A long-standing challenge in the US health care system is the provision of medical services to rural areas. This paper develops a structural spatial equilibrium model with heterogeneous physicians and uses it to explore the impact of policies on the geographical distribution of physicians. I allow for physicians’ preference to remain close to their residency location. I collect micro data from physicians’ directories on their medical school, residency, and first-job choices and use this new dataset to im- plement an instrumental variable approach to overcome endogeneity issues caused by the correlation between wages and unobserved amenities. I find that residents strongly prefer remaining close to their residency location. Current policies have led to 1.2% more physicians choosing rural areas. Switching the current policy focus toward higher salary incentives would lead to 6 times more primary care physicians choosing rural areas and a higher average quality of rural physicians.
This paper analyzes primary care physicians' (PCP) response to fee-for-service pricing along the urbanity margin. Data analysis of Medicare microdata documents that PCP provide more (remunerative) specialty procedures in less urban areas, where specialists are fewer. Using a structural random coefficients model, the demographic and time variation present in the data, and the policy-set reimbursement fees, this paper finds that an increase in reimbursement for a specialty procedure of ~$36 leads to PCP increasing their market share by 7-15% more than their urban counterparts, at the expense of specialists. Small metropolitan areas and very rural areas are the most affected.
“Physician Workforce Effect on Health” (Draft) [Latest Draft: June 2019]
The effect of the physician workforce on health outcomes of the American population is a key, yet still poorly understood relationship. Much evidence shows that cities attract both more physicians and healthier people, but whether these two facts are causally related is not as obvious. There are many variables, in fact, that are correlated with both the physician concentration and health outcomes. This paper will use an instrumental-variable approach as well as undentifiable public claims data files from New Hampshire and treatment-effects analysis to address this question. IV estimation indicates that an increase in one more physician per 10,000 saves, on average, 4.5 lives per 100,000 residents. The treatment-effects analysis that utilizes micro data confirms the IV results, with an average effect of 6.5 saved lives per physician for every 100,000 residents.
This work takes into consideration the fact that different technologies can be used in housing production, and that factory-built housing, as the most efficient production technology, should be used for the provision of affordable housing. However, due to high levels of regulations, its production is negligible. In this paper, we analyze different policies and their effect on the population in need of affordable housing. To be able to quantitatively measure the cost and savings across the two types of housing, we use micro-level data provided by R.S. Means to be able to price housing components exactly. Finally, we measure the welfare gains that could be obtained by switching to the more efficient production solution.
“A Thorough Analysis of Norwegian Cement Cartel Experience,” (Rough Draft)
The Norwegian cement market has experienced two events which are of particular interest to economists: its cartel period, from 1921 to 1968, and its subsequent monopoly period. The legality of both arrangements provides us with data on the price collusion and decision making of the companies affected unlike other industries where cartel periods happened illegally. This has of course attracted the attention of economists in the past. Röller, Steen, and Sørgard have all shown interest in this phenomenon and they have all provided some explanation of the move from a cartel to a monopoly, reaching the conclusion that the monopoly was actually highly welfare improving under some assumptions. This paper seeks to relax all of these assumptions and shows that the move to a monopoly was not actually as welfare improving as previously shown. Moreover, the improvement is actually due to the increase in prices, and is not welfare improving for consumers. To do so, I go back to the baseline model and increasingly relax more and more assumptions, showing the end change in results. In particular, starting from the baseline model, I make export prices depend on the quantity of cement exported. Then, I let marginal costs increase in time. I further show two generalizations: first, the case in which the export price is below marginal cost; second, the case in which marginal costs are company-specific. I conclude by taking the generalized version and calculate the effect of the change in welfare.
“Corporate Valuation: Measuring the Value of Companies in Turbulent Times” (Wiley 2016)
Buy the textbook I edited and contributed to here!
Other Work in Progress:
“An NP a Day Keeps the Doctor Away,” in progress.
In this paper, I analyze the effect of independent practice of nurse practitioners. In states where nurse practitioners can work independently (currently 23 out of 50), they tend to be more present in places where primary care physicians are lacking. By doing so, they can carry out more of those visits that physicians would normally carry out in cities. The location choice of nurse practitioners is therefore influenced by this margin when they are free to operate without supervision. First, I show that independent practices of nurse practitioners are particularly prevalent in rural areas and in health professional shortage areas in states where their independent practices are allowed. In states where this is not the case, nurse practitioners are concentrated in cities, as they are subject to physicians’ supervision. Because of this, policy changes that allow nurse practitioners to operate independently can be beneficial as they increase the provision of health care to areas which are not covered by physicians. However, the presence of nurse practitioners can subsequently deters physicians’ entry. Due to nurse practitioners’ more limited scope of practice compared to primary care physicians, the number of types of procedures carried out can in fact decrease. I conclude by analyzing which of these two effects is the strongest.
Older Work (Pre-PhD):
“High Product Complexity with Low Financial Literacy in a World of Rationally Bounded Individuals,” MSc Thesis, July 2015, Bocconi University (Draft)
This thesis combines the concepts of bounded rationality, product complexity, obfuscation, and trust to build a theoretical model which attempts to answer the following question: why do financially illiterate individuals hold highly complex financial products? In my thesis, investors have different degrees of sophistication, with unsophisticated investors being rationally bounded; financial institutions are able to exploit these differences to maximize their profits leading to discriminatory pricing of those product characteristics unobserved by unsophisticated individuals. Having to deal with the possibility of contamination between the two groups, financial institutions solve for the optimal time to reshuffle their product offerings, resetting the proportion of unsophisticated individuals back to the original one. Trust, proxied by the number of years the consumer has been dealing with the same financial intermediary, leads to high switching costs for unsophisticated investors. The latter are then eager to pay a higher price than the fair one, even after learning has occurred. If investors trust financial institutions, the latter are able to earn positive profits in equilibrium without the use of obfuscation (the absence of reshuffling of their products to eliminate the benefits unsophisticated investors obtain from learning). Limiting myself to a Bertrand setting, I can solve for equilibrium prices and consumers’ and financial institutions’ behavior when the complexity of products is not too high. Despite this restriction, the results found can be interesting if applied to products which can be easily compared across institutions, such as mortgages.