Working papers

Is physician cost-sharing an alternative to patient cost-sharing?

joint with Martin Salm

October 2024

also available as C.E.P.R. Discussion Paper 19578

This paper examines whether financial incentives for general practitioners (GPs) can effectively replace patient cost-sharing within a gatekeeper healthcare system in which the GP regulates access to specialist care. We estimate the effects of incentives provided to GPs by exploiting a field experiment in which treated GPs received a share of the cost savings they achieved. There was also a set of control practices. The patient population includes about 26,000 individuals and is largely representative of the overall population in the Netherlands. We estimate the effects of incentives for patients provided by a deductible using a simulated instrumental variables approach. For both sets of analyses, we use the same measures of healthcare costs, the same time period, and the same population of patients. Our findings show that the incentives provided to GPs reduce costs by 4.3 percent, mainly by decreasing less-urgent specialist care. A small patient deductible of 375 euros per year reduces costs by 27.8 percent in each month in which patients still have to pay out-of-pocket and 10.4 percent at the annual level, impacting all types of care. In contrast, physicians discriminate by type and priority of care. We do not find clear patterns for the interaction of incentives provided to GPs and patients. Our results suggest that GP incentives may complement but not fully replace patient deductibles in managing healthcare costs effectively.

Learning from Online Ratings

joint with Xiang Hui and Konrad O. Stahl

May 2024

also available as C.E.P.R. Discussion Paper 17006

the earlier February 2022 version with a different title is also available as CRC TR 224 Discussion Paper No. 267

Online ratings play an important role in many markets. However, how fast they can reveal seller types remains unclear. To study this question, we propose a new model in which a buyer learns about the seller’s type from previous ratings and her own experience and rates the seller if she learns enough. We derive two testable implications and verify them using administrative data from eBay. We also show that alternative explanations are unlikely to explain the observed patterns. After having validated the model in that way, we calibrate it to eBay data to quantify the speed of learning. We find that ratings can be very informative. After 25 transactions, the likelihood of correctly predicting the seller type is above 95 percent.

Patient cost-sharing and redistribution in health insurance

joint with Martin Salm and Suraj Upadhyay

January 2024

also available as C.E.P.R. Discussion Paper 18395 and IZA Discussion Paper No. 16778

Health insurance premiums often do not reflect individual health risks, implying redistribution from individuals with low health risks to individuals with high health risks. This paper studies whether more cost-sharing leads to less redistribution and to lower welfare of high-risk individuals. This could be the case because more cost-sharing increases out-of-pocket payments especially for high-risk individuals. We estimate a structural model of healthcare consumption using administrative data from a Dutch health insurer. We use the model to simulate the effects of a host of counterfactual policies. The policy that was in place was a 350 euro deductible. Our counterfactual experiments show that redistribution would decrease when the deductible would increase. Nonetheless, high-risk individuals can benefit from higher levels of cost-sharing. The reason is that this leads to lower premiums because both high-risk and low-risk individuals strongly react to the financial incentives cost-sharing provides.

• Twitter thread from 24 August 2023

How important are user-generated data for search result quality?

joint with Madina Kurmangaliyeva, Jens Prüfer, and Patricia Prüfer

May 2024

also available as C.E.P.R. Discussion Paper 17934

the earlier September 2022 version appeared as TILEC Discussion Paper No. 2022-16, CCP Working Paper No. 22-07, and was discussed in the CCP Policy Brief

Do search engines produce better results because their algorithm is better, or because they can access more data from past searches? We document that the algorithm of a small search engine can produce non-personalized results that are of similar quality to the dominant firm’s (Google) for certain types of search queries. Overall differences in the quality of search results are explained by searches for rare queries, which constitute 74% of the traffic in our data. We conduct an experiment, in which we keep the algorithm of a small search engine fixed and only vary the amount of data it uses as input. Our results show that giving small search engines access to more data about rare queries improves their quality. Mandatory data sharing by large search engines is a necessary, yet not a sufficient condition to increase the competitiveness of the search market.

Increasing price transparency in the Dutch health care market does not affect provider choice

joint with Maciej Husiatyński and Misja Mikkers

September 2024 (only small updates, new version is coming soon)

also available as C.E.P.R. Discussion Paper 15981

Price transparency is often viewed as an effective way to encourage price shopping and thereby lower health care expenditure. Using individual claims data for 6 frequent, non-emergency dermatological procedures, we estimate the short-run effect of unexpected publication of prices by a major Dutch health insurer on spending and provider choice. Visits to the price transparency website surged, but spending, the likelihood to visit a new provider, distance traveled, and type of provider visited remained unaffected.