5
however, and even though it is likely to be correlated with fundamental capabilities,
6
this
relationship may be weak.
In this paper, we draw on the idea in Lucas (1978) defining
capability as management and measure it directly.
Second, we relate to the literature on the drivers of management practices. One strand of this
has focused on the role of competitive frictions. There are a large number of papers pointing
to the importance of competitive pressure for productivity (e.g., the survey by Holmes and
Schmitz, 2010). One of the mechanisms is that a positive
impact of competition on
productivity may be due to competition increasing management quality (see Leibenstein,
1966 and the survey by Van Reenen, 2011). Other strands of literature analyzing drivers of
management have focused more on information and learning spillovers (e.g., Cai and Szeidl,
2017) and regulations (e.g., Bloom et al., 2019).
Third, we contribute to the large and growing literature on the effects of management on
productivity. Early studies used cross-sectional or occasionally panel data (e.g., Black and
Lynch, 2001;
Capelli and Neumark, 2001; Huselid, 1995; and Osterman, 1994). These
studies tend to find positive associations in the cross-sections, but they tend to disappear in
the panel dimension. Another group of studies focuses
on smaller numbers of firms,
sometimes even looking across sites in a single firm.
7
These “insider econometric” studies
tend to find more positive effects of management practices. More recent studies have used
randomized control trials and (more rarely) quasi-experiments (Giorcelli, 2019), which also
tend to find positive results.
8
Bruhn, Karlan, and Schoar (2018) is particularly relevant as it
focuses on Mexico and finds that firms that received management consulting exhibited a
much stronger job growth over the medium term (2-5 years), though they find heterogeneity
in the practices that have a high impact on firm performance.
6
In the model of Hsieh and Klenow (2009) firm-level revenue based TFP measured (TFPR) are unrelated to
fundamental firm capability, quantity based TFP (TFPQ) due to the unobservability of firm level prices. In more
general models, however, which allow for fixed costs of labor or adjustment costs there
is
a correlation between
TFPQ and TFPR (Bartelsman, Haltiwanger and Scarpetta, 2013 or Asker, Collard-Wexler and De Loecker
2014).
Nonetheless, the empirical and theoretical issues with TFP measurement make more direct
measurements attractive.
7
For example, Bartel, Ichniowski and Shaw (2007); Bandiera et al. (2005, 2007); Hamilton et al. (2003) and
Lazear (2000). See Bandiera et al. (2020) for a study focusing on CEOs and performance.
8
Examples of RCTs include Anderson et al. (2018), Bloom et al. (2013), Brooks et al. (2018); Fryer (2017);
Gosnell et al. (2020), Higuchi et al. (2017, 2019); Karlan, Knight and Udry (2015); McKenzie and Woodruff
(2013). See McKenzie and Woodruff (2017) for a survey and Bandiera et al. (2017) for a meta-study.