Motivated by the importance of including non-poor informal workers into Indonesia’s national health insurance scheme (JKN Mandiri), the Social Security Agency for Health (BPJS Kesehatan) and the National Development Planning Agency (Bappenas) collaborated with the Abdul Latif Jameel Poverty Action Lab South East Asia (J-PAL SEA) to design a pilot aimed at understanding which policy levers can be used to increase the take-up and retention of the program.
This report covers the findings from a Randomized Control Trial (RCT) conducted in Medan, North Sumatra and Bandung, West Java in 2015. The RCT tested the following three interventions:
1) Subsidies for the JKN Mandiri premium (Medan and Bandung);
2) Assisted registration at home as opposed to standard registration at BPJS Kesehatan office (Medan and Bandung);
3) Information on the financial implications of health insurance (Medan), on the waiting period from registration to actual enrollment (Bandung), and on the possibility of uture requirement of JKN membership for accessing administrative documents(Bandung).
For each of the interventions, we study three sets of outcomes. First, we study the effect of the treatments on enrollment. Second, we test whether the characteristics of the households who enroll are affected. Third, we analyze households’ premium payment behavior.
The main findings from the study are as follows:
1) Subsidies were successful at increasing take-up. As far as the characteristics
of the households who enrolled under different treatment arms are concerned,higher subsidies brought in families with better self-reported health and less likely to have prior inpatient history. There were no significant differences along other characteristics.
2) Offering onsite registration significantly increased both the probability that the household expressed willingness to enroll in the onsite registration treatment, many were not successful due to administrative barriers such as not having an updated Family Card. This suggests that administrative barriers may be an important factor hampering enrollment. The results show no evidence that onsite registration treatment attracted different types of households.
3) Neither the information on financial illustration, potential consequences of notsigning up, nor waiting period information increased actual take-up. There are no clear effects on selection of the information treatments.
The analysis presented in the report uses data collected through a baseline survey and administrative data on enrollment, claims and payments from BPJS. The administrative data is important to obtain accurate enrollment status of everyone in the study, including the respondents assigned as control group, health care service utilization and payments. The analysis uses an administrative dataset that the research team obtained from BPJS Kesehatan in May 2017, which covers utilization data until March 2017.
This report covers the findings from a Randomized Control Trial (RCT) conducted in Medan, North Sumatra and Bandung, West Java in 2015. The RCT tested the following three interventions:
1) Subsidies for the JKN Mandiri premium (Medan and Bandung);
2) Assisted registration at home as opposed to standard registration at BPJS Kesehatan office (Medan and Bandung);
3) Information on the financial implications of health insurance (Medan), on the waiting period from registration to actual enrollment (Bandung), and on the possibility of uture requirement of JKN membership for accessing administrative documents(Bandung).
For each of the interventions, we study three sets of outcomes. First, we study the effect of the treatments on enrollment. Second, we test whether the characteristics of the households who enroll are affected. Third, we analyze households’ premium payment behavior.
The main findings from the study are as follows:
1) Subsidies were successful at increasing take-up. As far as the characteristics
of the households who enrolled under different treatment arms are concerned,higher subsidies brought in families with better self-reported health and less likely to have prior inpatient history. There were no significant differences along other characteristics.
2) Offering onsite registration significantly increased both the probability that the household expressed willingness to enroll in the onsite registration treatment, many were not successful due to administrative barriers such as not having an updated Family Card. This suggests that administrative barriers may be an important factor hampering enrollment. The results show no evidence that onsite registration treatment attracted different types of households.
3) Neither the information on financial illustration, potential consequences of notsigning up, nor waiting period information increased actual take-up. There are no clear effects on selection of the information treatments.
The analysis presented in the report uses data collected through a baseline survey and administrative data on enrollment, claims and payments from BPJS. The administrative data is important to obtain accurate enrollment status of everyone in the study, including the respondents assigned as control group, health care service utilization and payments. The analysis uses an administrative dataset that the research team obtained from BPJS Kesehatan in May 2017, which covers utilization data until March 2017.