Robert McMillan

A Dynamic Model of Demand for Houses and Neighborhoods

This paper develops a dynamic model of neighborhood choice along with a computationally light multi-step estimator. The proposed empirical framework captures observed and unobserved preference heterogeneity across households and locations in a flexible way. The model is estimated using a newly assembled data set that matches demographic information from mortgage applications to the universe of housing transactions in the San Francisco Bay Area from 1994- 2004. The results provide the first estimates of the marginal willingness to pay for several non-marketed amenities – neighborhood air pollution, violent crime and racial composition – in a dynamic framework. Comparing these estimates with those from a static version of the model highlights several important biases that arise when dynamic considerations are ignored.

Tiebout Sorting and Neighborhood Stratification

Tiebout’s classic 1956 paper has strong implications regarding stratification across and within jurisdictions, predicting (in the simplest instance) a hierarchy of internally homogeneous communities, ordered by household income. In practice, urban areas tend to exhibit varying degrees of within-neighborhood mixing, likely attributable to departures from several standard Tiebout assumptions – the fact that households are influenced by more than public goods packages when deciding where to live, the heterogeneous nature of the housing stock, and the role of employment geography, given commuting costs are non-zero. To shed light on the way these factors influence observed residential mixing, this paper quantifies the separate contributions of employment geography and housing preferences in reducing neighborhood stratification. It does so using an equilibrium sorting model, estimated with rich Census micro-data. Simulations based on the model using credibly-identified demand estimates show that counterfactual reductions in commuting costs lead to marked increases in education segregation and, to a lesser degree, increases in income segregation, as households now find it easier to locate in neighborhoods with similar households. In contrast, turning off preferences for housing characteristics actually reduces income segregation, indicating that the nonuniform distribution of housing serves to stratify households based on ability-to-pay. Related, we show that differences in housing also help accentuate differences in the consumption of local amenities.

Adaptive space/frequency processing for distributed aperture radars

This paper deals a preliminary investigation into space-time-waveform adaptive processing for waveform diverse distributed apertures. The large baseline of such a distributed radar results in angular resolution that is orders of magnitude better than resolution of a monolithic system (single large radar) with the same power-aperture. This capability comes at the cost of grating lobes (multistatics with evenly spaced apertures) or high sidelobes (multistatics with randomly spaced apertures). This paper develops some preliminary solutions to these drawbacks associated with distributed apertures. In particular, the use of approximately logarithmic spacing with each aperture transmitting orthogonal waveforms provides excellent detection performance.

Distinguishing Racial Preferences in the Housing Market: Theory and Evidence

Given the extent of residential segregation on the basis of race and ethnicity in U.S. cities, it is unsurprising that a long line of research in social science has attempted to better-understand the causes and consequences of segregation. One prominent branch of that literature has used housing market data on the observed patterns of residential sorting and corresponding housing prices to make inferences about (i) the nature of household preferences for the racial composition of their neighborhoods and (ii) the extent to which segregation is driven by centralized discriminatory forces versus the decentralized location decisions of households, given their preferences (Zabel, this Volume, and Hite, this Volume).1

Choice and Competition in Education Markets

This paper presents a new approach for measuring the effects of competition on school performance. We use an equilibrium sorting model to generate an intuitive measure of the competition each school faces, captured by the slope of the school’s demand curve. We then show that this competition measure is positively related to school performance using rich Census data: a one standard-deviation increase in competitiveness leads to a 0.1 standard-deviation performance improvement, controlling for a host of other factors. This positive performance relationship is consistent with strong supply responsiveness, relevant to the school choice debate.

Teacher Performance and Accountability Incentives

This paper documents a new empirical regularity: teacher value-added increases within-teacher when accountability incentives are strengthened. That finding motivates a strategy to separate value-added into incentive-varying teacher effort and incentive-invariant teacher ability, combining rich longitudinal data with exogenous incentive-policy variation from North Carolina. Our estimates indicate that teacher effort and ability both raise current and future test scores, with ability having stronger effects. These estimates feed into a framework for comparing the cost-effectiveness of alternative education policies. For illustration, we show incentive-oriented reforms can outperform policies targeting teacher ability, given their potential to influence all teachers rather than a subset.