WPS7406


Policy Research Working Paper                      7406




            Are Large Firms Born or Made?
              Evidence from Developing Countries

                              Meghana Ayyagari
                             Asli Demirguc-Kunt
                             Vojislav Maksimovic




Development Research Group
August 2015
Policy Research Working Paper 7406


  Abstract
 This paper uses survey data from 120 developing countries                          in firm size and growth since birth over the firm lifecycle,
 to compare the role of institutions with firm characteristics                      whereas country factors dominate in predicting variation in
 at the time of creation of the firm in explaining the size,                        labor productivity over the firm lifecycle. The study also finds
 growth, and productivity of firms over their lifecycle. The                        that older firms are larger, partly because of the selection
 study finds that firm-level characteristics have comparable,                       of more efficient firms. The findings point to the impor-
 and sometimes even larger, power than institutional factors                        tance of initial founding conditions in explaining variations
 in predicting size and growth, but not productivity. In par-                       in size and growth over the firm lifecycle across countries.
 ticular, size at birth plays a key role in predicting variation




  This paper is a product of the Development Research Group. It is part of a larger effort by the World Bank to provide open
  access to its research and make a contribution to development policy discussions around the world. Policy Research Working
  Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at Ademirguckunt@
  worldbank.org, ayyagarim@gmail.com, and vmaksimo@rhsmith.umd.edu.




         The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
         issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
         names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
         of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
         its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.


                                                       Produced by the Research Support Team
            Are Large Firms Born or Made? Evidence from Developing Countries



Meghana Ayyagari                             Asli Demirguc-Kunt                           Vojislav Maksimovic




JEL codes: O1, L26, L25


Keywords: Firm productivity and growth; entrepreneurship; institutions; economic
development.
__________
*Ayyagari: School of Business, George Washington University, ayyagari@gwu.edu, Ph: 202-994-1292; Demirgüç-
Kunt: World Bank, ademirguckunt@worldbank.org, Ph: 202-473-7479; Maksimovic: Robert H. Smith School of
Business at the University of Maryland, vmaksimovic@rhsmith.umd.edu, Ph: 301-405-2125. We would like to
thank Miriam Bruhn, Roberto Fattal Jaef, Hiau Looi Kee, Aart Kraay, David McKenzie, and Sergio Schmukler for
their valuable comments and suggestions on an earlier version of this paper titled “Do firms in developing countries
grow as they age?” We thank Xiaoyuan Hu for excellent research assistance. .
I. Introduction

        Do developing country firms grow as they age? We know that in developed countries

such as the U.S., new businesses start small and, if they survive, grow fast as they age.1 But it is

not clear whether firms in developing countries face a similar size-age profile as in the US given

that they face entirely different financial environments and operating conditions.2 In this paper,

we examine lifecycle effects in firm size, growth, and productivity across 120 developing

economies. We are interested in three questions: Do firms grow over their lifecycle? Are

differences in the size and growth of firms better explained by the differences in institutions,

especially financial and legal, across countries, as a voluminous literature suggests, or by the

differences in firm level characteristics, in particular initial size at birth? How does firm

productivity change over the lifecycle and what explains this variation?

        A body of research in finance has established the importance of lifecycle explanations for

many fundamental corporate finance policies, including dividends (Fama and French (2001),

Grullon et al. (2002), DeAngelo, DeAngelo, and Stulz (2006), Denis and Osobov (2008)),

financing (Berger and Udell, 1990), stock valuations (Pastor and Veronesi (2003)) and

acquisitions (Maksimovic and Philips (2008), Arikan and Stulz (2011)). We also know that size

and age are closely related and are the best predictors of financing constraints (Hadlock and

Pierce (2010)). However, much less is known about if there is a lifecycle of firm size and what

factors explain the evolution of firm size and productivity with age.

        We start by examining whether older firms are larger, grow faster, and are more

productive than younger firms and whether these relationships vary across different countries



1
 See, for example, Evans, 1987 and Davis, Haltiwanger, and Schuh, 1996.
2
 An exception is Hsieh and Klenow (2014) who contrast growth trajectories in India and Mexico with those in the
USA.

                                                                                                                  2
and industries. We then assess the extent to which the observed variation in average size, growth

since birth, and productivity at different points in the firms’ lifecycle in different countries can be

explained by two explanatory factors: First, we explore how much country-level institutional

factors affect firm size, growth, and productivity at different ages since founding. Here we

quantify the explanatory power of these country-level institutional factors on firm indicators.

However, we recognize that institutions also affect the initial size distribution of firms and that

their influence through this latter channel may be very significant.

        Second, we estimate the extent to which these post-birth variables can be explained by

the firm’s initial size at founding. As the Lucas (1978) and Rauch (1991) models suggest, and

Klepper (2001) shows empirically, more capable entrepreneurs are expected to form larger

firms.3 Given that entrepreneurial talent is a likely predictor of firm performance, we condition

on initial firm size when predicting the role of institutions on subsequent growth. Furthermore,

as Klapper et al. (2006), Kerr and Nanda (2009), and Ayyagari et al. (2014b) show, initial firm

size is influenced by institutions within each country by affecting the financing of

entrepreneurial firms and the pool of potential educated entrepreneurs. Thus, the explanatory

power of initial size in the performance regressions also places an upper bound on the effect of

institutions through this channel.

        Using a variance decomposition approach, we examine differences between countries as

well as differences between different types of firms, and their interaction.4 We focus on the


3
  Maksimovic, Phillips, and Yang (2013) show initial size and productivity are predictors of future public firm
status, and hence growth and productivity.
4
  To examine the relative influence of firm effects compared with country effects, we use variance decomposition
analysis. Similar methods have been used in finance to examine the importance of country vs. firm level effects in
explaining property rights protection (Ayyagari, Demirguc-Kunt, and Maksimovic, 2008). The method is also well
established in corporate strategy studies in the context of decomposing profitability into corporate and industry
effects [Schmalensee (1985), Rumelt (1991), McGahan and Porter (1997, 2002), Khanna and Rivkin (2001)]. By
using this method, we can focus directly on the general importance of these effects in explaining firm size and
growth over lifecycle without any assumptions on structural analysis.

                                                                                                                     3
effect of country level institutions and initial firm size, but extend the analysis to consider other

firm-level characteristics, such as ownership and the firm’s legal organization, which while not

exogenous to the firm’s outcomes, nevertheless provide descriptive evidence of the association

between performance and firm characteristics. Finally, we examine the within-country

covariance between firm size and productivity at different points along the lifecycle to get a

sense of the extent of misallocation in developing countries.

        We use data from the World Bank Enterprise Surveys (ES) database, consisting of 44,870

formal firms5 in 118 countries, surveyed over the period 2006-2012. The surveys use

standardized survey instruments and a uniform sampling methodology to minimize measurement

error and to yield data that are comparable across countries. The strength of our data is that we

cover most of the developing countries in the world. We have the following main findings:

        First, our analysis shows that in developing countries, older firms are substantially larger

than younger firms. The average firm that is 40 years and older employs 5 times as many

workers as the average firm under the age of five. Our results are robust to considering only

manufacturing firms and taking into account survey sampling weights. We obtain similar

patterns when we look across country income groups. Our results are also robust to excluding

firms with government ownership, former socialist economies where there may be a discrepancy

in age reported due to large-scale privatization,6 and also to excluding firms that started

operations before countries’ independence. This finding also holds controlling for industry and

country fixed effects in a multivariate analysis. The significantly upward sloping age-size profile




5
  While the Enterprise Survey data surveys establishments, we use the terms firm and establishment interchangeably
since we restrict our analysis to single establishment firms.
6
  The establishments in the survey were explicitly told that if the establishment was privatized, then the
establishment year provided should refer to when the original government-owned establishment began operations

                                                                                                                 4
is pervasive in the vast majority of countries. In only less than 10% of the countries is the

average 40+ year firm not as big as firms younger than 5 years.7

         Second, we find that as firms age, they not only grow and employ more workers, but are

also much more productive. This is consistent with the literature (e.g. Pagano and Schivardi,

2003) suggesting that larger firms are more productive.

         Third, our analysis shows that firm-level characteristics dominate country-level

characteristics in predicting variation in size and growth over the lifecycle since birth. The firm-

level factors together explain between 34.6% of the variation in firm size in the case of mature

firms (20-39 years) and 51.1% of the variation in firm size in the case of young firms (<5 years).

Of the firm-level characteristics, size at birth dominates other factors such as location, industry,

ownership, and legal organization. In particular, size at birth has the largest explanatory power

explaining 52.2% of the variation in firm size for young firms, 37.3.3% for firms aged 5-19 years

and 27.2% for mature firms aged 20-39 years. In comparison, institutional factors matter less

than size at birth for explaining variation in size over the lifecycle. The total variation in size at

different points along the lifecycle that can be explained by country dummies ranges between 9%

(for firms in the 5-19 and 20-39 age bins) and 12.5% for young firms (<5 years).8




7
  However we also find a great deal of variation in the size-age gap across countries. In about 22% of the sample (27
countries), the size-age ratio is as large or larger than in the U.S. where Hseih and Klenow (2014) show that the
average 40 year old firm employs over seven times as many workers as the average firm five years or younger.
8
  Of the explainable variation in size at the country level, 24%-37% can be explained by using historical institutional
variables proposed in the literature such as legal origin (La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1998),
ethnic fractionalization (Easterly and Levine (1997), Alesina et al. (2003)) and latitude (Engerman and Sokoloff
(1997)). In smaller sub-samples where we also include settler mortality (Acemoglu, Johnson, and Robinson (2001)),
a proxy for the disease environment that shaped property rights protection during European colonization, the
historical institutional variables together explain nearly 71% of the explainable variation in size for firms in the 5-19
age bin. Similar results hold for growth, suggesting that the literature has been looking at first-order institutional
factors.


                                                                                                                       5
        We see similar patterns when we look at the variation in growth since birth. Size at birth

explains 3.2% of the variation in growth since birth for young firms, 3.8% for firms aged 5-19

years, and 4.8% of the variation for mature firms. Country dummies by contrast explain between

1.2% (for young firms) and 2.9% (for mature firms) of the variation in firm growth over the

lifecycle.

        Fourth, country-level characteristics are more important than firm-level characteristics in

predicting variation in labor productivity along different points in the lifecycle. The state of

institutions, particularly financial institutions, influences firm and industry performance and

growth.9 We find that country dummies explain 55.9% of the variation in labor productivity for

young firms, and 52% for firms aged 5-19 years and mature firms aged 20-39 years. Historical

institutional variables proposed in the literature such as legal origin (La Porta, Lopez-de-Silanes,

Shleifer, and Vishny, 1998), ethnic fractionalization (Easterly and Levine (1997), Alesina et al.

(2003)) and latitude (Engerman and Sokoloff (1997)) explain more of this explainable variation

for young firms (21%) than for mature firms (15%). In comparison, all the firm-level

characteristics (industry, location, ownership, legal organization, and size at birth) explain only

up to 10.5% of the variation in labor productivity at different points along the lifecycle.

