Revista de economía mundial 69, 2025, 197-221
ISSN: 1576-0162
DOI: http://dx.doi.org/10.33776/rem.v0i69.8240
Market Power and InvestMent In advanced econoMIes:
MonoPolIstIc and ‘segMented’ coMPetItIon
Poder de mercado e inversión emPresarial en economías
avanzadas: comPetencia monoPolística y comPetencia ‘segmentada’
Agustin Pedrazzoli
agustin.pedrazzoli@uam.es
Departamento de Estructura Económica y Economía del Desarrollo
Universidad Autónoma de Madrid
Ignacio Álvarez
nacho.alvarez@uam.es
Departamento de Estructura Económica y Economía del Desarrollo
Universidad Autónoma de Madrid
Santos M. Ruesga Benito
ruesga@uam.es
Departamento de Estructura Económica y Economía del Desarrollo
Universidad Autónoma de Madrid
Recibido: abril 2024; aceptado: agosto 2024
abstract
Recent studies have related market power in advanced economies to
stagnating business dynamism and investment. This article addresses this
issue at the firm level, using both markups and market shares as measures of
market power. While our results show that market shares have a negative effect
on investment, markups generally have a positive effect, which only becomes
negative for very high markups. The negative effect of high markups on
investment should not, however, be attributed to market leaders, but to smaller,
niche firms. Thus, the current state of market power in advanced economies
is one where “superstar firms” and niche firms or “segmented competition”
coexist, and together put downward pressure on investment rates.
Keywords: Market power, markups, investment behaviour, competition,
firm-level data.
resuMen
Estudios recientes han relacionado el creciente poder de mercado en las
economías avanzadas con el estancamiento de la inversión y el dinamismo.
Este artículo analiza esta relación a nivel microeconométrico, empleando la
cuota de mercado y el markup como medidas de poder de mercado a nivel
empresa. Nuestros resultados muestran que las cuotas de mercado tienen
un efecto negativo sobre la tasa de inversión, mientras que el efecto de
los markups es ambiguo, siendo positivo inicialmente, y solo negativo para
valores elevados del markup. Estos altos markups, sin embargo, se encuentran
típicamente en empresas de menor tamaño. En consecuencia, el estado actual
del poder de mercado en economías avanzadas se caracteriza por la presencia
de empresas ‘nicho’ o competencia segmentada que, junto con la existencia
de empresas pseudo-monopolistas, ejerce también una presión a la baja en la
tasa de inversión.
Palabras clave: Poder de mercado, markups, inversión empresarial,
competencia, microdatos.
JEL Classification/ Clasificación JEL: L10; L13; E22
Revista de economía mundial 69, 2025, 197-221
1. IntroductIon
The rise of market power in advanced economies has received much
attention in recent years. Different studies have documented strong increases
in market concentration rates (Gutierrez and Philippon, 2017; Grullon et al.,
2019; Autor et al., 2020), price markups (Diez et al., 2018; De Loecker et
al., 2020; Diez et al., 2021) and monopolistic rents (Gutierrez and Philippon,
2016; Brun and Gonzalez, 2017; Stiglitz, 2017; Eggertsson et al., 2021) over
the past decades, suggesting that markets have become less competitive and
business dynamism more stagnant. The standard narrative in most of these
studies is that higher market power increases monopolistic profit rates, reduces
investment and damages consumer surplus and labour shares.
One of the main limitations in previous studies is that they look at macro
or meso-level relations between market power and investment aggregates
(or at best they model firm-level behaviour that is consistent with aggregate
trends). Most of them do not empirically study how market power affects
investment behaviour at the firm-level, and generally assume the underlying
firm-level dynamics that would explain aggregate trends. In fact, to the best
of our knowledge, only Diez et al. (2018) estimate the effect of market power
on investment behaviour at the firm-level, and find that markups can have
a positive effect on investment which only becomes negative for very high
markups. While aggregate measures of market power may offer insight on
overall trends in advanced economies, a more granular look into sector and
firm-level dynamics is required to understand how the rise in market power
may affect investment overall.
This article takes a step in this direction. We study the relation between
market power and investment behaviour at the firm-level using data for over
13,000 firms in OECD countries, from 2012 to 2020. Our large dataset and
geographical coverage allows us to identify a generalised effect of market
power on investment in advanced economies, and our recent time window
helps characterise the current state of competition and market power (as
compared to studies that focus on how market power has changed in the past
decades). To do so, we consider two different indicators of market power at the
firm level: markups and market shares.
Our study presents an interesting result, which is that markups and market
shares are not generally positively related. Very high markups are more often
found in smaller firms. These smaller, high-markup firms, furthermore, appear
200 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
to play an important role in the overall rise in markups which the literature has
taken as evidence of declining competition. Thus, competition in advanced
economies today appears to be characterised not only by the presence of
large superstar firms, but also by the emergence of some form of ‘segmented
competition’ by which smaller firms manage to exert increasing market power
on niche segments of demand where they exhibit higher markups.
Another important result in our study is that markups, on their own,
do not seem to be a comprehensive measure of market power. If we were
to only consider markups as measures of market power, this concentration
of higher markups in smaller firms that would imply, contrary to standard
economic intuition, that larger firms tend to have less market power in general.
Instead, looking at the relation between markups and market shares reveals an
interesting trade-off by which firms with higher market shares tend to charge
moderate (not higher) markups. Again, very high markups are typically found
in firms with low market shares which seemingly represent niche firms that
cannot upscale their activity without lowering markups (Keil, 2017; 2019).
In consequence, since markups and market shares each reflect important
dimensions of market power, both need to be considered to have a full
understanding of market power at the firm level.
In relation to investment, our estimation shows that market power (market
shares and markups) does have a negative effect on investment, although in
the case of markups, this effect is found for very high markups only, as in
Diez et al. (2018). For most firms, with more moderate markups, it seems that
markups reflect growth opportunities or ‘post-investment rents’ (Aghion et al.,
2005) that stimulate investment behaviour. Given the negative effect of market
shares on investment, these ‘post-investment rents’ seem to wear off as firms
gain market, but also as they manage to target non-generalizable niche markets
with lower growth perspectives but a higher degree of market power relative to
consumers. The generalisation of this form of ‘segmented competition’ could
therefore contribute to putting downward pressure on aggregate investment
levels.
