Management and Misallocation in Mexico

 
 
B.3.3 Other characteristics of the firm 
Drive time to the border (hrs): 
As previously mentioned in the data sources, we calculated drive 
time to the border in hours using the drive time between the centroid of the municipality in which 
each firm is located to any of the three most important border crossings between Mexico and the U.S.: 
Tijuana, El Paso, and Nuevo Laredo. Then, we calculated the minimum value of these three drive 
times and that is the indicator we use in our regressions.
 
Share of exports from the 2014 Economic Census:
As we want to measure the exposure of each 
sector to the external sector, we wanted to obtain a measure from a different source than ENAPROCE 
2015. Therefore, we calculated this using the 2014 Economic Census. We calculate this share using 
information at the NAICS 6-digits level and it is calculated as exports divided by total sales. 
Therefore, a 1 percentage point change in this variable is a change of 0.01.
Foreign Direct Investment (FDI): 
As shown in Table A1 (Descriptive Statistics), we construct two 
alternative measures of FDI, FDI1, which takes a value of one if the firm has any participation of 
FDI, and FDI2, which indicates 50% or more of FDI in the firm.
Capital-per-worker: 
Capital stock of structures and equipment (at present or replacement value) 
divided by the total number of employees in the firm. As we are only considering one year (2014), 
our measure does not face the problems of differences in accounting methods over the years. 
Share of white-collar workers: 
It is calculated as the ratio of white-collar workers over total 
workers. 
Share of workers with a college degree: 
It is calculated as the ratio of workers that obtained a 
college degree over total workers. 
Population density:
Population density at the Metropolitan Area (MA) level was obtained by 
aggregating municipality data on population and the extension of the municipality (in square 
kilometers) for the municipalities that compose each MA. 
Average income:
Average income is obtained as an MA level average of the household per capita 
income estimates at the municipality level using the Small Areas Estimates Methodology, which 
combines the 2010 Population and Housing Census with information from the 2010 National Survey 
of Household Income and Expenses (ENIGH). 
City size: 
We follow the definition from the US Census and classify cities according to their 
population. 
Small urban area=population less than or equal to 200,000 
Medium urban area=population higher than 200,000 and less than or equal to 500,000 
Metropolitan area= population higher than 500,000 and less than or equal to 1,000,000 
Large metropolitan area= population higher than 1,000,000 
It is important to note that the definition from the OECD is very similar to the one we use. 

---

viii 
Regions: 
Regions used as controls in this paper are based on the classification used by INEGI on its 
National Account System to summarize the trimester indicator of state economic activity. The regions 
are defined as follows: