Ohio State nav bar

Dr. Joel E Cohen, Abby Rockefeller Mauzé Prof. of Populations and head of the Laboratory of Populations, Rockefeller University

Joel Cohen
February 10, 2015
12:30PM - 1:30PM
038 Townshend

Date Range
Add to Calendar 2015-02-10 12:30:00 2015-02-10 13:30:00 Dr. Joel E Cohen, Abby Rockefeller Mauzé Prof. of Populations and head of the Laboratory of Populations, Rockefeller University Population Projections Extrapolate Population History; Can Taylor's Law Help?Population projections assume that some characteristic of population history will continue in the future. We are exploring how well population history conforms to Taylor's law (TL) and whether TL can be used to evaluate alternative population projections. We are using Norway as an initial case study because of the high quality of its population data. From 1978 to 2010, the distribution of population density among the municipalities, counties, and regions of Norway (excluding Oslo) satisfied TL: for successive years, the variance of population density (among counties or other spatial units in a given year) was a power-law function of the mean of population density (among counties or other spatial units in a given year). The parameters of TL depended on whether spatial units were weighted equally, or by area, or by population. We are investigating two questions. How well do 6 projections of the spatial distribution of the Norwegian population published by Statistics Norway conform to the form and parameters of TL estimated from historical population data? If some of these projections conform to TL better than others, do the projections that are more consistent with TL describe Norway's actual population better than the others?This is joint work, partly published, with Meng Xu, University of New Haven, and Helge Brunborg, Statistics Norway. Cohen, Joel E., Xu, Meng and Brunborg, Helge  2013  Taylor's law applies to spatial variation in a human population.  Genus 69(1):25-60.  DOI 10.4402/genus-491. http://scistat.cilea.it/index.php/genus/article/view/491.   038 Townshend Institute for Population Research popcenter@osu.edu America/New_York public

Population Projections Extrapolate Population History; Can Taylor's Law Help?

Population projections assume that some characteristic of population history will continue in the future. We are exploring how well population history conforms to Taylor's law (TL) and whether TL can be used to evaluate alternative population projections. We are using Norway as an initial case study because of the high quality of its population data. From 1978 to 2010, the distribution of population density among the municipalities, counties, and regions of Norway (excluding Oslo) satisfied TL: for successive years, the variance of population density (among counties or other spatial units in a given year) was a power-law function of the mean of population density (among counties or other spatial units in a given year). The parameters of TL depended on whether spatial units were weighted equally, or by area, or by population. We are investigating two questions. How well do 6 projections of the spatial distribution of the Norwegian population published by Statistics Norway conform to the form and parameters of TL estimated from historical population data? If some of these projections conform to TL better than others, do the projections that are more consistent with TL describe Norway's actual population better than the others?

This is joint work, partly published, with Meng Xu, University of New Haven, and Helge Brunborg, Statistics Norway. Cohen, Joel E., Xu, Meng and Brunborg, Helge  2013  Taylor's law applies to spatial variation in a human population.  Genus 69(1):25-60.  DOI 10.4402/genus-491. http://scistat.cilea.it/index.php/genus/article/view/491.