The dissertation is a collection of three essays written on policy issues related to U.S. and Chinese healthcare systems. The first chapter, titled "User-generated Physician Ratings—Evidence from Yelp, " analyzes the effect of user-generated physician ratings from online sources on the healthcare market. They become increasingly popular among consumers, but since consumers typically lack the ability to evaluate clinical quality, it is unclear whether these ratings actually help patients. Using the universe of Yelp physician ratings matched with Medicare claims, I examine what information on physician quality Yelp ratings reveal, whether they affect patients' choices of physician, and how they influence physician behavior. Through text and correlational analysis, I show that although Yelp reviews primarily describe physicians' interpersonal skills, Yelp ratings are also positively correlated with various measures of clinical quality. Instrumenting physicians' average ratings with reviewers' "harshness" in rating other businesses, I find that a one-star increase in physicians' average ratings increases their revenue and patient volume by 1-2%. Using a difference-in-differences strategy, I test whether, in response to being rated, physicians order more substances that are desirable by patients but potentially harmful clinically. I generally do not find that physicians substantially over-prescribe. Overall, Yelp ratings seem to benefit patients—they convey physicians' interpersonal skills and are positively correlated with their clinical abilities, and they steer patients to higher-rated physicians. In the second chapter, titled "Consolidation of Primary Care Physicians and Healthcare Utilization, " (coauthored with Liran Einav, Jonathan Levin, and Jay Bhattacharya from Stanford University) we use administrative data from Medicare to document the massive consolidation of primary care physicians over the last decade, and its impact on patient healthcare utilization. Since patients' decisions to visit large or small organizations are likely endogenous, we employ two research designs that attempt to address this selection and isolate the causal effect of the physician organization size on patient healthcare utilization. The first takes advantage of the heterogeneity in the extent of primary care consolidation across healthcare markets, and the second exploits transitions of physicians across organizations. Our preferred specification suggests that visiting large physician organizations leads to a 16% reduction in the patient's healthcare utilization, and that this reduction is primarily driven by fewer primary care visits and lower number of inpatient admissions. In the third chapter, titled "Effects of Primary Care Management in Rural China, " (coauthored with Hui Ding and Karen Eggleston from Stanford University, Min Yu, Jieming Zhong, Ruying Hu, Xiangyu Chen from Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China, and Chunmei Wang, Kaixu Xie from Tongxiang Center for Disease Control and Prevention, Tongxiang, China) we turn our attention to the Chinese healthcare system. Health systems globally face increasing morbidity and mortality from chronic diseases, yet many—especially in low- and middle-income countries—lack strong primary care. We analyze China's efforts to promote primary care management for insured rural population with chronic disease using unique panel data for over 70,000 Chinese in 2011-2015. Utilizing plausibly exogenous variation in management intensity generated by administrative and geographic boundaries—villages within two kilometers distance but managed by different townships, we find that villagers with hypertension/diabetes residing in a township with more intensive primary care management had more primary care visits, fewer specialist visits, fewer hospital admissions, and lower inpatient spending. No such effects are evident in a placebo treatment year. Exploring the mechanism, we find that patients with more intensive primary care management exhibited better drug adherence. A back-of-the-envelope estimate suggests that the resource savings from avoided inpatient admissions substantially outweigh the costs of the program.