Likewise, the interaction between firm-level characteristics and country dummies or country-

level institutional variables also does not contribute much explanatory power.

        Finally, we try to understand to what extent selection processes may be driving our

results by studying the variation in the size-productivity covariance across countries.

Bartelsman, Haltiwanger, and Scarpetta (2013) show that this measure captures the extent of



9
  Rajan and Zingales (1998) show that financial development leads to comparatively faster growth rates in industries
normally dependent on external finance; Demirguc-Kunt and Maksimovic (1998) and Beck et al. (2005) show that
financial development promotes growth of individual firms .

                                                                                                                  6
resource mis-allocation and selection forces in an economy. Our results show that there is a great

deal of resource misallocation in developing countries since the covariance term is largely

negative. We also find that this misallocation seems to decrease with age since the average and

median size-productivity covariance appears to increase with age suggesting that on average the

unproductive firms exit so older firms are more productive.

       Our results contribute to the large literature on firm size and age (Dunne, Roberts, and

Samuelson (1989), Davis, Haltiwanger, and Schuh (1996), Cabral and Matta(2003), Foster,

Haltiwanger, and Syverson (2012) and advance the findings in Ayyagari, Demirguc-Kunt, and

Maksimovic (2014a) who find that mature firms have the largest share of employment in

developing countries. Overall our results suggest that across the world, firms start small and

grow as they age and this is consistent with the studies of the lifecycle of firms in developed

countries.

       This paper is also related to recent work investigating the importance of country versus

firm-level factors in explaining firm perceptions and behaviors. Ayyagari, Demirguc-Kunt, and

Maksimovic (2008) find that firm-level characteristics, such as legal organization and ownership

structure, are comparable with institutional factors in explaining variations in perceptions of

property rights protection. Doidge, Karolyi, and Stulz (2007) on the other hand find that most of

the variation in governance ratings across firms is explained by country characteristics rather

than by firm characteristics. They attribute this finding to the increased incentives of firms in

better legal environments to adopt better governance structures. Our results suggest that while

institutional variation across countries is important in understanding the size-age gap, firm-

specific factors are equally important if not more important in explaining the variation in size

and growth at different points in the firm lifecycle.



                                                                                                    7
         A number of caveats exist. Our results cover only the formal sector, that is, firms that are

formally registered and operating as legal entities. We also do not have data on micro enterprises

(less than 5 employees) in our sample. The dual economy view in development economics

(originally associated with Harris and Todaro (1970)) predicts that informal firms should look

very different from formal firms in terms of size, productivity, wages paid and the

industries/markets they operate in. Recent studies such as La Porta and Shleifer (2008) also show

that while informal firms account for a large portion of economic activity in developing

countries, growth and development come from the creation of highly productive formal firms.

Second, our data are only on the continuing/surviving firms and hence we have no data on firm

exit rates. On a related note, the surveys are stratified only by industry, firm size, and

geographical location and so we may not have a completely representative sample of firm ages,

though the firms within the strata are randomly sampled. Finally, the ES data are based on

survey data rather than individual country census. However, our enquiries with the survey

implementation team reveal that mean response rates across our sample of countries is 70%,10

which suggests that we have a representative sample of firms across economies. Overall, our

analysis provides an important first step in exploring cross-country differences in the lifecycle of

firms.


II. Data

         We use the World Bank Enterprise Surveys (ES), which are an ongoing initiative of the

World Bank to benchmark the investment climate in different countries across the world and to


10
  The 70% mean response rate across our sample of countries is superior to most other survey based studies. For
instance, Campello, Giambona, Graham, and Harvey (2011) report response rates of 3 to 7% in their survey of how
companies use credit lines during a financial crisis. Other studies in corporate finance report response rates between
7%-9% including Graham and Harvey (2001), Brav, Graham, Harvey and Michaely (2005), and Lins, Servaes, and
Tufano (2010).

                                                                                                                     8
analyze firm behavior and performance (see www.enterpriseseurveys.org for more details). This

is the only multi-country survey of its size and scope and has been used in numerous prior

studies.11

         The Enterprise Surveys survey from the universe of eligible firms obtained from the

country’s statistical office12 using stratified random sampling with replacement to generate a

sample representative of the whole non-agricultural private economy (so fully government

owned firms are excluded from the sampling universe) in the country. The surveys are stratified

according to three criteria: Sector of activity (population of industries include manufacturing

sectors, construction, services, transport, storage, communications, and computer and related

activities), Firm size (the strata include small firms (5-19 employees),13 medium firms (20-99

employees), and large firms (100 or more employees)), and Geographical location (selected

based on centers of economic activity in the country).

         While the ES have been produced since 2002, we restrict our sample to surveys

administered during 2006-2012 since these provide sampling weights that address the varying

probabilities of selection across different strata and are thus indispensable to making assertions

about the whole population.14 For countries that were surveyed twice during this period, we


11
   The Enterprise Surveys and their precursor, the World Business Environment Survey (WBES) have been used to
investigate a series of questions in finance and economics including the relation between property rights and
contracting institutions (e.g. Acemoglu and Johnson ( 2005)), investment climate and business environment
obstacles to growth (e.g. Beck et al. (2005) and Ayyagari et al. (2008)), firm financing patterns (e.g. Beck et al.
(2008), Cull and Xu (2005) and Ayyagari et al. (2010)), dispute resolution via courts (e.g. Djankov et al. (2003)),
corruption (e.g. Barth et al. (2009)) and tax evasion (Beck, Lin, and Ma, 2014).
12
   The master list of firms is sometimes obtained from other government agencies such as tax or business licensing
authorities. In some cases, the sampling universe is generated from lists maintained by the Chamber of Commerce
and business associations or marketing databases where registration is voluntary. In a few cases, the sample frame is
created via block enumeration.
13
   We have a few firms in the sample that report having less than 5 employees. All our results are robust to dropping
these firms. Our results are also robust to dropping firms that started out as informal enterprises and were formally
registered after starting operations.
14
   Most surveys contain three sets of weights – strict, median, and weak weights depending on the eligibility criteria
used to construct the sample universe. The survey implementation manual recommends the use of median weights
for cross-country comparisons which is what we use in this paper.

                                                                                                                     9
retain the most recent survey data for each country. Our results are also robust to retaining data

for the year with the largest number of firms surveyed.

        While the sampling unit in ES is the establishment,15 we restrict out analysis to single

establishment firms that comprise over 85% of the sample.16 Hence we use the terms firm and

establishment interchangeably. We drop missing observations on size and age, drop observations

where the date the firm began operations is before 1800 and obvious data errors (such as one

observation where date the firm began operations is reported to be 2205). This leaves us with a

final sample of 44,696 firms across 118 countries.17 The key variables used in our analysis are

the following:

        Firm Size is the number of permanent, full-time employees in the firm where permanent,

full-time employees are defined as all paid employees that are contracted for a term of one or

more fiscal years and/or have a guaranteed renewal of their employment contract and that work 8

or more hours per day. The survey asks firms to report the employment numbers at the end of the

last fiscal year.

        Size at Birth is the size of the firm when it first started operations. Size Ratio is the ratio

of current firm size to the size of the firm when it first started operations.

        Labor Productivity is the logarithm of the ratio of total annual sales (in USD) to number

of permanent, full time employees. While the ES report annual sales in local currency units




15
   In the Enterprise Surveys, the establishment is defined as a physical location where business is carried out and
where industrial operations take place or services are provided. In addition, an establishment must make its own
financial decisions, have its own financial statements separate from those of the firm, and have its own management
and control over its payroll.
16
   When we focus on multi-establishment firms, we find that they have similar size-age profiles as single-
establishment firms. The average 40+ multi-establishment firm is 4.5 times larger than the average multi-
establishment firm that is <5 years of age.
17
   The Enterprise Survey database was downloaded in Dec 2012.

                                                                                                                10
(LCU), we use end of period USD/LCU exchange rates from the International Financial

Statistics (IFS) to obtain sales in USD.

           Firm Age is defined as the number of years since the firm began operations in the

country. We compute age at the end of the last fiscal year to be consistent with the employment

question.

           We winsorize the top and bottom 1% of firm size, size at birth, and labor productivity

(before taking logs) to minimize the effect of outliers.

           Table1 presents the summary statistics and correlations between the main variables. Panel

A of Table 1 shows that the mean firm size in our sample is 59 employees. The mean firm size

ratio is 5.64. Thus, the average firm in our sample is 5.6 times its size when it started operations.

The ratio extends from a very small number to 860. The firms with ratios in excess of 100 are

spread out across countries and industries.18 For instance, there are two firms with a size ratio of

860, one is a Columbian firm surveyed in 2010 in the industry sector Metals and Machinery

which is privately held and started operations in 1940 with 1 employee and had 860 employees

in 2009 and the other is a publicly listed company in Mauritius in Non-metallic and Plastic

Materials surveyed in 2009 which started operations in 1953 with 1 employee. 55% of our

sample is composed of manufacturing firms and while the average log labor productivity is

10.08, there is wide heterogeneity as seen by the range between 5.81 and 17.2. Panel B shows

that older firms in our sample are larger and more productive. Firm Size Ratio and Age are also

positively correlated suggesting that firms grow as they age.




18
     Our results are robust to using median regressions rather than OLS for the specifications with size ratios.

                                                                                                                   11
       In the next two sub-sections, we describe the institutional and firm-level characteristics

we use for our variance decomposition analysis and the underlying motivation behind the choice

of these variables.


A. Country-Level Institutional Determinants

       In this sub-section, we describe the different country-level variables we use in the

variance decomposition analysis to capture the contribution of country factors in explaining firm

size and growth over the lifecycle. We begin with four exogenous institutional variables - Legal

Origin, Latitude, Ethnic Fractionalization, and Settler Mortality - that have been widely

considered in the literature to be determinants of property rights protections and with long-

lasting economic consequences. Here, we consider whether these historical institutions have a

significant explanatory power in predicting the size, growth, and productivity of firms over their

lifecycle. Below we first describe the hypotheses behind each of these institutional theories and

the variables we use to capture them.

