ESTIMATES OF THE PRICE ELASTICITY OF SWITCHING BETWEEN BRANDED AND GENERIC DRUGS

We estimate price elasticities of switching from branded to generic drugs for two widely used drugs: Prozac and Zocor. We find the price elasticity of switching varies by drug and is between 0.01 and 0.10. While elasticity estimates for Zocor are robust to the inclusion of controls for supply‐side factors, those for Prozac are not. Our results indicate consumers in managed care plans are most responsive to differences in out‐of‐pocket (OOP) cost, and we estimate that a 10% increase in the OOP cost difference between Zocor and generic Simvastatin increases an individual's probability of switching to the generic by approximately 0.3%. This would result in a modest total savings of $36,700 among our sample of 114,218 privately insured Zocor users. Our finding that individuals are relatively unresponsive to the lower prices caused by generic introduction implies that policies targeting supply‐side behavior are likely to have a larger effect on generic uptake than price‐based inducements. If generic‐uptake did occur immediately within the first 18 months after generic introduction, the total savings among individuals and insurance companies within our sample would be approximately $7 million for Zocor and $255,000 for Prozac. (JEL I11, I18)     

Price dispersion in atomistic competition

This note presents necessary conditions for non-degenerate price dispersion in a continuous atomistic market where buyers with price-elastic demand search sequentially for the lowest price, and firms maximize profit subject to a variable average cost function.     

Can a decline in search cost increase prices?

This paper provides a new explanation of why a decline in consumers’ price search cost may not lead to lower prices. In a duopoly with price competition, I show that when some consumers are captive to one firm, there may be a non‐monotonic relationship between search cost and market power; firms may charge high prices with higher probability and the average price charged may be higher when consumers’ price search cost falls below a critical level. Furthermore, when firms have asymmetric captive segments, expected prices charged by each firm may move in opposite directions as search cost declines.     

The economics of crowd out under mixed public/private health insurance

It is well known that public insurance sometimes crowds out private insurance. Yet, the economic theory of crowd out has remained unstudied. Here, I show that crowd out causes two countervailing effects: (a) the intensive margin effect-since high demanders are crowded out, the private market now has a larger proportion of low demanders on the intensive margin (The intensive margin are those who have already bought private insurance), and so will drop quality to lower the price to the low demanders liking; and (b) the extensive margin effect-before the public insurance expansion, the private sector had lowered quality to make insurance more affordable at the extensive margin (The extensive margin is the next group of people who would buy private insurance if the price decreased), but now that public insurance crowds out the extensive margin, quality can then be raised back up to the high demanders liking.If the extensive margin effect dominates, then a new phenomenon of push out occurs, in which crowd out causes the private sector to raise quality and to increase the number of uninsured low demanders not eligible for public insurance. If the intensive margin effect dominates, then crowd out will cause the private sector to lower quality, causing the phenomenon of crowd-in, in which the number of uninsured low demanders that take-up private insurance increases.These two countervailing effects have important implications for any government policy that desires to eradicate all uninsurance. First, if push out is dominant, then the private sector will respond to the public insurance by pushing out and leaving some people newly uninsured. If crowd-in is dominant, then all people can be insured and the government can do it at a lower-than-anticipated level of expansion due to the private sector crowding in.Received: April 2002, Accepted: February 2003, JEL Classification: I11, I38The views herein do not necessarily reflect the views or policies of AHRQ, nor the U.S. Department of Health and Human Services. I thank Pedro Pita Barros, Hugh Gravelle, and Lise Rochaix-Ranson, and participants at the 2nd Health Economics Workshop at the Universidade Nova de Lisboa for helpful comments.     

The effects of information costs on search and convergence in experimental markets

Nearly all markets contain some kind of friction, making it difficult to reach full efficiency. One ubiquitous source of market friction is the cost of obtaining pricing information. We determine how market performance as well as buyer and seller behavior are affected by the introduction of price information costs in experimental posted markets.Rothschild [Journal of Political Economy (1973) 1283; Journal of Political Economy (1974) 689] theorizes that search is negatively related to knowledge of the price dispersion and the cost of search, and positively related to market price dispersion. We find that market knowledge and the cost of search itself affect search, but we find no evidence supporting the role of price dispersion in search decisions.We also find evidence supporting Smith and Plott [Review of Economic Studies (1979) 133] and Walker and Williams (1988), both of which show price convergence comes from below in posted-bid markets and from above in posted-offer markets. High information costs tend to reduce or eliminate convergence. Average prices are above the market-clearing price for posted-offer markets, and below for posted-bid markets, and never cross the market-clearing threshold, also consistent with Walker and Williams.More generally, our results support the notion of symmetry between experimental posted-offer and posted-bid markets, broadening the relevance of experimental search research.     

