Skip to main content

Research

Explore our Research

Abstract

The ability to make accurate predictions relating to consumer preferences is a key factor of a digital firm’s success. Examples include targeted advertisements and, more broadly, business models relying on capturing consumers’ attention. The prediction technologies used to learn consumer preferences rely on consumer generated data. Despite the importance of data-driven technologies, there is a lack of knowledge about the precise role that data-scale plays for prediction accuracy. From a policy perspective, a better understanding about the role of data is needed to assess the risks that “big data” might pose for competition. This article highlights potential complementarities in algorithmic learning, which suggest data-scale advantages might be substantial. We analyze our hypothesis using search engine data from Yahoo! and provide evidence consistent with locally increasing returns to scale. The ability to make accurate predictions relating to consumer preferences is a key factor of a digital firm’s success. Examples include targeted advertisements and, more broadly, business models relying on capturing consumers’ attention. The prediction technologies used to learn consumer preferences rely on consumer generated data. Despite the importance of data-driven technologies, there is a lack of knowledge about the precise role that data-scale plays for prediction accuracy. From a policy perspective, a better understanding about the role of data is needed to assess the risks that “big data” might pose for competition. This article highlights potential complementarities in algorithmic learning, which suggest data-scale advantages might be substantial. We analyze our hypothesis using search engine data from Yahoo! and provide evidence consistent with locally increasing returns to scale..

Abstract

The ability to make accurate predictions relating to consumer preferences is a key factor of a digital firm’s success. Examples include targeted advertisements and, more broadly, business models relying on capturing consumers’ attention. The prediction technologies used to learn consumer preferences rely on consumer generated data. Despite the importance of data-driven technologies, there is a lack of knowledge about the precise role that data-scale plays for prediction accuracy. From a policy perspective, a better understanding about the role of data is needed to assess the risks that “big data” might pose for competition. This article highlights potential complementarities in algorithmic learning, which suggest data-scale advantages might be substantial. We analyze our hypothesis using search engine data from Yahoo! and provide evidence consistent with locally increasing returns to scale. The ability to make accurate predictions relating to consumer preferences is a key factor of a digital firm’s success. Examples include targeted advertisements and, more broadly, business models relying on capturing consumers’ attention. The prediction technologies used to learn consumer preferences rely on consumer generated data. Despite the importance of data-driven technologies, there is a lack of knowledge about the precise role that data-scale plays for prediction accuracy. From a policy perspective, a better understanding about the role of data is needed to assess the risks that “big data” might pose for competition. This article highlights potential complementarities in algorithmic learning, which suggest data-scale advantages might be substantial. We analyze our hypothesis using search engine data from Yahoo! and provide evidence consistent with locally increasing returns to scale..

Abstract

We analyze whether receiving care from higher-priced hospitals leads to lower mortality. We overcome selection issues by using an instrumental variable approach which exploits that ambulance companies are quasi-randomly assigned to transport patients and have strong preferences for certain hospitals. Being admitted to a hospital with two standard deviations higher prices raises spending by 52% and lowers mortality by 1 percentage point (35%). However, the relationship between higher prices and lower mortality is only present at hospitals in less concentrated markets. Receiving care from an expensive hospital in a concentrated market increases spending but has no detectable effect on mortality.

Abstract

We analyze whether receiving care from higher-priced hospitals leads to lower mortality. We overcome selection issues by using an instrumental variable approach which exploits that ambulance companies are quasi-randomly assigned to transport patients and have strong preferences for certain hospitals. Being admitted to a hospital with two standard deviations higher prices raises spending by 52% and lowers mortality by 1 percentage point (35%). However, the relationship between higher prices and lower mortality is only present at hospitals in less concentrated markets. Receiving care from an expensive hospital in a concentrated market increases spending but has no detectable effect on mortality.

Abstract

We analyze whether receiving care from higher-priced hospitals leads to lower mortality. We overcome selection issues by using an instrumental variable approach which exploits that ambulance companies are quasi-randomly assigned to transport patients and have strong preferences for certain hospitals. Being admitted to a hospital with two standard deviations higher prices raises spending by 52% and lowers mortality by 1 percentage point (35%). However, the relationship between higher prices and lower mortality is only present at hospitals in less concentrated markets. Receiving care from an expensive hospital in a concentrated market increases spending but has no detectable effect on mortality.

Abstract

We analyze whether receiving care from higher-priced hospitals leads to lower mortality. We overcome selection issues by using an instrumental variable approach which exploits that ambulance companies are quasi-randomly assigned to transport patients and have strong preferences for certain hospitals. Being admitted to a hospital with two standard deviations higher prices raises spending by 52% and lowers mortality by 1 percentage point (35%). However, the relationship between higher prices and lower mortality is only present at hospitals in less concentrated markets. Receiving care from an expensive hospital in a concentrated market increases spending but has no detectable effect on mortality.

Abstract

We characterize the revenue-maximizing information structure in the second price auction. The seller faces a classic economic trade-o§: providing more information improves the e¢ - ciency of the allocation but also creates higher information rents for bidders. The information disclosure policy that maximizes the revenue of the seller is to fully reveal low values (where competition will be high) but to pool high values (where competition will be low). The size of the pool is determined by a critical quantile that is independent of the distribution of values and only dependent on the number of bidders. We discuss how this policy provides a rationale for conáation in digital advertising.

Abstract

We analyzed Wisconsin court records from the period 2001–18 to document trends in hospital lawsuits to recover patients’ unpaid medical bills. These lawsuits increased 37 percent during this period, from 1.12 per 1,000 residents in 2001 to 1.53 per 1,000 residents in 2018, with lawsuits being disproportionately directed at Black patients and patients living in poorer and less densely populated counties.

Games and Economic Behavior
Abstract

We analyze nonlinear pricing with finite information. We consider a multi-product environment where each buyer has preferences over a d-dimensional variety of goods. The seller is limited to offering a finite number n of d-dimensional choices. The limited menu reflects a finite communication capacity between the buyer and seller.
We identify necessary conditions that the optimal finite menu must satisfy, for either the socially efficient or the revenue-maximizing mechanism. These conditions require that information be bundled, or "quantized," optimally. 
We introduce vector quantization and establish that the losses due to finite menus converge to zero at a rate of 1/n2/d_ In the canonical model with one-dimensional products and preferences, this establishes that the loss resulting from using the n-item menu converges to zero at a rate proportional to 1 /n2 .