       La Porta et al. (1998) argue that countries' legal systems differ in how much they protect

the rights of private investors vis-à-vis the state and minority shareholders. La Porta et al. (1998)

maintain that legal systems that evolve from common law traditions tend to support private

property rights. By contrast, countries established civil law systems as acts of policy. Such

systems tend to be designed to state administration, are more predictable, and are less likely to

favor individuals over the state, or to tailor decisions in ways that safeguard individual claimants

in specific instances. Subsequent research showed that legal origin has consequences not just for

the legal and regulatory framework of society but also for economic outcomes such as

government ownership of banks (La Porta et al. 2002), entry regulations (Djankov et al., 2002),




                                                                                                    12
regulation of labor (Botero et al. 2004), and financial development (La Porta et al. 1998,

Claessens and Laeven (2003), and Beck et al. (2005))

        Through these different channels, we expect legal origin should also have an impact on

firm size and growth over the lifecycle. For our sample of countries, we use Legal Origin, a set

of three dummy variables to capture the differences between three influential legal traditions -

British common law, French civil law, and Socialist law - using data from La Porta et al. (1998,

1999).19

        Acemoglu, Johnson, and Robinson (2001) argue that many countries, especially former

colonies, did not design the legal system to protect property rights. Instead, its purpose was to

facilitate the extraction of resources from the indigenous population. Thus, two systems with the

same legal origin may in practice offer very different protections. Acemoglu et al. (2001, 2002)

and Engerman and Sokoloff (1997) contend that European colonization offers a natural

experiment to test this hypothesis. Europeans set up extractive systems in colonies that were not

attractive for colonial settlement, either because of high settler mortality due to natural causes at

the time of colonization, or because the indigenous population was relatively large. In colonies

where settlement was feasible, countries set up the judicial systems to protect the property rights

of the settlers. This theory emphasizes the role of geography (latitude and natural endowments)

and disease environment (which affected the settler mortality) in shaping property rights.

        To measure geographical endowments, we use Latitude, which is the absolute value of

the latitude of the country scaled between zero and one, from La Porta et al. (1998). Countries

closer to the equator have a more tropical climate that is inhospitable to European settlers, and



19
  Three countries – Ethiopia, Iran, and Yemen – were re-classified from French Civil law in La Porta, Lopez-de-
Silanes, and Shleifer (2008) to English Common Law in Djankov, McLiesh, and Shleifer (2007). We retain the
original classification but our results are not significantly different with the change.

                                                                                                                  13
therefore these countries fostered “extractive” institutions. We use data on settler mortality from

Acemoglu, Johnson, and Robinson (2001). We define Settler Mortality as a measure of the death

rates for European settlers in former colonies and calculate it by the logarithm of annualized

deaths per thousand Europeans. Since we have data on Settler Mortality for a smaller sample, we

use Latitude for our main results and present the results on Settler Mortality as a robustness test.

       Easterly and Levine (1997) show that Ethnic Fractionalization is also an important

determinant of rent seeking and social polarization that affects property rights and social

institutions. Empirically, we use Ethnic Fractionalization, a measure of ethnic fractionalization

of the country compiled by Alesina et al. (2003).

       The explanatory power of the different variables may be affected due to nonlinearities

arising from the way we construct these variables as shown by Ayyagari, Demirguc-Kunt, and

Maksimovic (2008). Following Ayyagari et al. (2008), we rescale the variables by constructing a

four-point scale for Latitude, Settler Mortality, and Ethnic Fractionalization based on their

quantiles. We perform variance component analysis using this four-point scale. In the regression,

we enter dummies for each unique value of the rescaled variables.


B. Firm-Level Characteristics

       For the variance decomposition analysis, we consider the following firm-level

characteristics:

       First, we look at the initial conditions of the firm by looking at the size of the firm when

it first started operations, Size at birth. Theoretical models on entrepreneurial ability such as

Lucas (1978) and Rauch (1991) suggest that entrepreneurs of higher ability set up larger firms or

correspondingly managers with more ability run larger firms and employ more capital. Entry

size has also been shown to be important for survival based on the rationale that the relative

                                                                                                    14
efficiency of the entrant is discovered subsequent to entry through the process of learning from

actual market experience (e.g. Ericson and Pakes (1998), Jovanovic (1982), Geroski (1995), and

Sutton (1997)). Agarwal and Audretsch (2001) show that large entry size for survival matters

particularly in the formative stages of the industry lifecycle. Hence we look at the role of size of

birth though the focus in our paper is not on survival but relative size, growth, and productivity

over the lifecycle.

       Second, we look at the importance of firm location for lifecycle effects. Firm location has

been a much researched issue since Marshall (1890). Studies in economic geography have shown

that firms in urban clusters are more likely to outsource (Ono (2003)), vertically disintegrate

(Holmes (1999)), and innovate (Glaeser et al. (1992)). More recently a large finance literature

has used location as a proxy for growth opportunities (e.g. Almazan, De Motta, Titman, and

Uysal, (2010)) and shown the importance of location for acquisition behavior (Uysal, Kedia, and

Panchapagesan (2008); Kang and Kim (2008)) and corporate payout policies (John, Knyazeva,

and Knyazeva (2011)). Urban firms have also been shown to use more equity financing and

higher quality underwriters (Loughran and Schultz (2006)). In our context we apply this

framework to examine to what extent the firm location and its ensuing growth opportunities

influence size, growth, and productivity over the lifecycle. Location consists of five dummy

variables for Capital city, city with population over 1 million, population over 250,000 to 1

million, population over 50,000 to 250,000, and less than 50,000 population.

       Third, we look at the importance of ownership structure. Theories on the politics of state

ownership argue that state owned firms are less efficient and less profitable than privately owned

firms because of governments’ deliberate policy to transfer resources to supporters and weak

managerial incentives to reduce costs and (Shleifer and Vishny (1994); Shleifer (1998)).



                                                                                                   15
Empirically too, a large literature has shown that state owned firms under-perform private firms

(e.g. Boubakri and Cosset (1998); Dewenter and Malatesta (2001); Megginson and Netter

(2001); La Porta and Lopez-de-Silanes (1999); and Gupta (2005)), are less innovative (Ayyagari,

Demirguc-Kunt, and Maksimovic (2012)), and that investments of state owned firms are subject

to political interference (Dinc and Gupta (2011)).20 We would also expect the average size of

state owned firms to be larger than private and foreign firms both because of their social goals

and their inefficiencies. To capture Ownership, we use four dummy variables to classify firms

into State owned, Domestic owned, Foreign Owned, and Other ownership. The dummy variables

take the value 1 if the government, private domestic individuals/companies, private foreign

individuals/companies, or other ownership categories owned 50% or more of the company

respectively and 0 otherwise.

        Fourth we look at the role of legal organization structure. Recent corporate finance

research has shown that public firms are different from private firms in very fundamental ways.

These two categories of firms invest differently in response to demand shocks (Brav (2009);

Sheen (2009); Asker, Farre-Mena, and Ljunqvist (2010)), have different cash policies (e.g. Gao,

Harford, and Li (2012); Farre-Mensa (2012)), smooth dividends differently (Michaely and

Roberts (2012)) and listed firms take much more advantage of financial booms to grow by

acquisition (Maksimovic, Phillips, and Yang (2013)) while private acquirers pay significantly

less for targets than public acquirers (Bargeron, Schlingemann, Stulz, and Zutter (2008)). We

would thus expect public and private firms to have very different size, growth and productivity

profiles with age. More generally, Demirguc-Kunt, Love, and Maksimovic (2006) show a

significant interaction of institutions and incorporation status - incorporated firms grow


20
  A parallel literature has established the costs associated with government ownership in financial markets (Khwaja
and Mian (2005); Dinc (2005); Cole (2009); Sapienza (2004); Carvalho (2014))

                                                                                                                16
comparatively faster in countries with strong financial and legal institutions than in countries

with weak institutions. To examine these issues, we use Legal Organization, consisting of three

dummies for Publicly listed company, Privately held limited liability company, and an Other

category which includes Sole proprietorship, Partnership, Limited Partnership, and Other legal

status.

          Finally, we look at how the industry classification contributes to size, growth since birth,

and productivity over the lifecycle. Rajan and Zingales (1998) and Beck and Levine (2002) show

that differences in legal and financial systems affect the availability of external finance and the

growth of different industries in the manufacturing sector. We expect these differences to matter

as much or more when we look across different sectors. Hence, we also examine if belonging to

a particular Industrial Sector impacts the size, growth and productivity of firms over their

lifecycle. We use Sector, which consists of dummy variables for the following industrial sectors:

Manufacturing industries that includes Textiles, Leather, Garments, Food, Metals and

machinery, Electronics, Chemicals and Pharmaceuticals, Wood and furniture, Non-metallic and

plastic materials, Auto and auto components, and Other manufacturing; Retail and wholesale

trade; Hotels and restaurants; Other services; and Construction, Transportation and other

industries.

          Among the firm-level characteristics we consider, our focus is on size at birth since it is

the most exogenous variable. Other variables like sector and location are likely to be sticky at the

establishment level. Ownership structure and legal form are most likely to be influenced directly

by institutions as suggested by La Porta, Lopez-de-Silanes, and Shleifer (1999) and Demirguc-

kunt, Love, and Maksimovic (2006) respectively.




                                                                                                      17
III. Size, Growth Since Birth, and Lifecycle in Developing Countries

         In this section we explore the age effects associated with firm size, firm growth since

birth, and labor productivity. We begin by first looking at summary statistics on the relationship

between average firm employment in surviving plants and age in the cross-section in Figure 1.

We compute the mean employment in each country in each of the following 9 age bins: <5, 5-9,

10-14, 15-19, 20-24, 25-29, 30-34, 35-39, and 40+. We then compute an index for each country

where the index takes the value 1 for firms less than five years old. For all other age bins, the

index is the ratio of the mean employment in that age bin to the mean employment for firms

under five years of age. Figure 1 shows that in developing countries firms grow as they age.21

         The findings in Figure 1 are very robust. First, Figure 2 shows similar results when we

take sampling weights into account and compute average firm size as Total Employment/Total #

of firms. Second, the upward sloping size-age profile is pervasive in our sample of countries. In

only less than 10% of the countries are is the average firm aged 40 years or older not at least as

big as the average firm that is younger than five years old. 22 In Appendix A1, we present the

summary statistics of the firm size-age gap across different country income groups and

geographic regions. Across different classifications we find that firms grow as they age though




21
   Figure 1 shows that the firm that is 40 years or older on average employs nearly 5 times as many people as the
firm that is less than five years old. The 95% confidence intervals show that this number could vary between 4.17
and 5.64.
22
   In 10% of the sample (12 countries), the size-age ratio for the firm that is 40 years or older is 10 or over and in
20% of the sample, the size-age ratio is over 7, which is the number by which the average 40 year old firm in the US
is larger than the average firm that is five years or younger in the US (Hsieh and Klenow, 2012).

                                                                                                                   18
there is a great deal of variation in the size-age gap across countries. 23 Third, we find similar

results when we look at just manufacturing firms in Figure 3 as well as in Figure 4, where we

undertake additional robustness tests, including dropping countries with Socialist legal

tradition,24 dropping firms with any government ownership and dropping firms that were

established before the countries’ independence.