Costly Search, Capacity Constraints, and Bertrand Equilibrium Price Dispersion

This article analyzes the impact of transaction (search) costs and capacity constraints in an almost competitive market with homogeneous firms that compete on price. We characterize conditions under which Nash equilibria with price dispersion exist; in equilibrium, firms play pure strategies in prices and consumers adopt a symmetric mixed search strategy. Price dispersion is possible even though consumers all have the same search cost and valuation for the item and prices charged by all firms are common knowledge.     

Inflation and Price Adjustment: An Analysis of Microdata

I use large datasets on prices by products and stores from recent inflationary periods in Israel to compare simple menu cost models with simple uncertain and sequential trade (UST) models. The main empirical findings are (a) price erosion due to inflation explains only a tiny fraction of the variation in nonzero nominal price changes, (b) stores whose last nominal price change was relatively low are likely to choose a nominal price change that is relatively high, (c) stores that reduce their nominal price charge a lower price relative to stores that increase their nominal price, and (d) relative price variability is not related to inflation. Journal of Economic Literature Classification Numbers: E300, E310, D210.     

Estimation of hedonic price functions with incomplete information

Existence of persistent price dispersion suggests that some buyers find lower prices through search and information acquisition, while some sellers charge higher prices by gathering information on potential buyers. If buyers are not fully informed of the lowest price available in the market they end up paying a price higher than if they had full information. Similarly, if sellers are not fully informed about the highest price they could charge, they too suffer by receiving a price lower than had they had full information. This paper develops a hedonic price model that incorporates the effects of incomplete information on both sides of the market and obtains estimates of the discrepancies between market prices and buyers’ maximum willingness to pay and sellers’ minimum willingness to accept. Correlates of such price discrepancies are also explored. We apply the technique to a data set constructed from the American Housing Survey, and find that incomplete information has had a significant impact on housing prices.     

Decentralized international exchange

Abstract.  We utilize a random‐matching model to examine the relationships between market frictions and international trade. In our setting, an individual may choose to search abroad where she may have a cost advantage, but is less likely to meet potential trading partners, owing to higher market frictions. Interestingly, we find that international trade may be associated with lower welfare than autarky. We show how this is due to price distortions resulting from bargaining when there are opportunities for exchange across countries. JEL classification F10, C78, D83     

Dissertation abstract: Essays on pricing in experimental duopoly markets

This dissertation comprises three independent essays that analyze pricing behavior in experimental duopoly markets. The first essay examines whether the content of buyer information and the timing of its dissemination affects seller market power. We construct laboratory markets with differentiated goods and costly buyer search in which sellers simultaneously post prices. The experiment varies the information on price or product characteristics that buyers learn under different timing assumptions (pre- and post-search), generating four information treatments. Theory predicts that price information lowers the equilibrium price, but information about product characteristics increases the equilibrium price. That is, contrary to simple intuition, presence of informed buyers may impart a negative externality on other uninformed buyers. The data support the model's negative externality result when sellers face a large number of robot buyers that are programmed to search optimally. Observed prices conform to the model's comparative statics and are broadly consistent with predicted levels. With human buyers, however, excessive search instigates increased price competition and sellers post prices that are significantly lower than predicted. The second essay uses experimental methods to demonstrate the anti-competitive potential of price-matching guarantees in both symmetric and asymmetric cost duopolies. When costs are symmetric, price-matching guarantees increase the posted prices to the collusive level. With asymmetric costs, guaranteed prices remain high relative to prices without the use of guarantees, but the overall ability of guarantees to act as a collusion facilitating device depends on the relative cost difference. Fewer guarantees, combined with lower average prices, suggest that cost asymmetries may discourage collusion. The third essay investigates the effect of firm size asymmetry on the emergence of price leadership in a homogeneous good duopoly. With discounting, the unique subgame-perfect equilibrium predicts that the large firm will emerge as the endogenous price leader. Independent of the level of size asymmetry, the laboratory data indicates that price leadership by the large firm is one of the most frequently observed timings of price announcement. In most cases, however, it comes second to simultaneous price-setting. This tendency to wait for the other firm to announce its price is especially strong when the level of size asymmetry between firms is low. We attribute the lower than expected frequency of price leadership to coordination failure, which is further compounded by elements of inequity aversion. JEL Classification C91, D43, D83, L11 Dissertation Committee: Timothy Cason (Chair), Department of Economics, Purdue University Dan Kovenock, Department of Economics, Purdue University Stephen Martin, Department of Economics, Purdue University Marco Casari, Department of Economics, Purdue University     