This article is divided into seven sections. The following section reviews the
main findings in the literature on the rise of market power and its effects on
investment behaviour. Section three presents the data used for our analysis
and stylised facts in relation to markups and market shares. Section four
presents our empirical model and quantitative method used to estimate the
effect of market power on firm-level investment, and section five presents the
results of our estimation. In section six we discuss how these findings relate to
those of previous studies, and how they modify our interpretation of the rise of
market power in advanced economies, and section seven concludes.
2. InvestMent and Market Power
Our work is part of the vast literature that tries to explain the relationship
between market competition, investment and corporate growth. This research
201
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
agenda can be traced back to Bain (1956) and his “structure- conduct-
performance” (SCP) paradigm. The SCP model presents a framework for
empirical analysis that examines the effect of market structure on the evolution
of a given industry and on the performance of firms in that industry. According
to this framework, the structural characteristics of a certain market determine
the behavior of firms within that market and, therefore, their performance in
terms of innovation, investment, profits and other similar variables. In other
words, the SCP model assumes that there is a direct link between market
structure, firm behavior and performance. Later on, the New Industrial
Organization approach (Tirole, 1988) expanded and revised this theoretical
framework, introducing new insights when explaining firm behavior.
Within this literature, a large number of theoretical and empirical studies
have been developed analyzing, among other issues, the relationship between
market structure and firm strategy (Shubik and Levitan, 1980), the behavior
of firms in imperfectly competitive industries (Porter, 1981), the link between
industry concentration scenarios with market outcomes (Bresnahan and
Schmalensee, 1987) and the impact of market structure on competition
(Goolsbee and Syverson, 2008), innovation (Rafique-Hashmi and Van
Biesebroeck) or investment strategies (Cohen and Mazzeo, 2010). Specifically,
this literature on industrial organization considers that there are structural
reasons (economic, strategic and technological) that push firms to grow (Tirole,
1988): developing economies of scale, risk diversification, better access to
both capital and external sources of financing, reduced transaction costs or
greater market power.
The business strategy literature has also examined the motives that
explain corporate investment and growth. In particular, the “Resource-Based
View of the firm” is an approach that, building on Penrose’s (1959) work, has
been used to analyze the strategic resources a firm can exploit to achieve
sustainable competitive advantage. This literature has emphasized that
corporate growth and investment should be studied in terms of both internal
and external determinants, with the understanding that firms grow due to a
combination of factors: the efficient use of underutilized internal resources that
can be leveraged for growth, the accumulation of knowledge and experience
(“economies of learning”), the vision and ambition of the firm’s managers,
efficient management and external market conditions (increasing demand
for the firm’s products or services and access to external resources, including
finance, technology and strategic partnerships).
In this literature on industrial organization and corporate growth we find
different approaches to the relationship between competition and corporate
investment. For example, some authors point out that less competition in
markets reduces business investment due to a combination of different factors
(Porter, 1980): the lack of competition reduces the need to diversify and
expand into other markets and, in addition, companies in less competitive
markets tend to have higher returns without the need to reinvest their profits,
which reduces the pressure to innovate and the incentives to improve efficiency.
202 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
In the opposite direction, other authors (Tirole, 1988) point out that in
highly competitive markets, firms may be less willing to invest large sums due to
uncertainty about the payback of those investments and lower profit margins.
Moreover, firms already established in an industry, and with a dominant
position, may use investment as a strategic tool to create barriers and deter
the entry of new competitors.
All this literature on industrial organization and corporate growth provides
a theoretical and empirical framework of undoubted value for analyzing, from
a microeconomic point of view, the determinants that explain the behavior of
firms and, particularly, their investment decisions. According to this literature,
the relationship between competition and investment is complex and depends
on multiple factors, so that competition can either encourage or discourage
investment. Nevertheless, this framework must necessarily be complemented
by other literature −with a macroeconomic focus− when investigating the link
between market power and investment rates in advanced economies.
Investment rates in advanced economies have been slowing down over the
past decades (Gordon, 2012; Baldwin and Teulings, 2014; Summers, 2014;
FIgure 1. InvestMent rates In Major advanced econoMIes (Percentage oF gdP)
Source: IMF, World Economic Outlook data.
203
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
Gutierrez and Philippon, 2016; Grullon et al., 2019; Eggertsson et al., 2021).
In Figure 1, we see that Gross Capital Formation (as a percentage of GDP) has
fallen in most of the major advanced economies, typically presenting lower
levels in the last decade than those observed in the 1980s and 1990s.
While this is a well-established fact, the causes for this slowdown are less
clear. Different suggestions have been made, including structural headwinds
(Gordon, 2012), a lack of profitable investment outlets (Summers, 2014),
financialization of corporate management (Stockhammer, 2004; Davis,
2018; Tori and Onaran, 2020) or financial sector hypertrophy (Cecchetti and
Kharroubi, 2015).
The downward trend of the investment rate in the main advanced economies
has also been understood in the field of post-Keynesian economics as the result
of unexpected changes in demand (and thus possibly the result of equally
unexpected changes in the structure of the economy). In particular, from the
Sraffian supermultiplier approach (Serrano, 1995), it has been argued that a
lower investment rate may be the result of a lower growth rate of demand and,
particularly, a lower growth rate of the autonomous components of demand
(those expenditures that are neither financed by contractual wage income nor
can create capacity). Lower growth in autonomous demand would imply less
induced (capacity-creating) investment through the accelerator effect.
However, another important change that has taken place during these
decades has also been identified by the literature as a key factor: the rise in
market power, registered through both the increase in markups and the rise in
market concentration rates.
Gutierrez and Philippon (2017) find evidence of decreasing competition
and higher average price markups, and argue that rising concentration rates
can largely account for the investment gap (low investment despite high
Q-ratios) observed in the United States during the past decades. Grullon et
al. (2019) also find evidence of increasing market concentration rates among
listed firms in the United States since the 1980s and agree that market power
has become an important source of value for listed firms. In fact, Eggertsson et
al. (2021) argue that only by the combination of high market power and lower
natural interest rates can the observed investment slowdown be explained
when cost of funding has been historically low and profit rates historically high.