         In Table 2, we examine the empirical relationship between firm size and age in the above

figures in a regression setting. In Table 2, in col. 1, we regress Firm Size on age dummies and

find all the age coefficients to be positive and significant suggesting that older firms are larger

than younger firms. For ease of comparison with youngest firms, we add the constant (which

presents the average mean size for firms younger than 5 years old) to each of the age coefficients

and divide the sum by the constant. These ratios are reported in bold in square brackets below the

coefficient and standard errors. Thus we see that the average 40+ year old firm is 5.502 times the

size of the average firm that is younger than five years old. In col. 2, we repeat the analysis in

Col. 1 using industry dummies and country dummies to control for any time-invariant

unobserved heterogeneity across countries and industries. Once again we find that all the age

coefficients are positive and significant. Cols. 3 and 4 present results for manufacturing and non-

manufacturing sub-samples where we find a similar increase in firm size with age. Table 2 thus

shows that firms in developing countries grow as they age.25


23
   The first column in panel A presents the underlying data in Figure 2. Across income groups, we find the size-age
gap ranges from 4.8 times in the case of low income countries to 7.8 times in the case of high income countries.
Across regions in panel B, we find that the size-age gap for firms 40 years and older ranges from 1.56 in Middle
East and North Africa to 7.41 in Europe and Central Asia.
24
   We exclude countries with Socialist legal tradition because Ayyagari, Demirguc-Kunt, and Maksimovic (2008)
argue that these countries are fundamentally different from others in their perceptions of property rights protection.
One of the other concerns with the former Socialist economies might be that the data on firm year is noisy because
of mass privatizations though the firms were queried on when the firm first started operations even if it was under
state ownership. However in Figure 4 we find consistent results with our main finding when we drop former
Socialist economies.
25
   As shown in Appendix A2, we get very similar results if we were to repeat this analysis in terms of log(size).

                                                                                                                    19
         In Table 2, we are comparing the mean size of firms that are 40+ years old with the mean

size of firms that are younger than 5 years old. However, young firms could be varying in size

across years due to different macroeconomic factors, entry regulations, capital market openness,

etc. as well as across countries. In order to see if firms indeed grow as they age over their

lifecycle, we would want to compare the size of older firms at any given point in time with their

size when they were born. The Enterprise Surveys ask firms to report the number of full-time

employees when the firm started operations. In Figure 5, we normalize the average size of each

age cohort by the average size of that cohort at birth. We once again find an upward sloping

curve, consistent with our main finding that older firms in developing countries are larger than

their younger counterparts. To investigate this in a regression setting, we construct Size Ratio,

which is the ratio of each firm’s current employment level to the number of employees when it

started operations.

         Col. 1 of Table 3 presents a regression of Size Ratio on age dummies and shows that older

firms are much larger compared to their size when they first started operations.26 In Col.2 we add

in controls for industry and country and in Cols. 3 and 4 we look at sub-samples of manufacturing

and non-manufacturing industries. In all cases the age coefficients are positive and significant.

         Overall, Table 2 shows that the average older firm is larger than the average firm that is

less than five years old and Table 3 shows that the average older firm is larger than its size when

it first started operations. Together, Tables 2 and 3 and figures 1-5 show no evidence of a declining

lifecycle in our sample of firms.




26
  The constant shows that the average firm that is less than 5 years old is 2.92 times its size when it first started
operations whereas the coefficient of the 40+ age bin shows that the average 40+ year old firm is 9.64 times its size
when it first started operations.

                                                                                                                   20
        We next conduct the same analysis for labor productivity (in levels) instead of size.

Figure 6 plots the mean and 95% confidence intervals of the index of productivity for each age

bin. Figure 6 shows that labor productivity index increases from 1 to 2 between the <5 and 5-9

age bin (i.e. a 100% increase), declines thereafter and remains mostly flat but is still above 1, and

then again increases significantly in the last age bin of 40+ firms. Given the large confidence

intervals on this graph at all ages, we do not emphasize these results much. In Table 4, we repeat

the regression analysis in Table 2 but with log labor productivity as our dependent variable. All

the age coefficients are positive and significant suggesting that older firms have higher labor

productivity than younger firms. Each of the age dummy coefficients represent the % increase in

labor productivity relative to the reference category. The exact percentage changes are given by

the formula 100 * (exp(coefficient)-1) and are reported in bold in square brackets below the

coefficient and standard errors. Col. 1 of Table 4 shows that the labor productivity of the average

40+ firm is approximately double that of the average firm that is younger than five years old.

        To summarize, the findings in this section show that firms in developing countries grow

larger and are also more efficient as they age. The yet unanswered question is whether the

increase in firm size with age is a function of institutions in the country or if it is determined by

factors intrinsic to the firm? In the next section we attempt to answer this by analyzing what

firm and country level factors are key in determining the size, growth, and productivity over the

lifecycle.




                                                                                                    21
IV. Size, Growth Since Birth, and Firm Lifecycle – Analysis of Variance

        We use variance decomposition analysis to compare the relative importance of country

and firm effects in explaining variation in firm size, growth, and productivity over different

points in the lifecycle.

        To fix ideas, we start with a simple model of firm size evolution over the firm lifecycle.

If Sict is the size of firm i, in country c, at point t, we have:

                                                                                                 (1)

where Sic0 is size of the firm at birth, Ac is country effect, and for simplicity, the growth rate gic is

     gc = g0+g1c                                                                                 (2)

In terms of logs

        log Sit = log Ac + (g0 + g1c)t x log Si0                                                 (3)

        log Sit = log Ac + g0 x (t log Si0) + g1c x (t log Si0)                                  (4)

        That is, the size of a firm at point t can be decomposed into the scaling effect of initial

size in that country (log Ac), the universal growth adjusted effect of initial size (g0 x (t log Si0)) ,

and the country specific universal effect of size (g1c x (t log Si0)).

        To compare the relative importance of initial size versus country factors, we first estimate

equation (4) for the full sample of firms. We analyze this model using a regression-based,

simultaneous ANOVA approach that uses the standard assumptions of ordinary least squares.

We begin by estimating a restricted version of equation (4) where we include only the country

effect. The R-square of this regression provides an estimate of the proportion of the variation in

firm size, which is explained by the country-level institutional variable alone. That is, it provides

an upper bound for the amount that can be explained by that variable directly, and by other

variables that the institutional variable predicts. We next add initial size and compute the R-


                                                                                                       22
square to obtain an estimate of the proportion of the variation in firm size that is explained by

both the institutional variable and initial size together. We note that the difference in the R-

square between the new regression and the restricted regression does not provide an estimate of

the variation explained by the firm-specific variable by itself. Instead, it estimates the marginal

increment of explanatory power that we gain by adding the firm-specific variable. Finally, we

add an interaction term to provide for the possibility that initial firm characteristics may affect

firm size for some values of the institutional variable, but not for other values.

       In equation (4), for the country effect, apart from the base regression that uses country

dummies, we use the following four institutional variables – legal origin, latitude, settler

mortality, and ethnic fractionalization that are described in detail in the previous section. To

better understand lifecycle effects, we estimate equation (4) for different age buckets – young

firms (<5 years), mid-age (5-19 years), and mature firms (20-39 years). We also examine the

following firm characteristics in addition to initial size in specifications analogous to equation

(4) - age, ownership, legal organization, industry sector, and city location.

       Finally, we use alternate dependent variables including labor productivity and growth

since birth. Overall, our empirical model takes as given the classification of firms into firm

categories and countries and is essentially descriptive. In particular, although it posits the

existence of differences in responses across firms and countries, it offers no causal or structural

explanation for these differences.

       Tables 5-7 present the results of our ANOVA on Equation (4). In Table 5 we study how

country- and firm-level factors explain the variation in log firm size. In col. 1 we examine the

effect of age in the full sample and in cols. 2-7 we examine the effect of different factors in

different age bins. In the interest of clarity, we aggregate the nine age bins examined in Table 2



                                                                                                      23
into 3 age groups – young firms (<5), mid-age (5-19), and mature firms (20-39). We obtain

similar results if we were to classify mature firms as those that are 20+ years old rather than as

20-39 years old.

       We begin with a benchmark specification in which we use country dummies to model

institutional variation at the country level. This specification provides us with the upper bound

for the variation that can be explained at the country level. In alternate specifications, we replace

the country dummy with an institutional variable. In each case, we calculate the increment to

adjusted R-square with effects introduced in the following order: country, firm, and country-firm

interactions. Table 5 presents the contribution to the adjusted R-square when we enter each of the

institutional variables one at a time.

       When we look at the full sample in Col. 1, we see that country dummies explain 9.2% of

the variation in log firm size. Of the institutional variables we examine, Ethnic fractionalization

explains 2.3% of the variation in firm size and Legal Origin and Latitude each explain less than

1%. Together, the four institutional variables explain 2.9% of the variation in log firm size which

is 31.52% (=2.9/9.2) of the explainable variation at the country level. When we look across age

bins in cols. 2-4, we see that the explanatory power of country dummies ranges from 8.9% of the

variation in log firm size in the mid-age firms to 12.5% of the variation in log firm size in young

firms. The high explanatory power of ethnic fractionalization in the full sample seems to be

restricted to the sample with young firms (3.7%) and mid-age firms (2.1%). When entered

together, legal origin, ethnic fractionalization, and latitude explain 36% (=0.045/0.125) of the

explainable variation in log size among young firms, 37.08% of the explainable variation in log

size among mid-age firms and 23.91% of the variation in log size among mature firms. In cols.

4-6 we include settler mortality as one of the explanatory variables which reduces our sample



                                                                                                     24
size. Here again we see that ethnic fractionalization has the largest explanatory power in the

young firm (9.7%) and mid-age samples (4.1%). In the reduced sample, the institutional

variables together explain 67.91% of the explainable variation in log firm size in the <5 age bin,

71.43% of the explainable variation in log size among the mid-age firms and 62.71% of the

explainable variation in log size among the mature firms.

        At the firm-level, in the overall sample, age explains 6.8% of the variation in log size,

sector dummies explain 4.6%, location dummies explain 0.5%, ownership dummies explain

3.9%, legal organization dummies explain 10.1%, and log size at birth explains 35.7% of the

variation in log firm size at any point in time. Thus, size at birth explains the largest variation in

firm size at any given point in the lifecycle, far greater than any other country or firm-level

characteristics. Cols. 2-4 confirm this finding across different age groups - the size at the time the

firm started operations explains between 27.2% (for mature firms) to 52.2% (for the <5 age bin)

of the variation in log firm size. This also holds true in a reduced sample of firms in cols. 5-7

where we have data on settler mortality at the country-level.

        When we look at the other firm-level factors, in the <5 age bin, we find ownership

dummies, sector dummies, and legal organization dummies to be similar in their explanatory

power (5%). In the 5-19 and 20-39 age bins, legal organization dummies have the largest

explanatory power after size at birth, explaining 8.6% and 15.4% of the variation in log size.

        Overall, Table 5 shows that in explaining firm size over the lifecycle the variable with the

highest explanatory power is a firm-level factor – size at birth – rather than a country level

institutional variable.