TALKING,SEARCHING, AND PRICING*

I study the implications of interpersonal communication for incentives for consumers to acquire information and firms’ pricing behavior. Firms market a homogeneous product and choose its price; consumers acquire price information at some cost to themselves. Also, each consumer accesses the information acquired by a sample of other consumers—interpersonal communication. An exogenous increase in the level of interpersonal communication decreases the information that consumers acquire, and, when search costs are low, firms price less aggressively. In an extension, consumers may choose to invest in interpersonal communication at some cost. A decrease in the costs of interpersonal communication decreases firms’ competition.     

Price dispersion in the private health insurance industry: The case of Catalonia

This paper presents a vertical and horizontal product differentiation model that explains price dispersion among different kinds of health care insurance firms. Our model shows large insurance firms engaging in price competition with small mutual organizations that serve only a local area and charge lower premiums. We found that, although the market allows the entry of an excessive number of firms, the presence of local insurance companies increases social welfare by increasing the range of products available to consumers. Our conclusions are applicable to OECD countries in general although we rely on Catalonia's data.     

ASYMMETRIC EFFECTS OF MARITAL SEPARATION ON HEALTH INSURANCE AMONG MEN AND WOMEN

Despite widespread interest among policy researchers in the effect of job displacement on insurance loss, there is little focus in the literature on the insurance implications of a married person losing his or her spouse. Using a large household survey, this article finds that despite legislation aimed at protecting separating spouses, individuals remain at risk of losing health insurance in the event of marital disruption. This is especially true for wives who are enrolled in their husbands’ plans prior to marital termination. (JEL D13, I18)     

MEASURING HEALTH INSURANCE BENEFITS: THE CASE OF PEOPLE WITH DISABILITIES

Since 2012, the Congressional Budget Office has included an estimate of the market value of government‐provided health insurance coverage in its measures of household income. We follow this practice for both public and private health insurance to capture the impact of greater access to government‐provided health insurance for working‐age people with disabilities, whose market value rose in 2010 dollars from $11.7 billion in 1980 to $114.3 billion in 2012. We then consider the more general implications of incorporating estimates of the market price of insurance, equivalent to that provided by the government, into policy analyses in a post‐Affordable Care Act world. (JEL D31, H24, I18, J31)     

Who pays the high health costs of older workers? Evidence from prostate cancer screening mandates

Between 1992 and 2009, 30 US states adopted laws mandating that health insurance plans cover screenings for prostate cancer. Because prostate cancer screenings are used almost exclusively by men over age 50, these mandates raise the cost of insuring older men relative to other groups. This article uses a triple-difference empirical strategy to take advantage of this quasi-random natural experiment in raising the cost of employing older workers. Using Integrated Public Use Microdata Series data from the March Supplement of the Current Population Survey, I find that the increased cost of insuring older workers results in their receiving 2.8% lower hourly wages, being 2% less likely to be employed and being 0.7% less likely to have employer-sponsored health insurance.     

WILLINGNESS TO PAY IN CHINA'S NEW COOPERATIVE MEDICAL SYSTEM

Household survey data from China is used to analyze the determinants of households' willingness to pay to participate in the New Cooperative Medical System (NCMS). Wealthier households, households with at least one member who has been treated as an inpatient, and households in counties where NCMS is already established are willing to pay more for the program. Households who carry other forms of insurance, however, have a lower willingness to pay for the NCMS. We also find that the participation fee could be increased substantially to increase the size of the risk pool while scarcely affecting participation rates. (JEL I1, O1, D1)