Skip to content

Review_Mehlika

Jump to bottom
Jonathan Conning edited this page Oct 28, 2021 · 1 revision

Comments on Mehlika's proposal:

"Industrial agglomeration and wage inequality"

This research proposal sets out to investigate spatial inequality and in particular wage inequality between regions and industries which might be explained by a relationship between industrial agglomeration and inequality. At the simplest level agglomeration is associated with urbanization and population density. Empirical work cited investigates how well agglomeration effects and skills might explain regional variations in productivity and income, and an interesting question is how to tell those channels apart (e.g. are higher incomes a result of agglomeration or is it that the more educated and skilled move to cities).

There are different possible causal and non-causal mechanisms that might explain an association between agglomeration and the income distribution and some of these ideas tie back to core ideas in growth and development economics. The Kuznets curve posited that inequality would be at first rising (as productivity increases in the cities raised productivity there) and then falling as the world moved from agriculture (the hypothesized mechanism Kuznets had in mind seems to have been that rural to urban migration would over time tend to lower regional disparities) to cities. The persistence of high Agricultural Productivity Gaps however poses a challenge, and since Kuznets wrote the evidence is that inequality across regions is again rising. Cities continue to agglomerate, but now along new human-capital differentiated dimensions (Milanovic, 2016).

A number of empirical studies are cited. As the proposal clearly points out a key issue in this literature is that it is not clear whether the reason why incomes are rising in places of agglomeration is due to direct productivity gains from agglomeration (e.g. spillovers and economies of scale/scope.. as in Krugman 1991 and others).

This study proposes to extend the literature to empirically explore the "relation between wage dispersion in region r and year t with industry i and industrial agglomeration," and cites a study by Chen (2017) as possibly providing a guiding framework. Specifically she suggests collecting firm-level panel data on employment, wage, sales, sector (NAICS 4 digit), region (NUTS 3)" using Amadeus dataset of companies throughout Europe to run a regression of the form:

$$ WageDis_{i,r,t} = β_0 + β_1AggIndex_{i,r,t} + β_2Control_{i,r,t} + Fixed Effects + ε_{i,r,t} $$

where WageDis is the standard deviation of logarithmic wage across all firms in region r and year t, with sector i, and fixed effects are for regions, industry and year.

This is a very interesting research topic with space to contribute. A few comments and questions

  • Your proposal really needs to both explain some of the background literature and questions better and also narrow your own area of focus/attention. You should greatly expand on the relationship between inequality and agglomeration. Spell out some of the leading hypotheses and empirical stylized facts or findings. What are the main debates on this topic? Delve much deeper into the 'endogeneity' (and hence causal identification) issue you bring up.

  • Tells us of broad trends in Europe (where your dataset seems to apply to). Has there been growing wage inequality, is it mostly inter-sectoral (e.g. rural - urban) or intra-sectoral? Don't assume the reader knows any of this, spell it out.

  • Also try to explain what you hope to add with your study.. will you be one of the first to use this data source, will you use new indices or research-designs?

  • You mention that you might try different Agglomeration indices, and the the Ellison-Glaeser index only works at a national level. It wasn't clear why these can't be built for regions. Please explain and expand what kind of alternative indices you have in mind.

  • From what I understand, the EG index has been criticized as not really taking geography into account, and in particular by not accounting for the spatial dissimilarity of the administrative units (e.g. regions), for example Duranton and Overman (2005). Do you plan to explore more explicit geospatial elements or otherwise leverage the features that you might get from a geospatial system? Can you construct/employ any of these indices which make better use of distances and other features that might be extracted from a GIS system?

  • you don't say much about your controls... What variables there? Do political policies matter?

  • Spend more time presenting different theories behind the agglomeration phenomena and in particular how/why businesses would chose to locate.

  • Maybe exchange notes with Philipp who is looking at how urbanization might have been accelerated by unification. Something similar going on in Europe.. Krugman and many others have written about effects of EU on agglomeration and spatial re-allocation. Maybe questions such as these go somewhat beyond the framework you have in mind.. but maybe you can discuss as you discuss controls.

  • You mention the Chen study . It uses an IV approach which relies on the "Bartik" instrument from shift-share analysis. This method has come under some scrutiny (see also Jaeger et al) lately. One contribution might be how you intend to address and defend the necessary assumptions to do this right in the literature.

References cited

Ades, Alberto F., and Edward L. Glaeser. 1995. “Trade and Circuses: Explaining Urban Giants.” The Quarterly Journal of Economics 110 (1): 195–227.

Chen, Anping, Tianshi Dai, and Mark Partridge. 2017. “Agglomeration and Firm Wage Inequality: Evidence from China.”

Davis, Donald R., and Jonathan I. Dingel. 2019. “A Spatial Knowledge Economy.” American Economic Review 109 (1): 153–70. https://doi.org/10.1257/aer.20130249.

Duranton, Gilles, and Henry G. Overman. 2005. “Testing for Localization Using Micro-Geographic Data.” The Review of Economic Studies 72 (4): 1077–1106.

Ellison, Glenn, Edward L. Glaeser, and William R. Kerr. 2010. “What Causes Industry Agglomeration? Evidence from Coagglomeration Patterns.” American Economic Review 100 (3): 1195–1213.

Jaeger, David A, Joakim Ruist, and Jan Stuhler. 2018. “Shift-Share Instruments and the Impact of Immigration.” Working Paper 24285. Working Paper Series. NBER https://doi.org/10.3386/w24285.

Kuznets, S (1955), “Economic growth and income inequality”, American Economic Review, March, pp. 1-28.

Krugman, Paul (1991). Geography and trade, London MIT Press

Milanovic, B (2016), Global inequality: A new approach for the age of globalization, Harvard University Press.

Clone this wiki locally