        In Table 6, we study how country- and firm-level factors explain the variation in size

ratio i.e. how large the firms have grown since birth. Here we find that country dummies explain



                                                                                                    25
much less of the variation in growth than they do for log size. In the full sample in col.1, country

dummies explain only 2.3% of the variation in firm size and when we look across age groups this

ranges between 1.2% for the <5 age bin in col.2 to 2.9% for the mature firms in col. 4. When we

replace the country dummies with institutional variables, individually and together, they explain

less than 1% of the variation in size ratio. Together the institutional variables explain between

31-39% of the explainable variation in growth at different points along the lifecycle. In the

reduced sample with settler mortality, again the country dummies and institutional variables by

themselves do not have very high explanatory power. As a percentage of explainable variation at

the country-level, legal origin, ethnic fractionalization, latitude, and settle mortality together

explain 22% (=0.4/1.8) of the explainable variation in size ratio for young firms, 67% of the

explainable variation in size ratio for mid-age firms, and 62% of the explainable variation in size

ratio for mature firms.

       At the firm-level, again log size at birth is the most dominant factor in explaining

variation in size ratios over the lifecycle, explaining between 3.2% (<5 age bin) to 4.8% (20-39

age bin) of the variation. The other firm-level characteristics explain less than 1% of the

variation in size ratios for young, mid-age, and mature firms. When put together, firm

characteristics explain 4.6% of the variation in size ratio for young firms, 5.9% of the variation

in size ratio for mid-age firms and 7.5% of the variation in size ratio for mature firms. In a

reduced sample of firms in cols. 4-6, firm characteristics explain 5.6% of the variation in size

ratio for young firms, 5.8% of the variation in size ratio for mid-age firms and 8.6% of the

variation in size ratio for mature firms. Thus Table 6 shows that while size at birth is still the

dominant factor in explaining variation in firm growth in comparison to other characteristics, in




                                                                                                     26
absolute terms, the explanatory power is less than 4% in the full sample suggesting that neither

firm-level nor country-level factors explain much of the variation in firm growth since birth.

       In Table 7, we repeat the analysis in Tables 5 and 6 but look at variation in (log) labor

productivity. In col. 1 we see that country dummies explain 52.4% of the variation in labor

productivity. Across age groups, we see that the explanatory power of country dummies ranges

from 52% for mid-age and mature firms to 55.1% for young firms. Thus the maximum variation

that can be explained at the country-level is much higher for labor productivity than for size or

growth.

       When we replace the country dummies with institutional factors, legal origin has the

largest explanatory power over latitude and ethnic fractionalization explaining 4.8% of the

variation in labor productivity in the full sample, 8.6% of the variation in labor productivity for

young firms, 5.1% of the variation in labor productivity for mid-age firms and 3.2% for mature

firms. Together the institutional factors explain 6.9% of the variation in labor productivity in the

full sample which is only 13% (=6.9/52.4) of the maximum explainable variation at the country-

level. We also see that the institutional factors explain a larger variation in labor productivity for

young firms (11.8%) than for mature firms (7.6%). We see similar patterns in a reduced sample

with settler mortality where the institutional factors explain between 25.1% of the variation in

labor productivity for young firms, 20.6% for mid-age firms and 12.4% for mature firms. Legal

origin is again the dominant factor for young and mid-age firms whereas settler mortality is the

dominant factor for mature firms. These results show that institutional differences really matter

for explaining variations in labor productivity especially for young firms.

       At the firm level, size at birth, which had the largest explanatory power in explaining size

and growth since birth, explains less than 1% of the variation in labor productivity at different



                                                                                                    27
points along the lifecycle. For young firms, legal organization explains 4.5% of the variation in

labor productivity followed by location (3.6%) and industry sector (3.1%). For mid-age and

mature firms, legal organization again has the highest explanatory power – 1.6% and 2.7%

respectively – whereas all other firm characteristics explain less than 1% in these two age bins.

Together, the firm-level characteristics explain 10.5% of the variation in labor productivity in

young firms, 5.5% of the variation in labor productivity in mid-age firms, and 6% of the

variation in labor productivity in mature firms.

         Overall Tables 5-7 show that firm-level factors dominate country-level institutional

factors in explaining firm size and growth since birth over the lifecycle whereas country-level

factors are more important in explaining variation in labor productivity over the lifecycle. The

most dominant firm-level factor in explaining size and growth since birth is size at birth,

highlighting the importance of founding conditions in studying size and growth over firm

lifecycle. Size at birth plays a minor role in explaining variation in labor productivity.

         In unreported tests we explore the importance of country-size at birth interaction effects.

We find that this interaction explains 2.3% of the variation in size for young firms and decreases

with age explaining less than 1% of the variation in firm size for old firms. Thus size at birth is

an important explanatory variable across all countries.27




27
   The interaction of country dummies and size at birth dummies explains 3.7% of the variation in size ratios for
young firms, 0.4% of the variation in size ratios for mid-age firms, and 1.3% of the variation in size ratios for mature
firms. When we look at labor productivity, country dummies x size at birth dummies explain 1.9% of the variation
in labor productivity for young firms, 0.9% of the variation in labor productivity for mid-age firms and 0.4% of the
variation in labor productivity for mature firms.

                                                                                                                    28
V. Size-Productivity Covariance across Firm Lifecycle

        The previous two sections have shown that across developing countries, older firms are

larger and grow as they age, but are not much more productive than younger firms. We also find

that initial starting conditions specifically size at birth, plays a significant role in explaining the

variation in size and growth since birth. However, country-level institutional factors dominate

firm-level factors in explaining variations in labor productivity.

        In this section, we try to understand to what extent our results may be driven by selection

processes. A recent literature in economics has argued that the large differences in productivity

between rich and poor countries can be explained by heterogeneity in firm-level productivity

which can, in turn, be attributed to the resource mis-allocation in developing countries

(Bartelsman, Haltiwanger, Scarpetta (2004), Hsieh and Klenow (2009), Alfaro, Charlton, and

Kanczuk (2008), and Midrigan and Xu (2014)). Bartelsman, Haltiwanger, and Scarpetta (2013)

show, that the within-industry covariance between size and productivity is a robust measure of

this mis-allocation. The underlying logic behind this measure is as follows: In the absence of any

distortions, the traditional models of firm size distribution (e.g. Lucas, 1978 and Melitz, 2003)

predict a positive correlation between size and productivity so that larger firms are more

productive. However distortions in developing countries affect both resource misallocation (too

many resources are devoted to small unproductive firms) and selection processes (highly

productive firms may exit and low productivity firms may be allowed to operate) which lead to a

great deal of variation in the size-productivity relation across countries. This variation is then

captured by the cross country variation in the covariance between size and productivity.

        We examine how the strength of this covariance varies with firm lifecycle and which

institutional and policy variables explain the variation in this covariance. We supplement this



                                                                                                      29
measure with a standard deviation of labor productivity to assess the role of misallocation. We

focus on labor productivity because Bartelsman et al. show that the within-industry dispersion of

labor productivity is much larger than the within-industry dispersion of total factor productivity.

        In constructing the covariance term, we follow Bartelsman et al. (2013) who exploit the

cross-sectional decomposition of productivity developed by Olley and Pakes (1996).

Specifically, Olley and Pakes (1996) decompose total industry productivity (weighted average of

firm-level productivity with shares of industry output as weights) into the unweighted average of

plant level productivities and the sample covariance between productivity and output share. We

implement the Olley-Pakes (OP) decomposition at the country-age bin level. That is, for firm i

belonging to age bin a, if firm-level productivity is LPia, and wia is firm i’s share of employment

in age bin a, then total productivity in that age-bin is given by:




This can be decomposed into two terms as:


                                                ∆    	       		   ∆   	



                                                         ∆    ∆




where     	    	    represent unweighted mean share and unweighted mean productivity

respectively. The second term is the OP covariance measure.

        In our data, we construct the covariance measure in each country-age bin as the

difference between the weighted average labor productivity in the country-age bin and the

unweighted average labor productivity.

                                                                                                  30
        Table 8 presents the summary statistics for the standard deviation of labor productivity

and for the covariance term. The summary statistics show that the mean and median standard

deviation in log labor productivity increase with age suggesting that the heterogeneity in labor

productivity increases with firm age. When we look at the size-productivity covariance term, we

find that the mean and median across countries and ages is negative suggestive of resource

misallocation and selection issues.28 For instance, the median covariance term of -0.858 for

young firms suggests that the labor productivity in the median country in the average young firm

is 86% lower than it would be if employment shares were randomly allocated. However, the

mean and median values of the OP covariance term also seem to become less negative with age

indicating that the older firms in our sample are more efficient.29 For instance, the median

covariance term of -0.263 for mature firms suggests that the labor productivity in the median

country in the average mature firm is 26% lower than it would be if employment shares were

randomly allocated. The increase in the size-productivity covariance over the lifecycle provides

suggestive evidence that the size-age profiles established in sections A and B are partly driven

by the selection of more efficient firms.



VI. Conclusion

        The role of size and age in influencing firm dynamics is crucial to understanding

productivity differences across firms and countries. In this paper, we use unique survey data

across 118 developing countries to examine the relationship between firm size and age in the



28
   For comparison sake, the OP covariance term for labor productivity reported by Bartelsman et al. (2013) for the
United States is 0.51 which implies that the industry index of labor productivity in the average US manufacturing
industry is 50% higher than it would be if employment shares were randomly allocated within industries.
29
   For comparison sake, the OP covariance term for labor productivity reported by Bartelsman et al. (2013) for the
United States is 0.51 which implies that the industry index of labor productivity in the average US manufacturing
industry is 50% higher than it would be if employment shares were randomly allocated within industries.

                                                                                                                 31
formal sector. We find that the average 40 year old firm in developing countries is about 5 times

larger than the average firm less than 5 years of age. Our results are robust to a number of checks

in smaller samples including looking at just the manufacturing sector, dropping former socialist

economies, dropping firms with any government ownership and dropping firms that started

operations before the countries’ year of independence.

        We then try to understand the factors that explain firm size, growth, and productivity over

the firm lifecycle using a variance decomposition analysis, which allows us to compare country

versus firm-level factors without positing a structural or causal explanation. Our results show

that while country institutional factors are important in explaining labor productivity, firm-level

factors are comparable and even dominate country-level variables in explaining size and growth

over the lifecycle. In particular, initial size at birth has significant explanatory power in

explaining variation in firm size and firm size ratios of firms.

        To understand the selection forces at play and the extent of resource mis-allocation in

developing countries, we study the size-productivity covariance at different ages. We find that

the resource mis-allocation is the worst for young firms and decreases with age. Hence, older

firms are larger, partly due to selection of more efficient firms.

        In summary, the paper shows that across the world, on average, new firms in the formal

sector start small and grow as they age. We find no evidence of declining lifecycles in the formal

sector, suggesting that older firms are important contributors to employment. Initial conditions of

a firm, specifically its size at birth, play an important role in explaining variation in size and

growth over firm lifecycle. Productivity differences, however, are explained mostly by

institutional differences.