Another set of studies also show evidence of rising market power by looking
at the evolution of average price markups (De Loecker and Eeckhout, 2018;
Diez et al., 2018; De Loecker et al., 2020; Diez et al., 2021). Following a method
for markup estimation developed by De Loecker and Warzynski (2012), these
studies find a strong increase in price markups over the past decades in both
advanced and emerging economies. Since markups are observed at the firm
level, in contrast to market concentration rates, these studies can also look
into the distribution of markups across firms to identify where market power
is growing. De Loecker et al. (2020) and Diez et al. (2021) find that the rise in
average markups has been mainly driven by high-markup firms growing faster
204 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
than other firms and by a greater dispersion in the distribution of markups
(more firms enjoying increasingly higher markups).
Other studies, however, have called into question the standard narrative
found in these studies, arguing that higher market power is not necessarily
responsible for lower investment rates, and could even potentially contribute to
stimulating investment levels. Autor et al. (2020) argue that the rise in aggregate
markups could, in fact, be consistent with the emergence of more efficient,
highly innovative ‘superstar’ firms, which would actually bring investment rates
up. Diez et al. (2018) estimate a panel data regression model using firm-level
data from listed firms of 74 countries, from 1980 to 2016, and find that
higher markups do not have a consistently negative effect on investment. Their
study finds a non-monotonic inverted-u shaped relation between markups and
investment at the firm level, similar to the relation between competition and
innovation found in Aghion et al. (2005). According to their results, increases
in market power initially have a positive effect on investment (‘post-investment
rents’). Only as markups reach higher values does the effect wear off, eventually
becoming negative, as most of the literature expects.
Looking at market power more broadly, that is, taking both markups and
market concentration into consideration, Davis and Orhangazi (2021) also
find an ambiguous relation between market power and investment. Focusing
on the industry level relation between market concentration rates, markups,
investment behaviour and profitability in the United States, the authors find
that more concentrated markets (typically considered to be less competitive)
actually tend to have higher investment levels, and do not always present
higher markups. In fact, the authors find cases of high market concentration
rates with low markups, low profit rates and average investment rates, and
yet other cases in which firms enjoy high markups in low concentration
markets (indicating some form of market power that is not related to market
concentration) with no clear relation to investment.
In summary, the relation between the rise in market power and the
observed investment slowdown in advanced economies is not clear, and the
empirical evidence is somewhat inconclusive. It should be noted that most of
these studies are done at the aggregate (macro or meso) level, without looking
directly into underlying firm-level dynamics that explain how firms manage to
obtain higher levels of market power, and how that affects their competition-
related incentives to invest. We believe that studying these dynamics empirically
and at the firm level, can offer important insights for understanding the relation
between the rise in market power and the aggregate investment slowdown.
3. MarkuPs and Market shares
3.1. data
To study these relations at the firm level we use annual financial data
obtained from the ORBIS database for listed firms of 35 OECD countries
205
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
during the years 2012-20201. Financial statements in ORBIS have been
standardised and are comparable across jurisdictions and, while the availability
of historical data in ORBIS is lower than in other datasets such as Compustat,
its geographical coverage is much larger and has been used for this reason
in other studies concerned with multi-country firm-level analyses (Gal, 2013;
Gopinath et al., 2017; Diez et al., 2021).
For our study, we focus on nonfinancial corporations2 that present
consolidated financial statements. Financial data are deflated using the GDP
deflator. Additionally, firm-level data typically require some treatment in
order to eliminate anomalous or irrelevant observations. Since we are only
considering listed firms, we eliminate any observations for years in which the
firm was not yet listed or had already been delisted. We also drop observations
that represent a large jump in firm assets or sales3 since these are typically
residual observations representing a firm’s starting years or last years. Firms
with zero or negative profits, sales or capital stock throughout the entire
observation window are also eliminated. Finally, we remove observations in
the top and bottom 1 per cent of the distribution for each variable. As a result,
we end up with a panel of over 100,000 observations for 13,000 firms from
2012-2020.
Price markups and investment rates can be calculated directly for each firm-
year observation in our sample. We estimate markups as profit margin (sales
revenue minus costs of goods sold) relative to costs of goods sold. Though
many studies have used the method for estimating markups developed by
De Loecker and Warzynski (2012), this method has recently been called into
question (Bond et al., 2021; Doraszelski and Jaumandreu, 2021). Doraszelski
and Jaumandreu (2021) argue that the estimation of output elasticities
required to calculate these markups are not robust to differences in demand
across firms or time, thus leading to biased estimated of price markups.
Additionally, the entire estimation procedure relies on cost minimisation of a
representative production function, which requires the strong assumption of
common production technologies across firms within the same market.
While our markup is not an exact measure of price relative to marginal
cost, similar measures are commonly used in the literature as approximations
given the lack of more detailed information (Aghion et al., 2005; Gutierrez
and Philippon, 2017; Grullon et al., 2019; Davis and Orhangazi, 2021). More
importantly, as we will show later on, our stylised facts obtained using profit
margin relative to costs of goods sold are largely in line with those found in
previous studies using the De Loecker and Warzynski method (De Loecker et
al., 2020; Diez et al., 2021).
1 Given the relatively short observation window available in ORBIS at the time we extracted our
data, we include year 2020 in our sample, despite the pandemic, to use all the available historical
observations.
2 Firms in Financial Services (NAICS code 52) and Real Estate (NAICS code 53) are not considered.
3 Following Bloom et al. (2004) and Tori and Onaran (2020).
206 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
Estimations of market shares present a different set of complications since
direct calculation of market shares using firm-level data are typically unreliable
in markets where listed firms only represent a small portion of the market (Diez
et al., 2018; Grullon et al., 2019; Davis and Orhangazi, 2021). To mitigate this
effect we calculate market shares of listed firms by considering sales of both
listed and non-listed firms available in ORBIS. However, even after including
non-listed firms, many markets still present unreasonably high concentration
rates which is probably due to incomplete data collection rather than reflecting
true underlying market structures (see Table 2 in the Appendix). To further
avoid possible biases introduced by these firms, we calculate pseudo market
shares for a given year considering only the largest 20 firms in markets with at
least 20 firms4. In consequence, the number of market-share observations is
reduced to roughly one third of our initial sample (see Table 3 in the Appendix
for summary statistics).
