Strategies to increase health insurance enrollment
In 2015, out-of-pocket health spending contributed to pushing an estimated 89.7 million people into extreme poverty.1 The recent Covid-19 pandemic has also worsened access to care, likely increasing the estimated 3.8-5 billion people who lacked access to essential health services in 2017.2,3
In order to increase access to key health services and reduce the risk of poverty due to health care costs, there has been an increasing trend toward government-run health insurance schemes and universal health coverage in recent decades.4 A key factor in these programs’ ability to increase access to medical care, protect individuals from financial hardship, and be financially sustainable insurance schemes is take-up. Individuals adopt health insurance if they expect the benefits of enrolling outweigh their individual costs.
A review of 11 randomized evaluations in 10 primarily low- and middle-income countries shows that reducing health insurance costs or increasing awareness of benefits can increase insurance adoption. When it comes to costs, research shows that reduced premiums, reduced co-pays, and simplified procedures for enrolling, using, or paying for insurance can increase take-up. In regards to benefits, interventions that build trust in the insurance scheme or inform people about their health risks are effective at increasing insurance take-up. Providing information to help individuals compare the costs and benefits of insurance has generally not been impactful outside of the United States.
Reducing expected costs: More people sign up for health insurance when premiums are lower.
Reducing premiums for key populations, such as informal workers, has been an effective way to reduce expected costs and increase enrollment in most cases. Five out of six studies found that reducing the price of health insurance premiums for public or private insurance increased coverage (Figure 1)      . Despite large impacts on enrollment, however, these subsidies were still far from achieving universal coverage.
In contrast, a 25 percent discount for informal workers in Vietnam had no impact on take-up of the government-run scheme . The lack of impacts may have been due to the smaller discount compared to other studies as well as a coinsurance policy, in which individuals had to pay a percentage of the cost for certain expenses.
Offering individuals the opportunity to try a health insurance product at a subsidized rate enables them to learn about the product’s value and informs their future enrollment decisions after subsidies expire.
Two randomized evaluations in Indonesia  and Ghana  found that after experiencing insurance coverage, individuals chose to keep their health insurance even after subsidies expired. For example, informal workers above the poverty line in Indonesia who received a full subsidy for public health insurance premiums were 4.6 percentage points (86 percent) more likely to have coverage three months after the subsidy ended relative to those not offered the subsidy . Individuals were still 3.9 percentage points (58 percent) more likely to have coverage eight months after it ended. Those who kept their coverage after the subsidy expired were more likely to have filed claims and more likely to have made health care visits in the first year, despite having better self-reported health than those who did not maintain coverage. This suggests that the subsidy helped individuals learn more about their personal health, the quality of covered health care, or other health insurance benefits.
Two studies from Nicaragua  and Kenya  demonstrate that enrollment is likely to decline after subsidies expire in instances in which individuals have a negative experience or minimal engagement. This may include instances when reimbursements are difficult, costs remain high, individuals remain mistrustful of institutions, or the insurance does not cover high-quality care.
In Nicaragua, fewer than 10 percent of those who enrolled in health insurance were still covered six months after the subsidy expired . Results highlight important implementation issues. For example, some people who had enrolled never received proof of their enrollment, were dissatisfied with health care quality at the covered facilities, or did not trust the program. Improving administrative processes, trust in insurance providers, and the quality of covered care is important if policymakers hope for temporary subsidies to increase insurance coverage in the long run.
Complex enrollment or reimbursement processes are significant barriers to take-up.
Evidence from two randomized evaluations highlights that hassles associated with enrollment or reimbursement can deter enrollment  . In Indonesia, at-home assistance increased enrollment from 7.8 percent to 11.3 percent . The success rate in enrolling among those who attempted to enroll was relatively low, however: requirements to produce official documentation to prove family size and composition remained a major barrier.
Still, working with other actors to simplify enrollment or coordinate health insurance plans may introduce confusion, skepticism, or mistrust. In Nicaragua, researchers randomly assigned some informal workers who received an insurance subsidy to enroll at a local microfinance institution (MFI) . This was meant to be more convenient, given the high rate of microfinance use among informal workers in Nicaragua. However, this reduced enrollment compared to workers who enrolled at a government office: many individuals did not understand why MFIs were involved, while others felt MFIs were less trustworthy and could charge hidden fees.