                                                                                                     32
References

Acemoglu, D., Johnson, S., Robinson, J. A., 2001, “The Colonial Origins of Comparative
Development: An empirical investigation,” American Economic Review, 91, 1369-1401.

Acemoglu, D., Johnson, S., Robinson, J. A., 2002, “Reversal of fortunes: Geography and
Institutions in the making of the modern world income distribution,” Quarterly Journal of
Economics, 117 (4), 1231-1294.

Acemoglu, D., Johnson, S., 2005. “Unbundling Institutions” Journal of Political Economy
113(5), 949-995.

Agarwal, R., Audretsch, D. B., 2001. “Does Entry Size Matter? The Impact of the Life Cycle and
Technology on Firm Survival” Journal of Industrial Economics 49(1), 21-43.

Alesina, A., Devleeschauwer, A., Easterly, W., Kurlat, S., Wacziarg, R., 2003,
“Fractionalization,” Journal of Economic Growth, 8(2).

Alfaro, L., Charlton, A., Kanczuk, F., 2008. “Plant-Size Distribution and Cross-Country Income
Differences.” National Bureau of Economic Research Working Paper Series No. 14060

Almazan, A., De Motta, A. D., Titman, S., Uysal. V., 2010, “Financial Structure, Acquisition
Opportunities, and Firm Locations” Journal of Finance 65(2), 529-563.

Arikan, A. M., Stulz, R., 2011. “Corporate Acquisitions, Diversification, and the Firm’s
Lifecycle.” NBER Working Paper No/ 17463.

Asker, J., Farre-Mensa, J., Ljungqvist, A., 2015. “Corporate Investment and Stock Market
Listing: A Puzzle” Review of Financial Studies 28(2), 342-390.

Ayyagari, M., Demirguc-Kunt, A., Maksimovic, V., 2008. How well do Institutional Theories
Explain Firms Perception of Property Rights, Review of Financial Studies 21(4), 1833-1871.

Ayyagari, M., Demirguc-Kunt, A., Maksimovic, V., 2010. “Formal versus Informal Finance:
Evidence from China.” Review of Financial Studies, 23(8).

Ayyagari, M., Demirguc-Kunt, A., Maksimovic, V., 2014a. “Who creates jobs in developing
countries?” Small Business Economics 43(1), 75-99.

Ayyagari, M., Demirguc-Kunt, A., Maksimovic, V., 2014b. “What determines entrepreneurial
outcomes in emerging markets? The Role of Initial Conditions ” SSRN Working Paper Series.

Bartelsman, E., Haltiwanger, J., Scarpetta, S., 2004. "Microeconomic evidence of creative
destruction in industrial and developing countries." The World Bank, Policy Research Working
Paper No. 3464, December.



                                                                                               33
Bartelsman, E., Haltiwnager, J., and Scarpetta, S., 2013. “Cross-Country Differences in
Productivity: The Role of Allocation and Selection.”American Economic Review 103(1): 305-
334.

Bargeron, L. L., Schlingemann, F. P., Stulz, R. M., Zutter, C. J., 2008. “Why do private acquirers
pay so little compared to public acquirers?” Journal of Financial Economics 89(3), 375-390.

Barth, J., Lin, C., Lin, P., Song, F., 2009. “Corruption in bank lending to firms: cross-country
micro evidence on the beneficial role of competition and information sharing.” Journal of
Financial Economics 91, 361-388.

Beck, T., Demirguc-Kunt, A., Laeven, L., Levine, R., 2008. "Finance, Firm Size, and Growth,"
Journal of Money, Credit and Banking, vol. 40(7), 1379-1405.

Beck, T., Demirguc-Kunt, A., Maksimovic, V., 2005, Financial and Legal Constraints to Firm
Growth: Does Firm Size Matter? Journal of Finance, 137-177

Beck, T., Levine, R., 2002. "Industry Growth and Capital Allocation: Does Having a Market- or
Bank Based System Matter?" Journal of Financial Economics, 64 (2): 147-180.

Beck, T., Lin, C., Ma, Y., 2014. “Why do Firms Evade Taxes? The Role of Information Sharing
and Financial Sector Outreach.” Journal of Finance 69(2), 763-817.

Berger, A. N., Gregory F. U., 1990. "Collateral, Loan Quality, and Bank Risk.” Journal of
Monetary Economics 25, 21-42..

Botero, J. C., Djankov, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2004. “The
Regulation of Labor” Quarterly Journal of Economics, 119 (4), 1339-1382.

Boubakri, N., Cosset, J., 1998. “The financial and operating performance of newly privatized
firms: Evidence from developing countries”, Journal of Finance 53, 1081– 1110

Brav, A., Graham, J. R., Harvey, C., Michaely, R., 2005. Payout Policy in the 21st Century,
Journal of Financial Economics 77(3), 483-527.

Brav, O. 2009. “Access to Capital, Capital Structure, and the Funding of the Firm” Journal of
Finance 64(1), 263-308.

Cabral, L., Mata, J., 2003. "On the Evolution of the Firm Size Distribution: Facts and
Theory," American Economic Review 93(4): 1075-90.

Campello, M., Giambona, E., Graham, J. R., Harvey, C. R., 2011. “Liquidity Management and
Corporate Investment During a Financial Crisis,” Review of Financial Studies 24(6): 1944-1979.

Carvalho, D., 2014. “The real effects of government-owned banks: Evidence from an emerging
market.” The Journal of Finance 69 (2), 577{609.

                                                                                                   34
Claessens, S., Laeven, L., 2003. "Financial Development, Property Rights, and Growth." Journal
of Finance, 58(6): 2401-2436.

Cole, S., 2009. “Financial Development, Bank Ownership, and Growth: Or, Does Quantity
Imply Quality?” Review of Economics and Statistics 91, 33-51.

Cull, R., Xu, L. C., 2005. “Institutions, Ownership, and Finance: The Determinants of
Investment among Chinese Firms,” Journal of Financial Economics 77, 117-146.

Davis, S., Haltiwanger, J., Schuh, S., 1996. Job Creation and Destruction.
Cambridge, MA: MIT Press.

DeAngelo, H., DeAngelo, L., Stulz, R. 2006. Dividend policy and the earned/contributed capital
mix: A test of the lifecycle theory. Journal of Financial Economics 81:2, 227–54

Demirguc-Kunt, A., Love, I., Maksimovic, V., 2006. “Business environment and the
incorporation decicion.” Journal of Banking and Finance 30(11), 2967-2993.

Demirguc-Kunt, A., Maksimovic, V., 1998. “Law, Finance, and Firm Growth” Journal of
Finance 53(6), 2107-2137.

Denis, D. J., Osobov, I., 2008. Why do firms pay dividends? International evidence on the
determinants of dividend policy. Journal of Financial Economics 89(1), 62–82.

Dewenter, K. L., Malatesta, P. H., 2001. “State-owned and privately owned firms: An empirical
analysis of profitability, leverage, and labor intensity”, American Economic Review 91, 320–334

Dinc, S. 2005. “Politicians and banks: Political influences on government-owned banks in
emerging markets.” Journal of Financial Economics 77 (2), 453 { 479.

Dinc, S., Gupta, N., 2011. “The Decision to Privatize: Finance and Politics” Journal of Finance
66(1), 241-269.

Djankov, S., McLiesh, C., Shleifer, A., 2007. “Private Credit in 129 Countries” Journal of
Financial Economics 12(2), 77-99.

Djankov, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2002. “The Regulation of Entry”
Quarterly Journal of Economics, 117 (1), 1-37.

Djankov, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2003. “Courts,” Quarterly Journal
of Economics, 118 (2), 453-517.

Doidge, C., Karolyi, A. G., Stulz, R., 2007. “Why do countries matter so much for corporate
governance?” Journal of Financial Economics 86, 1-39.

                                                                                                 35
Dunne, T., Roberts, M., Samuelson, L., 1989. "The Growth and Failure of U.S.
Manufacturing Firms," Quarterly Journal of Economics 104(4), 671-98.

Easterly, W. Levine, R., 1997, “Africa’s growth tragedy: policies and ethnic divisions,”
Quarterly Journal of Economics, 112, 1203-1250.

Engerman, S., Sokoloff, K., 1997, “Factor endowments, institutions, and differential paths of
growth among new world economies,” in S.H. Haber (ed.), How Latin America Fell Behind,
Stanford University Press, Stanford CA, pp. 260-304.

Ericson, R., Pakes, A., 1998. “Empirical implications of alternative models of firm dynamics”
Journal of Economic Theory 79(1), 1-45.

Evans, David S, 1987, The Relationship between Firm Growth, Size, and Age: Estimates for 100
Manufacturing Industries, Journal of Industrial Economics 35(4), 567-581.

Fama, E. F., French, K. R., 2001. Disappearing dividends: Changing firm characteristics or lower
propensity to pay? Journal of Financial Economics 60(1), 3–43

Farre-Mensa, J., 2012. Comparing the cash policies of public and private firms. Unpublished
working paper. Harvard University

Foster, L., Haltiwanger, J. C., Syverson, C., 2012. The Slow Growth of New Firms: Learning
about Demand? National Bureau of Economic Research Working Papers No. 17853.

Gao, H., Harford, J., Li, K., 2013. “Determinants of corporate cash policy: Insights from private
firms” Journal of Financial Economics 109, 623-639.

Geroski, P. A. 1995. “What do we know about entry?” International Journal of Industrial
Organization 13(4), 421-440.

Glaeser, E., Kallal, H., Scheinkman, J., Shleifer, A., 1992. “Growth in Cities” Journal of
Political Economy 100, 1126-1152.

Graham, J. R., Harvey, C. R., 2001. “The Theory and Practice of Corporate Finance: Evidence
from the Field”, Journal of Financial Economics 60, 187-243

Grullon, G., Michaely, R., Swaminathan, B., 2002. Are dividend changes a sign of firm
maturity? Journal of Business 75(3), 387–424.

Gupta, N., 2005. “Partial Privatization and Firm Performance” Journal of Finance Vol. LX(2),
987-1015.

Hadlock, C., Pierce, J. R. 2010. “New Evidence on Measuring Financial Constraints: Moving
Beyond the KZ Index.” Review of Financial Studies 23(5), 1909-1940.

                                                                                                36
Harris, J. R., Todaro, M. P., 1970. “Migration, Unemployment and Development: A Two‐ Sector
Analysis.” American Economic Review 60, no. 1: 126‐42.

Holmes, T., 1999. “Localization of Industry and Vertical Disintegration” Review of Economics
and Statistics 81(2), 314-325.

Hsieh, C. T., Klenow, P., 2009. “Misallocation and Manufacturing TFP in China and India.”
Quarterly Journal of Economics 124, 1403-1448.

Hsieh, C. T., Klenow, P., 2014."The Lifecycle of firms in India and Mexico,”
Quarterly Journal of Economics 129(3), 1035-1084.