3.2. MarkuPs
As we show in Figure 2, both the median markup and the sales-weighted
average markup in our sample appear to have risen steadily during in the past
decade. The sales-weighted average markup can be affected by composition
effects such as high-markup firms becoming larger or large firms increasing
their markups (De Loecker et al., 2020). In our case, we are more interested in
a generalised rise in markups that could have a significant effect on firm-level
investment behaviour overall, and not just for a specific subsample of firms.
For this reason, the rise in the median markup is a more significant trend for
us, since it shows that market power has increased for most firms in advanced
economies.
Previous studies have typically associated these rising markups to the
predominance of pseudo-monopolists, large multinationals or ‘superstar’ firms
(Gutierrez and Philippon, 2017; Autor et al., 2020; De Loecker et al., 2020)
since they find rising markups to be related to greater market concentration
or weaker competition. The standard reasoning is that higher markups can
only be enjoyed if firms manage to beat competition and gain higher market
shares. In fact, we typically expect firms to turn any competitive advantage
they may have into higher market shares first, in order to ease off their survival
constraint, and only later would they be able to start raising prices without fear
of losing their dominant position.
Figure 3 shows the evolution of concentration rates5 and median markups
for major advanced economies. Since our estimation sample contains more
than 30 countries, we chose to present these trends only for the most relevant
4 The United States Census Bureau calculates industry-wide HHIs using the top 50 firms in each
industry. This criterion, however, would have significantly reduced the number of markets in our
sample.
5 Concentration rates are calculated as the Herfindahl-Hirschmann Index, using our approximate
measure of market shares.
207
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
economies. We can see that markups show a growing trend in most cases, but
contrary to what is typically expected, we do not find a clear relation between
markups and concentration rates. This is better accounted for in the next
section where we look at the firm-level relation between markups and market
shares.
3.3. Market shares
As explained above, to obtain more reliable estimates of market shares
we calculate the share of total sales among the top 20 firms in each market6.
Figure 4 shows the relation between markups and market shares at the firm
level for all firms in our sample over the period 2012-2020. This figure reveals
an interesting pattern. First of all, we typically find higher markups in firms with
low market shares. Only in a few exceptional cases do firms with high market
shares manage to enjoy very high markups (these seem to be cases of pseudo-
monopolists or ‘superstar’ firms).
However, the relation between market share and markups is not directly
negative. Instead, what we find is that, as firms gain market shares, their
markups tend to gravitate towards more moderate values. In other words,
there appears to be a certain trade-off by which firms that intend to reach
high market shares either need to keep prices in line with competition (Shaikh,
6 Even considering this restricted sample, many firms have a small (or very small) market share after
applying this threshold, as shown in Figure 4.
FIgure 2. evolutIon oF average MarkuPs
Source: Own elaboration using data from ORBIS.
208 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
2016; Kiel, 2017; 2019) or face diminishing returns to scale, and where very
high markups can only be enjoyed by smaller ‘niche’ firms that find it difficult
to upscale their activity and at the same time enjoy a high degree of market
power on consumers (Diez et al., 2021).
In any case, this nonlinear relation implies that market shares and markups
are not necessarily interchangeable measures of market power. Davis and
Orhangazi (2021) argue that market concentration rates can be insufficient
FIgure 3. evolutIon oF Market Power In Major advanced econoMIes
Source: Authors’ own elaboration using data from ORBIS
209
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
measures of market power if there are high-markup firms in low-concentration
markets. Similarly, if market shares or market dominating positions are also an
expression of some form of market power (as most of the literature seems to
accept), then market power cannot be fully accounted for by only considering
markups. To analyse the effect of market power on investment behaviour at the
firm level we should, therefore, take both dimensions into account.
4. eMPIrIcal Model and estIMatIon Method
In order to see how each of these dimensions of market power affects
investment behaviour, we estimate a firm-level investment function. The
determinants of firm-level investment have been the subject of extensive debate
in economic literature (Jorgenson, 1971; Abel and Blanchard, 1986; Blundell
et al., 1992; Chirinko, 1993; Fazzari and Petersen, 1993; Carruth et al., 2000;
Brauman and Kopcke, 2001), and there is a general consensus that profitability,
demand and cost of funding are important determinants of investment behaviour.
In consequence we estimate the following regression model:
(1)
FIgure 4. Market shares and MarkuPs (observatIons For the PerIod 2012-2020)
Source: Own elaboration using data from ORBIS.
210 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
Where I/K represents firm i’s investment expenditure relative to capital
stock (property plant and equipment) at time t, Q represents the Q-ratio
(calculated as stock market capitalisation relative to total assets) as a measure
of firm profitability, S/K represents sales volume relative to capital stock as
a measure of demand pull and D/K represents the debt to capital ratio as a
measure of financial fragility and cost of external funding. To this equation
we add two measures of firm-level market power: markup represents price
markups, defined as profit margin relative to costs of goods sold and market
share represents share of total sales volume among the largest 20 firms in a
given sector and country.
We include a lag of the dependent variable, since investment is better
specified as a dynamic process containing an autoregressive component and
a set of exogenous variables (Bond and Meghir, 1994; Brauman and Kopcke,
2001; Tori and Onaran, 2020). All other regressors are also lagged since a
firm’s investment decisions at time t is expected to be made in the previous
period or on the basis of information available to the firm at the start of the
current period (Abel and Blanchard, 1986; Orhangazi, 2008). Additionally,
since we are working with panel data, we include time dummies ( ) to control
for time fixed effects.
To avoid any potential individual fixed effects (ηi) from biasing our estimated
coefficients, we estimate the coefficients in our model using the system-GMM
estimator with instrumental variables as in Arellano and Bover (1995) and
Blundell and Bond (1998). Sector and country dummies are not included
since they are dropped from the first-difference equation and are collinear to
individual fixed effects in the level equation. Instead, we use country-sector
cluster-robust standard errors7 to control for cross-sectional correlation and
heteroscedasticity in the error term. Finally, observations are also weighted
by country-sector so that coefficient estimates do not mainly reflect empirical
relations observed in countries or sectors with more observations (Love, 2003;
Cameron and Trivedi, 2010).