While policymakers should explore opportunities to simplify processes and improve convenience, it is important to consider individuals’ trust in possible partner institutions and how the involvement of additional actors could have negative impacts.
Increasing expected benefits: Patients will not maintain enrollment if they receive unsatisfactory health care.
Exposure to low-quality care under health insurance led to dropout in Burkina Faso  . A community-based health insurance scheme paid health centers a flat rate for taking care of insured patients, regardless of the health services provided. This disincentivized health care workers to treat insured patients and lowered the quality of care , in turn reducing demand for insurance .
If individuals have access to other forms of health care without health insurance, such as traditional medicine, or are distrustful of health care providers, demonstrating the potential higher quality of care under insurance may be particularly important. In setting up these schemes, policymakers should also think carefully about payment structures so that providers do not offer those with insurance lower-quality care.
Simplifying the cost-benefit comparison: Information can help individuals better understand and compare the costs and benefits of health insurance. However, information alone may not be enough to increase enrollment in contexts where important barriers to enrollment remain, particularly in LMICs.
Comparing the costs and benefits of health insurance, including covered services, enrollment processes, premiums, subsidies, and penalties for not having coverage, can be difficult if information about the policy is not transparent. The impact of information is influenced by the credibility of the source and access to care absent insurance, among other factors.
In two studies in the United States, a country with a notoriously complex health insurance market, letters with information on enrollment, estimated costs, and tax penalties for not enrolling had small-to-moderate positive impacts on enrollment, ranging from a 0.91-16.7 percent increase  .5 In one of these studies, information was most impactful for those who had recently become ineligible for their previous free government health insurance . For households with higher income, which were thus eligible for smaller subsidies relative to lower-income households, information about available subsidies actually had a small negative impact on take-up . This suggests that information needs to be customized to the relevant needs of particular subgroups.
Two out of three randomized evaluations in China , Indonesia , and Nicaragua  suggest that information on health insurance programs may have little impact on enrollment decisions for informal workers in LMICs. In China only, health insurance information increased take-up for rural-urban migrants without employment contracts, increasing coverage by 3.2 percentage points (23 percent) . This suggests that while there may be information barriers to health insurance enrollment in contexts outside the United States, other barriers, such as the “hassle cost” of enrolling or the premium cost, may be more influential on an individual’s decision to take up health insurance. The impact of information may differ with the availability of potential substitutes .
Most of the studies reviewed examined the impact of one or two interventions at a time. Because many barriers exist, it is likely that multiple interventions need to be implemented simultaneously to achieve universal take-up: large subsidies combined with information, simplified procedures, and a guarantee of accessible, high-quality care appear necessary but may not be sufficient.
Abdul Latif Jameel Poverty Action Lab (J-PAL). 2021. "Strategies to increase health insurance enrollment." J-PAL Policy Insights. Last modified May 2021. https://doi.org/10.31485/pi.3550.2021
Asuming, Patrick Opoku, Hyuncheol Bryant Kim, and Armand Sim. “Selection and Treatment Effects of Health Insurance Subsidies in the Long Run: Evidence from a Field Experiment in Ghana.” Working Paper, March 2020. Research Paper.
Banerjee, Abhijit, Amy Finkelstein, Rema Hanna, Benjamin A. Olken, Arianna Ornaghi, and Sudarno Sumarto. 2021. "The Challenges of Universal Health Insurance in Developing Countries: Experimental Evidence from Indonesia's National Health Insurance." American Economic Review, 111 (9): 3035-63. Research Paper | Evaluation Summary
Fischer, Torben, Markus Frölich, and Andreas Landmann. 2018. “Adverse Selection in Low-Income Health Insurance Markets: Evidence from a RCT in Pakistan.” IZA Discussion Papers 11751. Research Paper.
Domurat, Richard, Isaac Menashe, Wesley Yin. “The Role of Behavioral Frictions in Health Insurance Marketplace Enrollment and Risk: Evidence from a Field Experiment.” Working Paper, 2020. Research Paper.
Goldin, Jacob, Ithai Z. Lurie, and Janet McCubbin. 2021. “Health insurance and mortality: Experimental evidence from taxpayer outreach.” Quarterly Journal of Economics 136, no. 1: 1-49. Research Paper.
Giles, John, Xin Meng, Sen Xue, and Guochang Zhao. 2021. “Can Information Influence the Social Insurance Participation Decision of China’s Rural Migrants?” Journal of Development Economics 150, no. 102645. Research Paper.