John, K., Knyazeva, A., Knyazeva, D., 2011. “Does geography matter? Firm location and
corporate payout policy” Journal of Financial Economics 101(3), 533-551.

Jovanovic, B. 1982. “Selection and evolution in industry” Econometrica 50, 649-670.

Kang, J., Kim, J., 2008. “The Geography of Block Acquisitions” Journal of Finance 63(6), 2817-
2858.

Kerr, W., Nanda, R., 2009. “Democratizing Entry: Banking Deregulations, Financing Constraints,
and Entrepreneurship.” Journal of Financial Economics 94(1), 124-149.

Khanna, T., Rivkin, J., 2001. “Estimating the Performance Effects of Business Groups in
Emerging Markets,” Strategic Management Journal 22(1), 45-74.

Khwaja, A. I., Mian, A., 2005. “Do Lenders Favor Politically Connected Firms? Rent Provision
in an Emerging Financial Market” Quarterly Journal of Economics 120(4), 1371-1411.

Klapper, L., L. Laeven, Rajan, R., 2006. “Entry regulation as a barrier to entrepreneurship.”
Journal of Financial Economics 82, 591-629.

Klepper, 2001. “Employee Startups in High-Tech Industries” Industrial & Corporate Change
10(3), 639-67

LaPorta, R., Lopez-de-Silanes, F., 1999. “Benefits of privatization—Evidence from Mexico”,
Quarterly Journal of Economics 114, 1193–1242.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998, “Law and finance,” Journal
of Political Economy, 106, 1113-1155

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. “Corporate Ownership around the World,”
Journal of Finance 54 (2), 471-517.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2002. “Government ownership of banks,”

                                                                                                37
Journal of Finance 57, 265-301.

La Porta, R., Shleifer, A., 2008. "The Unofficial Economy and Economic Development,"
Brookings Papers in Economic Activity, 275-352.

Lins, K. V., Servaes, H., Tufano, P., 2010, “What drives corporate liquidity? An international
survey of cash holdings and lines of credit”, Journal of Financial Economics 98(1), 160-176.

Loughran, T., Schultz, P., 2005. “Liquidity: Urban versus Rural Firms” Journal of Financial
Economics 78(2), 341-374.

Lucas, R. E. 1978. “On the size distribution of business firms.” Bell Journal of Economics 9,
508-523.

Maksimovic, V., Phillips, G. 2008. “The Industry Life Cycle, Acquisitions and Investment: Does
Firm Organization Matter? Journal of Finance 63(2), 673-708.

Maksimovic, V., Phillips, G., Yang, L. 2013. “Private and Public Merger Waves” Journal of
Finance 68(5), 2177-2217.

Marshall, A., 1890. Principles of Economics. London: Macmillan.

McGahan, A. M., Porter, M.E., 1997. “How Much Does Industry Matter, Really?” Strategic
Management Journal, 18, 15-30.

McGahan, A. M., Porter, M.E., 2002. “What Do We Know About Variance in Accounting
Profitability?” Management Science, 48 (7), 834-851.

Megginson, W. L., Netter, J. 2001. “From state to market: A survey of empirical studies on
privatization,” Journal of Economic Literature 39, 321–389.

Michaely, R. and M. R. Roberts. 2012. “Corporate Dividend Policies: Lessons from Private
Firms” Review of Financial Studies 25(3).

Midrigan, V., Xu, D. Y., 2014. “Finance and Misallocation: Evidence from Plant-Level Data”
American Economic Review 104(2), 422-458.

Melitz,M., 2003. “The Impact of Trade of Intra-Industry Reallocations and Aggregate Industry
Productivity” Econometrica 71(6), 1695-1725.

Olley, S. G., Pakes, A., 1996. “The Dynamics of Productivity in the Telecommunications
Equipment Industry” Econometrica 64(6). 1263-1297.

Ono, Y., 2003. “Outsourcing Business Services and the Role of Central Administrative Offices”
Journal of Urban Economics 53(3), 377-395.



                                                                                                 38
Pagano, P. and Schivardi, F. (2003). Firm Size Distribution and Growth. Scandinavian Journal
of Economics 105(2), 255-274.

Pastor, L., Veronesi, P., 2003. Stock valuation and learning about profitability. Journal of
Finance 58(5), 1749-1790.

Rajan, R. and Zingales, L., 1998, Financial Dependence and Growth. American Economic
Review 88 (3), 559-586.

Rauch, J. E., 1991. “Modelling the informal sector informally.” Journal of Development
Economics 35, 33-47.

Rumelt, R., 1991. “How Much Does Industry Matter?” Strategic Management Journal, 12 (3),
167-185.

Sapienza, P., 2004. “The effects of government ownership on bank lending.” Journal of
Financial Economics 72 (2), 357-384.

Schmalensee, R., 1985. “Do Markets Differ Much?” American Economic Review, 75 (3), 341-
351.

Sutton, J., 1997. “Gibrat’s legacy.” Journal of Economic Literature 35, 40-59.

Sheen, A., 2009. “Do Public and Private Firms Behave Differently? An Examination of
Investment in the Chemical Industry” Working Paper UCLA.

Shleifer, A., 1998. “State versus Private Ownership” Journal of Economic Perspectives 12(4),
133-150.

Shleifer, A., Vishny, R., 1994. “Politicians and Firms” Quarterly Journal of Economics 109(4),
995-1025.

Uysal, V., Kedia, S., Panchapagesan, V., 2008. “Geography and Acquirer Returns” Journal of
Financial Intermediation 17(2), 256-275.




                                                                                               39
Figure 1: Firm Employment by Age – Sample Estimates in 120 Developing Countries
              6        5
                                                          Sample Estimates
   Employment(Age<5=1)
        3     2
              14




                           <5        5-9       10-14       15-19       20-24      25-29   30-34   35-39   40+
                                                                        Age
                           Source: Enterprise Surveys across 120 developing countries




Figure 2: Firm Employment by Age - Population Estimates in 120 Developing Countries

                                                        Population Estimates
              6        5
   Employment(Age<5=1)
        3     2
              14




                           <5        5-9       10-14       15-19       20-24      25-29   30-34   35-39   40+
                                                                        Age
                           Source: Enterprise Surveys across 120 developing countries



                                                                                                                40
Figure 3: Firm Employment by Age – Manufacturing in 120 developing countries
              8      6
                                                             Manufacturing
   Employment(Age<5=1)
   2       4  0




                         <5         5-9       10-14       15-19        20-24     25-29      30-34      35-39        40+
                                                                        Age
                         Source: Enterprise Surveys across 120 developing countries




Figure 4: Firm Employment by Age – Robustness Tests
            5
   Employment(Age<5=1)
    2      31       4




                         <5        5-9        10-14      15-19         20-24    25-29      30-34       35-39        40+
                                                                        Age
                                     Drop former Socialist countries                     Drop firms with govt. ownership
                                     Drop firms formed before independence

                         Source: Enterprise Surveys across 120 developing countries




                                                                                                                           41
Figure 5: Firm Employment by Age - Normalize by Average Employment at Birth
                                                                8
   Avg Employment of cohort/Avg Employment of cohort at birth
                                                                6
                                                                4
                                                                2




                                                                    <5        5-9        10-14      15-19       20-24      25-29    30-34   35-39   40+
                                                                                                                 Age
                                                                    Source: Enterprise Surveys across 120 developing countries




Figure 6: Labor Productivity by Age
                                                                5
   Labor Productivity(Age<5=1)
      2          3          4                                   1




                                                                    <5         5-9       10-14      15-19       20-24       25-29   30-34   35-39   40+
                                                                                                                 Age
                                                                    Source: Enterprise Surveys across 120 developing countries




                                                                                                                                                          42
             Table 1: Summary Statistics and Correlations – Cross-country Sample
The variables are defined as follows: Firm Size is the number of permanent, full-time employees in the firm where
permanent, full-time employees are defined as all paid employees that are contracted for a term of one or more fiscal
years and/or have a guaranteed renewal of their employment contract and that work 8 or more hours per day. Firm
Size Ratio is the firm size of each firm scaled by the number of full-time employees when it started operations.
Firm Age is defined as the number of years since the firm began operations in the country. Manufacturing is a
dummy variable that takes the value 1 for firms in the manufacturing sector and 0 otherwise. Labor Productivity
(Log) is the logarithm of the ratio of total annual sales (in USD) to number of permanent, full time employees.



Panel A: Summary Statistics
                                                                Std.
  Variable                               N         Mean         Dev.             Min         Max
  Firm Size                            44696       59.35        125.02             3         860
  Log (Firm Size)                      44696        3.09         1.27             1.10      6.76
  Firm Size Ratio                      38932        5.64        18.15            0.006       860
  Firm Age                             44696       15.73        15.37              0         209
  Manufacturing                        44676        0.55         0.50              0          1
  Labor Productivity (Log)             31020       10.08         1.95             5.81      17.36



Panel B: Correlations
                                                    Log (Firm
                                                                    Firm Size
                                   Firm Size          Size)                              Firm Age    Manufacturing
                                                                      Ratio
Log (Firm Size)                    0.760***
Firm Size Ratio                    0.351***         0.312***
Firm Age                           0.259***         0.309***           0.134***
Manufacturing                      0.103***         0.162***             0.007           0.141***
Labor Productivity(Log)            0.084***         0.146***           0.046***          0.090***      -0.031***
*, **, and *** represent significance at 10%, 5%, and 1% levels respectively




                                                                                                                   43
                                          Table 2: Firm Size and Age
This table shows results from the following regression: Firm Size = α + β1 Age Dummies + β2 Industry Dummies + β3Country
Dummies + β4Year Dummies + e. Firm Size is the number of permanent, full-time employees in the firm where permanent,
full-time employees are defined as all paid employees that are contracted for a term of one or more fiscal years and/or have a
guaranteed renewal of their employment contract and that work 8 or more hours per day. Age Dummies consist of 9 age
dummies for the following age bins: <5(reference category), 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, and 40+.
Manufacturing is a dummy variable that takes the value 1 for firms in the manufacturing sector and 0 otherwise. Year
Dummies consist of dummies for each of the years in which the Enterprise Survey is conducted. All regressions are estimated
by ordinary least squares. Robust standard errors are reported in parentheses. The square brackets in col. 1 contains a ratio of
the mean size of the particular age bin to the mean size in the age bin with youngest (<5) firms.
                                   1                      2                    3                      4
                               Firm Size             Firm Size             Firm Size              Firm Size
                                                                         Manufacturing      Non-Manufacturing
5-9                          12.805***              11.081***             14.584***                6.626***
                               (1.178)               (1.178)               (1.948)                  (1.389)

                               [1.452]
10-14                        28.611***              22.272***             30.422***               12.945***
                               (1.453)               (1.464)               (2.356)                  (1.751)
                               [2.010]
15-19                        35.393***              28.656***             32.279***               24.981***
                               (1.782)               (1.832)               (2.687)                  (2.504)
                               [2.250]
20-24                        35.030***              32.225***             38.536***               25.506***
                               (2.412)               (2.488)               (3.302)                  (3.824)
                               [2.237]
25-29                        42.667***              41.593***             46.368***               36.210***
                               (2.969)               (3.049)               (3.927)                  (5.029)
                               [2.507]
30-34                        60.010***              55.428***             61.000***               49.566***
                               (4.148)               (4.147)               (5.358)                  (6.383)
                               [3.119]
35-39                        65.241***              63.114***             73.153***               48.247***
                               (4.808)               (4.811)               (6.395)                  (7.076)

                               [3.304]
40+                          127.496***            120.580***            135.790***               94.946***
                               (3.852)               (3.895)               (5.004)                  (6.324)
                               [5.502]
Constant                      28.318***             27.983***             53.847***                  1.734
                               (0.758)               (4.703)               (17.833)                 (2.987)

Fixed Effects                   None         Country, Industry, Year    |------------- Country, Year --------------|
N                               44696                 44676                 24564                    20112
Adjusted R-square               0.064               0.115                 0.108                      0.094
*, **, and *** represent significance at 10%, 5%, and 1% levels respectively.




