100%(2)2 out of 2 people found this document helpful
This preview shows page 1 - 8 out of 66 pages.
ECON280DLecture 5: Heterogeneous Agents, SortingCecile GaubertU.C. BerkeleyECON280D. Spring 2019. C. GaubertLecture 5-Sorting1 / 57
Spatial SortingIUntil now, we have seen frameworks withIHomogenous workersIHomogenous firmsICities / regions differ ‘quantitatively’, not ‘qualitatively’ECON280D. Spring 2019. C. GaubertLecture 5-Sorting2 / 57
Spatial SortingIUntil now, we have seen frameworks withIHomogenous workersIHomogenous firmsICities / regions differ ‘quantitatively’, not ‘qualitatively’IIn reality, the composition of cities varies a lotIAnd varies systematically with city characteristics (e.g., size)ECON280D. Spring 2019. C. GaubertLecture 5-Sorting2 / 57
Example: educational attainment varies a lot across citiesShare of population 25 and older with bachelor’s degree or higher0.28 −0.630.23 −0.280.19 −0.230.16 −0.190.14 −0.160.07 −0.14Data source:American Community Survey, 2005-2009, Series S1501Plot:CBSAs for maptileECON280D. Spring 2019. C. GaubertLecture 5-Sorting3 / 57
And covaries with city characteristicsPopulations of three educational groups across US metropolitan areas-1-.50.51Log (demeaned) skilled population share1012141618Metropolitan log populationBachelor's degreeCollege dropoutHigh school graduate-.5-.250.25.5Log (demeaned) skilled population share1012141618Metropolitan log populationBachelor's degreeCollege dropoutHigh school graduateData source:2000 Census of Population microdata via IPUMS-USAECON280D. Spring 2019. C. GaubertLecture 5-Sorting4 / 57
Sorting mattersSkills and sectors are strongly linked to cities’ sizes:(a)Confounds inference: Agglomeration benefits vs compositional effects(b)Confounds counterfactuals: Making NYC 10x larger raises finance’s share ofnational employment and GDP(c)Questions exclusion restriction of Bartik-type instrumentsITheory: all locations produce a homogeneous goodIEmpirics: exploit variation in industrial composition to estimate modelparameters via shifts in local labor demandIShift-share instrument: local composition×national changesIWhat variation does the instrument exploit?(d)Also: agents heterogeneity (e.g. high vs low skill) necessary to studydistributional implications of shocks and policies, beyond aggregate effectsECON280D. Spring 2019. C. GaubertLecture 5-Sorting5 / 57
Continuum of ‘types’ and assignment modelsIA series of papers study sorting in spatial equilibrium with acontinuum of‘types’(see review in Behrens and Robert-Nicoud (Handbook chapter2015)IContinuum of workers skill: Behrens, Duranton, Robert-Nicoud (2014), Davisand Dingel (2016, 2018)IContinuum of firms productivities a la Melitz: Gaubert(2018)IAssignments of heterogeneous individuals/firms to cities, with endogenouscity characteristics determined in equilibriumIThese papers rely on tools from theassignment literatureIPopularized in the trade literature by Costinot (2009)IContinuum case can be quite tractableIProduction, or preferences?