In line with previous studies on investment behaviour we expect the Q-ratio
to be positively related to investment, since profitability will lead to high returns
on investment and facilitate external funding (Blundell et al., 1992; Bond and
Meghir, 1994). We also expect the sales-to-capital ratio to be positively related
to investment, since sales largely represents the strength of demand for a firm’s
goods or services and this will typically lead to greater investment through
a sales-accelerator mechanism (Brauman and Kopcke, 2001; Orhangazi,
2008; Tori and Onaran, 2020). The debt-to-capital ratio is expected to have
a negative effect on investment, since higher indebtedness will signal financial
fragility and increase external funding costs (Fazzari and Petersen, 1993;
Nikolaidi, 2014; Tori and Onaran, 2020). However, indebtedness could also
7 The use of both countries and 2-digit NAICS codes was necessary to ensure a sufficiently large
number of clusters following Roodman (2009).
211
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
have a positive effect on investment if firms mainly rely on external funds to
finance investment, and the final effect could be ambiguous.
As for our measure of market power (market shares and markups), we
expect them both to have a negative effect on investment. Generally speaking,
a higher degree of market power (higher market shares) is expected to reduce
incentives to invest since firms face lower competitive pressures. Additionally,
in line with most of the literature (Gutierrez and Philippon, 2017; Grullon et
al., 2019; De Loecker et al., 2020) we expect higher markups to also reflect
lower competition, and therefore, have a negative effect on investment, even
if the effect may initially be positive (Diez et al., 2018). Importantly, since
market shares and markups are not directly related (as we saw in section
3), we expect each variable to reflect the effect of different forms of market
power on investment behaviour (high market shares indicating market
dominant positions, and high markups mainly reflecting niche firms in forms of
‘segmented competition’).
5. estIMatIon results
Table 1 presents the results of our estimation. In addition to the base
model containing both markups and market shares (column I), we estimate the
model again including only markups (column II) or market shares (column III).
Furthermore, since using market shares greatly reduces our estimation sample
by restricting the sample to the largest 20 firms in each market, we also explore
the effect of markups on the full, unrestricted sample to take advantage of the
whole set of observations. In this case, we estimate two models: one using
only markups and another using markups and squared markups to test for any
nonlinear effects such as those found in Diez et al. (2018). Non-linear effects
of both market shares and markups are also explored on the restricted sample
and results are presented in Appendix Table 4.
The p-values of the Hansen test in all three models show no sign of
overidentification8 and the autocorrelation tests also show no sign of
autoregressive behaviour in the error term. Beginning with the traditional
determinants of firm-level investment, we find the lagged term of investment
rate to be positive and significant, as expected, in all models. However, while
the Q-ratio, the sales-to-capital ratio and the debt-to-capital ratio present the
expected signs in most cases, only the coefficient for the Q-ratio seems to
be statistically significant. Considering that our sample consists of firms from
many different sectors and countries where demand elasticities of investment
may vary and where financial fragilities related to costs of funding may appear
at different levels of indebtedness, this is not entirely surprising.
8 Instruments were limited to two lags for the first-difference equation to avoid instrument
proliferation.
212 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
table 1. estIMatIon results
Variables I II III IV V
Markups and
shares Market shares Markups Markups
(full sample)
Markups
squared
(full sample)
(I/K)-1 0.2032*** 0.1961*** 0.2448*** 0.2289*** 0.2262***
(0.0585) (0.0584) (0.0553) (0.0391) (0.0380)
Q-1 3.8690* 2.5221** 6.2210*** 5.5805*** 5.2002***
(2.1093) (1.1136) (1.6075) (1.4103) (1.3303)
(S/K)-1 0.0127 0.0378 0.0599 0.0878 0.0872
(0.0803) (0.0421) (0.0748) (0.0927) (0.0979)
(D/K)-1 0.1472 0.0714 0.1018 -0.0006 -0.0062
(0.1202) (0.0769) (0.1137) (0.0698) (0.0839)
Market share-1 -0.7321** -0.6442
(0.3057) (0.3998)
Markup-1 1.1977* 0.7197 0.1298 1.5905***
(0.6482) (0.6772) (0.2384) (0.5786)
Square Markup-1 -0.0524***
(0.0182)
Constant 16.2583*** 19.3911*** 7.7158*** 10.1694*** 8.9831***
(3.6214) (4.0311) (2.0531) (2.1905) (2.1595)
Observations 18,111 19,021 18,111 65,940 65,940
Number of id 3,300 3,419 3,300 10,496 10,496
Cluster variable country-sector country-sector country-sector country-sector country-sector
Num. Clusters 147 149 147 423 423
Year FE YES YES YES YES YES
Num. Instru-
ments 127 107 107 107 127
Hansen test 0.380 0.223 0.101 0.197 0.310
AR1 0.000 1.55e-08 4.55e-10 0.000 0.000
AR2 0.725 0.830 0.710 0.200 0.247
Wald Chi-square 1082 1248 1845 2864 2856
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
If we look at column I, where we estimate the full model using model market
shares and markups, we find that both these variables also have a significant
effect, indicating that market power plays an important role in firm-level
investment. Interestingly, however, only the effect of market share is negative
(reducing investment by 0.73 percentage points for every one-point increase
in market share). In contrast, markups are estimated to have a positive effect
on investment (increasing investment by 0.12 point per each 10-percentage-
point increase in markups), even after controlling for demand pressure (S/K)
and profitability (Q). These effects are also found in columns II and III, where we
estimate the model using each measure of market power separately. However,
213
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
while market shares continue to have a negative effect (column II) and markups
a positive effect (column III), these effects are only statistically significant when
we control for both measures of market power.
These results highlight the different aspects or dimensions of market power
captured by each variable. When firms are able to charge higher markups, this
leads to stronger investment levels, possibly because they wish to upscale their
activity in order to obtain a stronger position within their market or because
their higher markups are the result of efficiency gains related to innovation.
However, and as we saw in the second section, when firms reach dominant
positions in their market (represented by higher market shares), their level of
investment weakens since additional gains from upscaling activity or innovating
are likely to become smaller, and the lack of competition reduces the pressure
to innovate, the incentives to improve firm’ efficiency and the need to diversify
and expand into other markets (Porter, 1980; Gutiérrez and Philippon, 2016
and 2017; Grullon et al., 2019).
