpublished] and the performance of the sector is inadequate . In addition, the quality of care in Indonesia is low [8–10]. Merely increasing the total number of staff or the number who are permanent civil servants without addressing the more systemic issues  is unlikely to raise the quality or the overall performance of the system.
Finally, this increase in the overall number of providers in the system also results in an increase in the private sector facilities - across the 3 districts, the 680 additional staff are associated with 511 additional facilities, 369 solo-provider facilities and 142 additional multi-provider facilities. As is already the case, the quality of these additional services (public or private) is also likely to be limited and there is an urgent need for the public sector to take its stewardship functions seriously. Even so, the district governments have few resources devoted to oversight of the quality of care. An increase in the PNS staff alone without serious efforts to monitor service quality is unlikely to lift the mediocre performance of the sector.
Whilst this increase in overall staff levels (and, indirectly, facilities) and in the number and proportion of permanent civil servants might be applauded as an attempt to improve the low density of health service providers in Indonesia  its effect on health system performance is likely to be limited because it is not part of an overall coherent approach to improving the performance of the health sector . At the same time the health information system is unable to provide the district level information needed to track changes and understand human resources for health at the district level. Because policy for human resources in health is weak and civil service reform has stalled, the central, provincial and district governments appear to be operating independently with respect to human resources in health. For political reasons the central government decided to convert various forms of contract workers (mostly PTT for doctors and midwives and only local contract for nurses) to PNS. The West Java Provincial government has been using its own resources to fund extra doctors and other health staff through a provincial contract scheme and BHL (a provincially-funded scheme which is found only in West Java). The districts have decided to maintain flexibility of hiring through the use of 'new' categories of local contracts. Each level of government responds to a different constituency and independently of the other. The human resources for health policy is uncoordinated and weak. Further, there is no overall health strategy which addresses the health problems of at least the next 30 years to provide a context for the development of policy about its most important asset, human resources. Indonesia needs both the strategy and the policy as soon as possible.