Dr. Farmer has considerable policy experience with natural resource managers. His interests in integrated modeling across multiple disciplines motivates concerns about how human and natural resource models are used together. Most works find that existing methods of integration under-estimate gross system uncertainty.
Dr. Farmer seeks higher resolution data to uncover latent heterogeneity in human and natural system to apply stronger methods to track and to better manage gross system uncertainty. Interest in "Modeling by Precaution" for long run planning has led to interests from integrated watershed management, health promotion effects of installed infrastructure, and small to add higher valuedniche markets for small prodded operating in ecologically sensitive areas to sustIn the local ecology and rural economy toward higher profits and lower risk. For these topics, Dr. Farmer uses numerical integration through Bayesian models to uncover gross uncertainty in forward planning. Dr. Farmer mentors five graduate students and has worked in the last three years with economics, mechanical engineers, soil scientists, public health scientists and regional scientists. He brings together disciplines of simulation programming, regional economic development and public administration to help policy-makers apprehend complex policy.
||JOUNRAL Ross, J., Farmer, M.C., Lipscomb, C. 2011. "Inconsistency in Welfare Inferences from Distance Variables in Hedonic Regressions" The Journal of Real Estate Finance and Economics. 32.4. 435-460|
||This paper uses Monte Carlo simulations to show that significant and efficient hedonic parameter estimates on variables that measure distances to landmarks can be achieved by alternative specifications. The identity of the true landmark(s) responsible for home price response remains largely latent. Information contained in multiple hedonic function specifications using other than the true landmark(s) are virtually statistically equivalent.rn rnWe demonstrate a robust method to estimate a single ‘best,’ or optimal, location; but this estimated optimum may reflect an underlying value for proximity to that location or simply some weighting of multiple locations to the household. rnrnAlternative specifications cannot be statistically well-distinguished as distance variables are measured along identical dimensions (e.g. longitude and latitude). The concern of this work is not about the capacity to specify an aggregate influence of location on housing price; but the inability to identify the components of that effect. This compromises welfare interpretations that rely on inferences from the value of an environmental amenity (disamenity) as the distance to environmental amenities (disamenities) on housing prices. rn|
||Jounal: Michael C. Farmer, Robert D. Cox, Mioamiao Wang, Marty Middleton. A method for simulating the prevalence of drought and below-normal precipitation; implications for rangeland management. Submitted working paper, revise and resubmit.|
||1. Punctuated drought events and protracted low rainfall sequences may require altering management strategies on semi-arid and arid rangelands. Many rangeland plans do not adequately account for either situation. Relatively accepted protocols already exist for estimating the uncertainty of overall drought risk, but estimating the distribution of protracted sequences of low rainfall is often outside the core skill set of most ecologists and rangeland managers.rn2. We suggest a method for adapting a random walk Metropolis Hastings (MH) technique, currently used when estimating the distribution of single events, to estimate the distribution of sequences, such as years of low rainfall. This approach should allow ecologists and rangeland managers to account for protracted sequences of low precipitation by adapting the basic tools already in wide use. We demonstrate this method by estimating rainfall sequences at Post, Texas, USA.rn3. The estimated frequency of drought equal to or worse than the worst drought on record at Post is 127 years out of 11,168 (or 1.08% of the time).rn4. The total number of years that fully exceed the worst low rainfall sequences of record is 955 years in the chain, or 8.55% of the time. Such sequences of low rainfall have had devastating effects in some areas, and understanding their potential frequency should allow managers to better prepare. These outcomes challenge range managers and other applied ecologists to prepare for outcomes that have not occurred during their lifetimes.rn|
||Jounral: Belasco, E., Farmer, M.C., Lipscomb, C. Forthcoming 2012. “Using a Finite Mixture Model of Heterogeneous Households to Delineate Housing Submarkets. ” Journal of Real Estate Research|
||Housing submarkets have been delineated by the characteristics of the house, location or home price but also, at times, by the characteristics of the individual households that occupy a given area. In our work, the variables that help to distinguish submarkets require direct household data, which can be quite cumbersome to acquire. This analysis suggests that the data requirement is actually quite modest. Household data can be extracted from a very small sample if survey response rates are high and the statistical method is powerful. We use a finite mixture model to identify latent submarkets from household demographics that estimates a separate hedonic regression equation for each submarket. The method is a relatively robust empirical tool to extract submarkets from demographic information with far less effort than suspected for two reasons: those marginally key physical characteristics, location or home prices that strongly differentiate markets in a given area are highly correlated to particular demographic mixtures. Also households can belong to several submarkets with different degrees of similitude. Individuals that strongly proto-type a given submarket most heavily define that submarket. Finally, results replicate the submarket sorting, with greater clarity, of a prior study in the same area that used more comprehensive data.|
Energy and Natural Resource Policy SECONDARY
Environmental Policy PRIMARY
Policy Analysis and Evaluation PRIMARY
LAND USE CHANGE MANAGEMENT
BAYESIAN CORSS MODEL INTEGRATION