Furthermore, the opposite effect of markups and market shares, together
with their lack of significance in columns II and III, is not surprising if we consider
the relation between these two variables found in section 3. While markups and
market shares are inversely related in the extremes, firms with low markups
can have high or low market shares and firms with low market shares can have
high or low markups, leading to an ambiguous relation between market power
and investment if we do not control for both measures of market power. In
terms of the overall effect of market power on investment behavior that we find
in the results shown in column I, we can consider different situations. Firms with
high market shares tend to have moderate markups, such that the negative
effect of market shares on investment is likely to dominate. Similarly, firms with
high markups tend to have lower market shares, such that the positive effect of
markups on investment is likely to dominate. In the case of firms with low (high)
markups and low (high) market shares the dominating effect is ambiguous,
but, in any case, higher markups will generally have a positive effect for similar
levels of market shares (indicating higher profitability of investment), and
market shares will generally have a negative effect for similar levels of markups
(indicating lower growth opportunities).
Since the results in column I are limited to a subsample of larger markets
(given that we cannot measure market shares for firms in markets with few
observations), we also explore the relation between markups and investment
behavior considering the entire observation sample (columns IV and V). As in
Diez et al. (2018), the relation between markups and investment appears to
be non-monotonous and inverse-u shaped. When the squared markup term is
omitted (column IV), markups again have a positive but not significant effect
on investment. However, once the squared term is included (column V), we find
that markups have a positive and significant effect on investment that becomes
negative at high markup levels as a result of the negative effect found in the
squared markup term.
214 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
As we mentioned earlier, Diez et al. (2018) explain these results by
comparing them to the inverse-u shape relation between competition and
innovation found in Aghion et al. (2005). Initially, firms in more competitive
markets (lower markups) have strong incentives to invest (innovate) in order
to escape competition. By investing/innovating firms become more efficient,
leading to higher markups and creating a positive feedback mechanism between
investment and markups. Only as firms reach higher levels of market power do
these returns from investment/innovation begin to wear off, eventually leading
to lower levels of investment.
The problem with this explanation is that it, again, identifies higher
markups with higher market shares. However, this is not consistent with our
evidence on the relation between markups and market shares in large markets
(as we saw in section 3), where market leaders do not typically have very high
markups. If high markups represent firms escaping competition, our data seem
to suggest that this is done by successful market segmentation rather than by
gaining market share. In fact, beating competition to greater market shares
apparently involves a trade-off by which firms lower markups to keep in line
with competition rather than charging increasingly higher markups.
Table 4 in the Appendix shows the estimation of non-linear effects on the
restricted sample (used in columns I-III of Table 1), where we find the squared
terms of market shares and markups to not be significant. Thus, when we
remove smaller (niche) firms from large markets we find that the non-linearity
of markups is no longer significant. In consequence, we believe the negative
effect found in column V for very high markups (the squared markup term) is
more likely to be related to the fact that niche firms do not expect high returns
from growing their activity and therefore keep investment levels low. If ‘post-
investment rents’ represented by markups wear off as firms gain market share,
as suggested by Aghion et al. (2005), this would more generally appear to
involve a movement towards the left-hand side of the inverted-u shape, as a
result of the apparent trade-off between market shares and markups.
For example, if a small firm with high markups is an innovator, it is likely to
gain market share but lower its markup in the process. The negative effect of
a higher market share (which we see in column I), together with a decreasing
markup, will eventually bring investment levels down. If a small firm with high
markups is instead a niche firm, it is likely to continue targeting specific market
segments and attempt to raise markups while lowering investment (resulting in
the negative effect on the squared markup term found in column V).
6. dIscussIon
Our estimation results largely describe two main scenarios where rising
market power can lead to lower investment levels. The first scenario is related
to rising concentration rates. Higher market shares are found to have a negative
effect on investment behaviour and, therefore, greater concentration rates
will likely lead to lower investment levels. However, contrary to the standard
215
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
narrative, we find that these high market shares do not typically lead to very
high markups, but entail a trade-off between markups and market shares that
seems more in line with dynamics described in Shaikh (2016)’s theory of real
competition than with the notion of pseudo-monopolists or ‘superstar’ firms
(Autor et al., 2020; Diez et al., 2021).
The second scenario can thus be characterised as one where a larger
number of niche firms manages to charge very high markups on well-defined
market segments while not aspiring to become industry leaders. Very high
markups are also found to have a negative effect on investment behaviour and,
therefore, in a scenario like this, the rise in high-markup firms would also lower
aggregate investment rates.
Interestingly, if the rise in aggregate markups over the past decades has
mainly been driven by high-markup segments, as shown in Diez et al. (2021), it
would seem that the nature of market power in advanced economies is better
described both by a proliferation of niche firms that act as monopolists in
their specific market segments together with a growing accumulation of market
power and surplus extraction in the hands of large corporate behemoths. We
would thus be witnessing the development of a kind of ‘segmented competition’
with growing market power for firms in more narrowly defined product markets
or niches.
Given the developments in information technologies, big-data, advertising
techniques and core-competencies-oriented corporate strategy (which would
greatly favour market segmentation strategies), together with the fact that
globalisation has increased the degree of competition faced by large market
dominant firms, this outcome is not entirely surprising. In fact, it is possible
that both of these scenarios are actually interdependent. Market segmentation
can be a valid survival strategy for market laggards in more concentrated
markets where competing for higher market shares via price undercuts is more
complicated, or where only market dominant firms have the financial capacity
to carry out efficiency-improving innovations and will carry out pre-emptive
mergers to defend their market shares (Bryce and Dyer, 2007).
While both scenarios would entail an overall negative effect on investment
behaviour, correctly identifying how market power is being exerted is relevant
in terms of its effects on other aspects of economic performance (such as
productivity growth, innovation, consumer surplus or income distribution) and
to accurately design pro-competition policy measures. More importantly, by
only looking at the evolution of markups, it is not possible to determine how
firms are reacting to changes in market structure and competitive pressures
and how these may be driving rises in markups.
7. conclusIons
In this article we look at market power and investment at the firm level. Our
study offers two main results. First, we find that high markups are generally
found in smaller and not larger firms. While this does not rule out the existence
216 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
of ‘superstar’ firms or pseudo-monopolists that may enjoy a strong market
dominant position, our data show that higher markups are more common in
firms with lower market shares which could either be emergent innovators or
niche firms that cannot upscale their activity and at the same time retain these
high markups. Market leaders, in contrast, tend to gravitate towards moderate
markups, and thus, in order to accurately measure market power (at the micro
and macro level) both dimensions –markups and market share– should be
taken into consideration.
