Hello everybody,
I hope you are well.
To complete the empirical part of my thesis entitled "Agglomeration economies and concentration of industrial activities in Morocco", I must mobilize techniques of spatial econometrics through the calculation of concentration indices.
I currently have an Excel database that contains the company, its location, its sector and its workforce.
My question is how to transfer my database as a shapefile file on a map to be able to map the concentration of industries and calculate the concentration indexes (gini index, local and global moran index).
I would like to tell you that I intend to calculate the concentration indices of companies and their workforce.
Best regards.
http://siteresources.worldbank.org/INTWDRS/Resources/477365-1327525347307/8392086-1327528510568/WDR09_10_Ch04web.pdf
https://www.jyu.fi/ersa2003/cdrom/papers/361.pdf
http://xgzlez.webs.uvigo.es/research_archivos/AppliedEconomics.pdf
https://ageconsearch.umn.edu/bitstream/190899/2/WP01-010.pdf
Dear Rachid,
Following up on previous answers, I suggest that you read the following papers regarding different approaches and how to use STATA for estimating spatial econometric models:
[1] Arauzo-Carod, J-M. and Manjón-Antolín, M. (2012), (Optimal) spatial aggregation in the determinants of industrial location, Small Business Economics, 39, 645-658.
[2] Drukker, D., Prucha, I. and Raciborski, R. (2013), Maximum likelihood and generalized spatial two-stage least-squares estimators for a spatial-autoregressive model with spatial-autoregressive disturbances, The Stata Journal, 13, 2, 221, 241.
[3] Elhorst, J. Paul (2010), Applied Spatial Econometrics: Raising the bar, Spatial Econometric Analysis, 5:1, 9-28.
[4] LeSage J. and Pace K. (2011), Pitfalls in Higher Order Model Extensions of Basic Spatial Regression Methodology, The Review of Regional Studies, 41, 13-26.
[5] Pisati, M. (2001), Tools for spatial data analysis, Stata Technical Bulletin, 60, 21-37.
Pisati (2001) explains the stata procedures for the Moran's I and Geary's spatial autocorrelation tests.
Following the procedures described in Elhorst (2010) and LeSage and Pace (2011) is a good benchmkark, which goes from the simplest spatial models up to the SDM model, as stated in the answer above provided by Esteban. These models are usually estimated in STATA using the "spatreg" command (Pisati, 2001). From there, and depending on your analysis, it could be a good idea to dwell into more general models, such as the SARAR model (see Drukker et al., 2013). These are particularly useful if you run diagnostic tests that provide evidence of spatial dependence in the error term. In this case you are required to create a contiguity matrix using the "spmat" command, which then allows you to estimate the SARAR model using the "spreg" command. If the spatially lagged error term coefficient is statistically significant, the SARAR model generally provides a good fit. Finally, it is usually a good a idea to estimate a general model which includes spatially lagged explanatory variables, since this allows you to compare the SARAR model with, e.g., the SDM model. The Stata matrix environment "mata" can then be used to compute estimates of the marginal effects.
I hope this helps.
Sincerely,
José Gaspar
thank you all for your helpful answers and sorry for the late reaction.
- Badges
- Science topic
- Similar topics
- Indexes