7
from English without any modifications. For Services,
13
it was slightly adapted to reflect
differences in operations – for example, the question about “problems
in the production
process” was replaced by a question about “problems with customer service” (see Appendix
A for an English translation of the services questions). The reason to exactly replicate the US
MOPS was to maximize the comparability of Mexican management practices to the US
MOPS 2010 and 2015 samples. After data cleaning, we end up with a sample of 16,100 firms
with ten or more employees and 8,497 microenterprises with fewer than ten employees.
14
For every firm, a management score was constructed following Bloom et al. (2019). The
responses for each of the 16 questions included in the management section are normalized to
obtain a score that ranges between zero and one, where the more structured management
practices (the ones
that are more specific, formal, frequent, or explicit) are associated with a
value of one, while the less structured practices obtain a value of zero. The responses located
in between these two extreme values obtain a fraction depending on the number of categories.
Once every question is scored, we calculate an unweighted average of the 16 questions to
construct the management score. We also separate the overall management index into two
sub-indexes assessing monitoring practices and human resource practices (i.e., incentives and
targets). Our index of structured management practice is not necessarily “better” or “worse,”
although, as we shall see, they are strongly correlated with a wide variety of performance
measures.
For the case of microenterprises, only four questions from the US MOPS were included in
the shorter survey form. The correlation between the management
score calculated with this
subsample of questions and the overall management score for the non-microenterprises
sample is 0.86, indicating that this short score is a reasonable approximation for the overall
score. To reduce respondent burden, we did not ask questions on various other aspects such
as exports and FDI. Consequently, the main part of the analysis
drops microenterprises,
although we do show robustness to including them in the Online Appendix.
13
The definition of sectors (Manufacturing and Services) is shown in the Online Appendix (Appendix B).
14
See Online Appendix (Appendix B) for further details on the construction of the sample.