Secondly, while higher market shares are estimated to negatively affect
investment behaviour, we only find a negative effect on investment for
markups in the case of firms with very high markup values, as in Diez et al.
(2018). However, given the observed trade-off between markups and market
shares, the negative effect of very high markups on investment we believe can
better be explained by niche firms that have high market power (markups) in
specific market segments and low incentives for investment given their less
generalizable business models, rather than by market leaders whose post-
investment rents have worn off (Aghion et al., 2005; Diez et al., 2018).
Interestingly, if the aggregate rise in market power observed in other
studies has been driven mainly by these higher markup segments of the
population (De Loecker et al., 2020), then the current state of market power
in advanced economies is perhaps more accurately described as one where
firms are becoming increasingly successful in targeting specific segments
of the market on which they charge higher markups, a sort of ‘segmented
competition’, in addition to the existence of ‘superstar’ or pseudo-monopolist
firms. This distinction is relevant in terms of its potential effects on consumer
surplus, income distribution, innovation or productivity growth and particularly
in terms of the effectiveness of pro-competition policy measures developed by
governments.
We should note, however, that our study presents some limitations due
to data availability, specifically regarding estimations of market shares across
the entire sample of firms. Future studies could help determine if the trade-
off between market shares and markups we find in our data is also observed
in a broader set of markets or industries, or if, perhaps, there are different
competition regimes, as suggested by Davis and Orhangazi (2021), that give rise
to other interrelations between markups, market shares and investment. While
our study shows a general relation between market power and investment at
the firm level, the existence of different competitive regimes could entail more
specific patterns, which could prove to be important in determining how micro-
level relations in terms of market power translate into aggregate investment
levels. Additionally, future studies could also consider the relation between
market power and investment behavior in different moments of the business
cycle or under different demand regimes, such as the pandemic or the Global
Financial Crisis.
217
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
reFerences
Abel, A.B., and Blanchard, O.J. (1986). Investment and Sales: Some Empirical
Evidence. NBER working paper 2050. Cambridge, MA: National Bureau of
Economic Research.
Aghion, P., Bloom, N., Blundell, R., Griffith, R., and Howitt, P. (2005). Competition
and innovation: An inverted-U relationship. The quarterly journal of
economics, 120(2), 701-728.
Arellano, M., and Bover, O. (1995). Another look at the instrumental variable
estimation of error-components models. Journal of econometrics, 68(1), 29-
51.
Autor, D., Dorn, D., Katz, L. F., Patterson, C., and Van Reenen, J. (2020). The fall
of the labor share and the rise of superstar firms. The quarterly journal of
economics, 135(2), 645-709.
Bain, J. (1956). Barriers to New Competition. Harvard University Press,
Cambridge, MA.
Baldwin, R., and Teulings, C. (eds). (2014). Secular Stagnation: Facts, Causes,
and Cures. London: Centre for Economic Policy Research.
Bloom, N., Klemm, A., Newton-Smith, R., and Vlieghe, J. (2004). Technical
Appendix: Working with IBES forecast data. Accessed at www.bankofengland.
co.uk/wp/index.html
Blundell, R., Bond, S., Devereux, M., and Schiantarelli, F. (1992). Investment
and Tobin’s Q: Evidence from company panel data. Journal of econometrics,
51(1-2), 233-257.
Blundell, R., and Bond, S. (1998). Initial conditions and moment restrictions in
dynamic panel data models. Journal of econometrics, 87(1), 115-143.
Bond, S., and Meghir, C. (1994). Dynamic investment models and the firm’s
financial policy. The review of economic studies, 61(2), 197-222.
Bond, S., Hashemi, A., Kaplan, G., and Zoch, P. (2021). Some unpleasant
markup arithmetic: Production function elasticities and their estimation from
production data. Journal of monetary economics, 121, 1-14.
Brauman, R., and Kopcke, R. (2001). The performance of traditional
macroeconomic models of businesses’ investment spending. New England
economic review, 2001(2), 3-39.
Brun, L., and González, I. (2017). Tobin’s Q and Inequality. Available at SSRN
3069980.
Bryce, D.J., and Dyer, J.H. (2007). Strategies to crack well-guarded markets.
Harvard Business Review, 85(5), 84-92.
Cameron, A.C., and Trivedi, P.K. (2010). Microeconometrics using stata (vol. 2).
College Station, TX: Stata press.
Carruth, A., Dickerson, A., and Henley, A. (2000). What do we know about
investment under uncertainty?. Journal of economic surveys, 14(2), 119-
154.
Cecchetti, S.G., and Kharroubi, E. (2015). Why does financial sector growth
crowd out real economic growth?. Available at SSRN 2615882.
218 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
Chirinko, R.S. (1993). Business fixed investment spending: Modeling strategies,
empirical results, and policy implications. Journal of economic literature,
31(4), 1875-1911.
Cohen, A. and Mazzeo, M. J. (2010). Investment strategies and market structure:
An empirical analysis of bank branching decisions, Journal of Financial
Services Research, 38, 1-21.
Davis, L. (2018). Financialization and the non-financial corporation: An
investigation of firm-level investment behaviour in the United States.
Metroeconomica, 69(1), 270-307.
Davis, L., and Orhangazi, Ö. (2021). Competition and monopoly in the U.S.
economy: What do the industrial concentration data show?. Competition &
Change, 25(1), 3–30.
De Loecker, J., Eeckhout, J., and Unger, G. (2020). The rise of market power
and the macroeconomic implications. The quarterly journal of economics,
135(2), 561-644.
De Loecker, J., and Eeckhout, J. (2018). Global market power. NBER working
paper 24768. Cambridge, MA: National Bureau of Economic Research.
De Loecker, J., and Warzynski, F. (2012). Markups and firm-level export status.
American economic review, 102(6), 2437-2471.
Diez, F.J., Fan, J., and Villegas-Sánchez, C. (2021). Global declining competition?.
Journal of international economics, 132, 103492.
Diez, F.J., Leigh, D., and Tambunlertchai, S. (2018). Global Market Power and its
Macroeconomic Implications. IMF Working Papers, 2018(137). Washington
DC: International Monetary Fund.
Doraszelski, U., and Jaumandreu, J. (2021). Reexamining the De Loecker &
Warzynski (2012) method for estimating markups. CEPR Discussion Paper
DP16027. London: Center for Economic Policy Research.
