The adoption of telemedicine, precipitated by the COVID-19 pandemic, is raising both worries about care overuse and reduced quality and hopes for improved convenience and access for immobile and rural populations. Balicer, Einav, Rashba, and Zeltzer estimate the effects of increased access to telemedicine following widespread adoption during the March-April 2020 lockdown period in Israel. They focus on the post-lockdown period, which was characterized by a partial (and temporary) return to normalcy. Using a difference-in-differences framework, the researchers compare primary care episodes before and after the lockdown, between patients with high and low access to primary care, defined based on their main primary care physician's recent adoption. Increased access to telemedicine results in visits involving slightly more tests and follow-ups, and slightly fewer prescriptions. Results are consistent with reduced diagnostic certainty in the absence of physical examination. However, analyzing specific conditions, Balicer, Einav, Rashba, and Zeltzer find no evidence for missed diagnoses. Telemedicine access is also associated with a modest increase in patient demand for care, but not in total healthcare cost. Overall, evidence suggests that telemedicine may complement in-person care without compromising care quality or raising costs.
Retirement savings abandonment is a rising concern connected to defined contribution systems and default enrollment. Goodman, Mukherjee, and Ramnath use tax data on Individual Retirement Accounts (IRAs) to establish that in 2017, 2.7% of 72.5 year-old account holders in total abandoned $790 million; the median abandoned account held $5,400. A regression discontinuity design enabled by account thresholds for automatic rollover IRAs reveals that abandonment is 10 times higher in these default plans. State unclaimed property records show a limited role to reduce abandonment. Goodman, Mukherjee, and Ramnath nest their findings in a model of retirement savings featuring forgetting to derive implications for passive and active savers.
Some consumers lack the cash needed to pay for medical care. As a result, they either delay care until they can pay for it or they forgo the care altogether. To test for such a possibility, Gross, Layton, and Prinz study the distribution of monthly Social Security checks among Medicare Part D enrollees. When Social Security checks are distributed, prescription fills increase by 6-12 percent. In that sense, drug consumption of low-income Medicare recipients is "liquidity sensitive." The researchers then study recipients who transition onto a program that eliminates copayments. When those recipients do not face copayments, their drug consumption becomes less liquidity sensitive. That finding implies that, beyond risk protection, generous insurance also provides recipients with the ability to consume healthcare when they need it rather than when they have cash. Further, Gross, Layton, and Prinz find that recipients whose drug consumption is most liquidity sensitive exhibit price elasticities of demand that are twice the size of the average elasticity, suggesting that more-generous insurance causes recipients both to re-time prescription filling and also to start filling prescriptions that they otherwise would not fill. They present a stylized model that uses this finding to call into question the conventional interpretation of demand-response to price as solely inefficient moral hazard.
This paper was distributed as Working Paper 27977, where an updated version may be available.
The poor live paycheck to paycheck and are exposed to strong cyclical income fluctuations. Akesaka, Eibich, Hanaoka, and Shigeoka investigate whether such income fluctuations affect risk preference among the poor in the US. If risk preference temporarily changes around payday, optimal decisions made before payday may no longer be optimal after, which could reinforce poverty. By exploiting Social Security payday cycles in the US, the researchers find that risk preference among the poor relying heavily on Social Security changes around payday. The deterioration of mental health rather than cognitive decline before payday may play a role. Akesaka, Eibich, Hanaoka, and Shigeoka find similar evidence among the Japanese elderly.
This paper was distributed as Working Paper 28784, where an updated version may be available.
Zhong theoretically analyzes and empirically identifies the optimal default savings rate in automatic enrollment retirement saving plans. The model suggests the optimal default is determined by two counterbalancing forces: If individuals tend to adhere to the default option and procrastinate in making an active decision, the optimal default should be high to encourage people to opt out of the default. If individuals tend to actively undersave, the optimal default rate should be low to encourage people to stay at the default. Based on the model, Zhong derives an explicit formula for the optimal default as a function of reduced-form sufficient statistics that can be empirically identified. He estimates the statistics measuring individual adherence to the default rate using exogenous increases in the default savings rate of OregonSaves, the first state-sponsored auto-enrollment plan in the U.S. Using survey data from OregonSaves-eligible workers, the researcher quantifies the degree of undersaving if workers actively switch to a non-default rate. Combining estimates from administrative and survey data with the optimal default formula, Zhong finds that the optimal default savings rate is 7% of income.