Eggertsson, G.B., Robbins, J. A., and Wold, E.G. (2021). Kaldor and Piketty’s
facts: The rise of monopoly power in the United States. Journal of monetary
economics, 124, 19-38.
Fazzari, S.M., and Petersen, B.C. (1993). Working capital and fixed investment:
new evidence on financing constraints. The RAND journal of economics,
24(3), 328-342.
Gal, P.N. (2013). Measuring Total factor productivity at the firm level using OECD-
ORBIS. OECD Economics Department Working Papers 1049. Washington
DC: OECD Publishing.
Gopinath, G., Kalemli-Özcan, Ş., Karabarbounis, L., and Villegas-Sanchez, C.
(2017). Capital allocation and productivity in South Europe. The quarterly
journal of economics, 132(4), 1915-1967.
Gordon, R.J. (2012). Is US economic growth over? Faltering innovation confronts
the six headwinds. NBER working paper 18315. Cambridge, MA: National
Bureau of Economic Research.
Gutierrez, G., and Philippon, T. (2016). Investment-less growth: An empirical
investigation. NBER working paper 22897. Cambridge, MA: National Bureau
of Economic Research.
219
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
Gutierrez, G., and Philippon, T. (2017). Declining Competition and Investment in
the US. NBER working paper 23583. Cambridge, MA: National Bureau of
Economic Research.
Grullon, G., Larkin, Y., and Michaely, R. (2019). Are US industries becoming more
concentrated?. Review of finance, 23(4), 697-743.
Jorgenson, D.W. (1971). Econometric studies of investment behaviour: A survey.
Journal of economic literature, 9(4), 1111-1147.
Keil, J. (2017). Explaining the concentration-profitability paradox. Review of
political economy, 29(2), 209-231.
Keil, J. (2019). Is there a causal effect of concentration on persistent profitability
differentials?. Industrial and corporate change, 28(2), 241-257.
Love, I. (2003). Financial development and financing constraints: International
evidence from the structural investment model. The review of financial
studies, 16(3), 765-791.
Nikolaidi, M. (2014). Margins of safety and instability in a macrodynamic model
with Minskyan insights. Structural change and economic dynamics, 31, 1-16.
Orhangazi, Ö. (2008). Financialisation and capital accumulation in the non-
financial corporate sector: A theoretical and empirical investigation on the
US economy: 1973–2003. Cambridge journal of economics, 32(6), 863-
886.
Penrose, E. (1959). The Theory of the Growth of the Firm, Oxford University
Press, Oxford.
Porter, M. (1981). The contributions of industrial organization to strategic
management, Academy of management review, 6(4)
Rafique-Hashmi, A. and Van Biesebroeck, J. (2016). The Relationship between
Market Structure and Innovation in Industry Equilibrium: A Case Study of the
Global Automobile Industry, The Review of Economics and Statistics, 98 (1)
Roodman, D. (2009). How to do xtabond2: An introduction to difference and
system GMM in Stata. The stata journal, 9(1), 86-136.
Shaikh, A. (2016). Capitalism: Competition, conflict, crises. Oxford: Oxford
University Press.
Shubik, M. and Levitan, R. (1980) Market Structure and Behavior, Cambridge,
Harvard University Press, MA and London, England.
Stiglitz, J. (2017). Inequality, stagnation, and market power: the need for a new
progressive era. Roosevelt Institute Working Papers 19. New York: Roosevelt
Institute.
Stockhammer, E. (2004). Financialisation and the slowdown of accumulation.
Cambridge journal of economics, 28(5), 719-741.
Summers, L.H. (2014). U.S. Economic Prospects: Secular Stagnation, Hysteresis,
and the Zero Lower Bound. Business economics 49(2): 65–73.
Tirole, J. (1988). The Theory of Industrial Organization. The MIT Press,
Cambridge, MA.
Tori, D., and Onaran, Ö. (2020). Financialization, financial development and
investment. Evidence from European non-financial corporations. Socio-
Economic Review, 18(3), 681-718.
220 Agustin Pedrazzoli · Ignacio Álvarez · Santos M. Ruesga Benito
aPPendIx
table 2. nuMber oF Market observatIons by FIrM count
Market size Markets Market-year observations
Full sample 1,683 14,649
At least 20 firms 219 1,614
At least 50 firms 79 568
table 3. suMMary statIstIcs
Variable Mean Std. Dev. Observations
I/K 22.065 overall 22.998 N = 95087
between 20.935 n = 12910
within 15.296 T-bar = 7.365
Q 1.144 overall 1.333 N = 90850
between 1.299 n = 12053
within 0.697 T-bar = 7.538
S/K 16.266 overall 36.693 N = 100295
between 37.082 n = 13449
within 18.953 T-bar = 7.457
D/K 8.299 overall 21.827 N = 101758
between 21.765 n = 13504
within 12.748 T-bar = 7.535
Markup 1.418 overall 3.014 N = 93841
between 3.249 n = 12787
within 1.446 T-bar = 7.339
Market share 5.749 overall 8.797 N = 27876
between 8.559 n = 4249
within 1.958 T-bar = 6.561
221
Market Power and InvestMent In advanced econoMIes: MonoPolIstIc and ‘segMented’ coMPetItIon
revIsta de econoMía MundIal 69, 2025, 197-221
table 4. non-lInear eFFects on restrIcted saMPle
Variables I II
Market share Markups
(I/K)-1 0.1911*** 0.2590***
(0.0616) (0.0575)
Q-1 2.9028*** 6.5030***
(1.0251) (1.7632)
(S/K)-1 0.0393 0.0590
(0.0466) (0.0753)
(D/K)-1 0.0670 0.1137
(0.0823) (0.1275)
Market share-1 -0.0236
(0.3143)
Square Market share-1 -0.0046
(0.0067)
Markup-1 -0.2639
(1.0541)
Square Markup-1 0.0394
(0.0578)
Constant 14.9373*** 8.2360***
(2.5952) (1.7904)
Observations 19,021 18,111
Number of id 3,419 3,300
Cluster variable country-sector country-sector
Num. Clusters 149 147
Year FE YES YES
Num. Instruments 127 127
Hansen test 0.189 0.179
AR1 2.63e-08 1.72e-09
AR2 0.846 0.744
Wald Chi-square 1082 2160
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