                                                                                                                              44
                                        Table 3: Firm Growth and Age
This table shows results from the following regression: Firm Size Ratio = α + β1 Age Dummies + β2 Industry Dummies +
β3Country Dummies + β4Year Dummies + e. Firm Size Ratio is the firm size of each firm scaled by the number of full-time
employees when it started operations. Age Dummies consist of 9 age dummies for the following age bins: <5(reference
category), 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, and 40+. Manufacturing is a dummy variable that takes the value 1
for firms in the manufacturing sector and 0 otherwise. Year Dummies consist of dummies for each of the years in which the
Enterprise Survey is conducted. All regressions are estimated by ordinary least squares. Robust standard errors are reported in
parentheses.


                                 1                        2                         3                        4

                          Firm Size Ratio          Firm Size Ratio           Firm Size Ratio         Firm Size Ratio


                                                                             Manufacturing         Non-Manufacturing


5-9                          0.868***                 0.686***                   0.418*                  0.870***
                              (0.146)                  (0.156)                   (0.248)                  (0.202)
10-14                        2.690***                 1.957***                  1.439***                 2.458***
                              (0.219)                  (0.227)                   (0.284)                  (0.359)
15-19                        4.531***                 3.719***                  2.619***                 5.035***
                              (0.331)                  (0.336)                   (0.394)                  (0.575)
20-24                        4.061***                 3.434***                  3.387***                 3.203***
                              (0.399)                  (0.413)                   (0.576)                  (0.519)
25-29                        4.438***                 4.026***                  3.718***                 4.413***
                              (0.544)                  (0.574)                   (0.733)                  (0.992)
30-34                        5.031***                 4.315***                  3.715***                 5.296***
                              (0.516)                  (0.519)                   (0.609)                  (1.006)
35-39                        5.622***                 5.186***                  5.817***                 3.681***
                              (0.683)                  (0.675)                   (0.941)                  (0.796)
40+                          9.638***                 8.621***                  9.568***                 6.238***
                              (0.832)                  (0.822)                   (1.130)                  (0.977)
Constant                     2.922***                   0.423                   1.262***                 0.898***
                              (0.112)                  (0.533)                   (0.253)                  (0.185)

Fixed Effects                  None            Country, Industry, Year          |---------- Country, Year ---------|
N                              38932                   38914                      21342                    17572
Adjusted R-square               0.018                  0.039                      0.037                    0.041
*, **, and *** represent significance at 10%, 5%, and 1% levels respectively.




                                                                                                                             45
                              Table 4: Labor Productivity and Age
This table shows results from the following regression: Labor Productivity (Log) = α + β1 Age Dummies + β2
Industry Dummies + β3Country Dummies + β4Year Dummies + e. Labor Productivity (Log) is the logarithm of
the ratio of total annual sales (in USD) to number of permanent, full time employees. Age Dummies consist of 9
age dummies for the following age bins: <5(reference category), 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39,
and 40+. Manufacturing is a dummy variable that takes the value 1 for firms in the manufacturing sector and 0
otherwise. Year Dummies consist of dummies for each of the years in which the Enterprise Survey is conducted.
All regressions are estimated by ordinary least squares. Robust standard errors are reported in parentheses. The
square brackets in col. 1 is the % change in firm size given by 100 x (exp(b)-1) where b is the respective age
coefficient.

                                 (1)                   (2)                      (3)                     (4)
                          Labor Productivity    Labor Productivity       Labor Productivity      Labor Productivity
                                (Log)                 (Log)                    (Log)                   (Log)
  5-9                         0.212***              0.138***                 0.100***                  0.151***
                               (0.037)               (0.025)                  (0.034)                   (0.037)
                             [23.615%]
  10-14                       0.417***              0.207***                 0.151***                  0.246***
                               (0.037)               (0.026)                  (0.035)                   (0.040)
                             [51.740%]
  15-19                       0.460***              0.231***                 0.176***                  0.256***
                               (0.040)               (0.028)                  (0.038)                   (0.044)
                             [58.407%]
  20-24                       0.388***              0.253***                 0.220***                  0.261***
                               (0.049)               (0.036)                  (0.045)                   (0.065)
                             [47.403%]
  25-29                       0.411***              0.241***                 0.194***                  0.313***
                               (0.056)               (0.041)                  (0.049)                   (0.086)
                             [50.833%]
  30-34                       0.445***              0.253***                 0.204***                  0.341***
                               (0.064)               (0.046)                  (0.055)                   (0.091)
                             [56.049%]
  35-39                       0.538***              0.291***                 0.267***                  0.345***
                               (0.076)               (0.054)                  (0.064)                   (0.106)
                             [71.258%]
  40+                         0.762***              0.413***                 0.499***                  0.261***
                               (0.045)               (0.035)                  (0.043)                   (0.066)
                             [114.256%]
  Constant                    9.756***              7.353***                 7.774***                  8.579***
                               (0.029)               (0.114)                  (0.165)                   (0.134)
  Fixed Effects                 None           Country, Industry, Year           |--------Country, Year------------|
  N                             31020                 31014                    17638                    13376
  Adjusted R-sq                 0.011                 0.538                    0.550                     0.493
*, **, and *** represent significance at 10%, 5%, and 1% levels respectively.




                                                                                                                       46
                                                Table 5: Firm Size and Lifecycle – Analysis of Variance
The table documents how firm level variables and country level variables contribute to the adjusted R-square of the following regression model when they are entered one at a time:
Log Firm Size = α + β1 Firm Characteristic Dummies + β2 Country Characteristic Dummies + e. Log Firm Size is the logarithm of the number of permanent, full-time employees in the
firm where permanent, full-time employees are defined as all paid employees that are contracted for a term of one or more fiscal years and/or have a guaranteed renewal of their
employment contract and that work 8 or more hours per day. The country characteristic is captured by using one of the following variables at the country level: Country dummies,
Legal Origin, Ethnic Fractionalization, or Latitude. Legal Origin takes the value one if Common law, two if French civil law, and three if Socialist law countries. Ethnic
Fractionalization is the probability that two randomly selected individuals in a country are not from the same ethnic group. Latitude is the absolute value of the latitude of the country
scaled between zero and one. The firm effect is captured by one of the following variables: Age, Sector, Location (City Size), Legal Organization, Ownership, and Size at birth. Age is
the age of the firm since it started operations. Sector is 14 industrial sector dummies. Location is 5 dummies for capital city, city with population over 1 million, city with population
over 250,000 to 1 million, city with population 50,000 to 250,000, and city with less than 50,000 population. Legal Organization is one of three values that reflects whether the firm is
organized as a publicly listed company, privately held limited liability company or one of the following - sole proprietorship, partnership, or another alternative form. Ownership is one
of four values reflecting whether the owner (>=50% ownership) is the state government; private domestic individuals, companies or organizations; private foreign individuals,
companies or organizations; or another ownership form. Size at birth is the log of firm size when the firm started operations. We rescale Ethnic Fractionalization, Latitude, and Settler
Mortality on a four-point scale using quantiles.
                                                                               1            2              3               4               5              6               7

   Age Groups                                                             All          Young         Mid-Age           Old           Young         Mid-Age           Old
                                                                                        (<5)          (5-19)         (20-39)          (<5)          (5-19)         (20-39)
   Country characteristics                                                                                                                                              
   Country Dummies                                                       0.092          0.125          0.089          0.092          0.187           0.105          0.059
   Legal Origin Dummies                                                  0.008          0.018          0.017          0.014          0.056           0.04           0.015
   Ethnic Fractionalization Dummies                                      0.023          0.037          0.021          0.009          0.097           0.041          0.009
   Latitude Dummies                                                      0.004          0.018          0.009           0.01          0.035           0.01           0.001
   Settler Mortality Dummies                                                                                                         0.044           0.024          0.016


   Legal Origin, Latitude, Ethnic Fractionalization Dummies              0.029          0.045          0.033          0.022                                             
                                                                                                                                                                        
   Legal Origin, Latitude, Ethnic Fractionalization, Settler
   Mortality Dummies                                                                                                                 0.127           0.075          0.037


   Firm-level characteristics                                                                                                                                           
   Age                                                                   0.068
   Sector Dummies                                                        0.046          0.051          0.041          0.026          0.069           0.033          0.017
   Location (City Size) Dummies                                          0.005          0.002          0.002           0.01          0.014           0.009          0.008
   Ownership Dummies                                                     0.039           0.05          0.044          0.026          0.084           0.052          0.038
   Legal Organization Dummies                                            0.101           0.05          0.086          0.154          0.097           0.104          0.148
   Log (Size at birth)                                                   0.357          0.522          0.373          0.272          0.577           0.429          0.273
   All firm-level characteristics together                               0.423          0.511          0.389          0.346           0.62           0.487          0.37
   N                                                                    33982           6119           20144           5724           2234           8352           3896


                                                                                                                                                                                      47
                                             Table 6: Firm Growth and Lifecycle – Analysis of Variance
The table documents how firm level variables and country level variables contribute to the adjusted R-square of the following regression model when they are entered one at a time:
Size Ratio = α + β1 Firm Characteristic Dummies + β2 Country Characteristic Dummies + e. Firm Size Ratio is the firm size of each firm scaled by the number of full-time employees
when it started operations. The country characteristic is captured by using one of the following variables at the country level: Country dummies, Legal Origin, Ethnic Fractionalization,
or Latitude. Legal Origin takes the value one if Common law, two if French civil law, and three if Socialist law countries. Ethnic Fractionalization is the probability that two randomly
selected individuals in a country are not from the same ethnic group. Latitude is the absolute value of the latitude of the country scaled between zero and one. The firm effect is
captured by one of the following variables: Age, Sector, Location (City Size), Legal Organization, Ownership, and Size at birth. Age is the age of the firm since it started operations.
Sector is 14 industrial sector dummies. Location is 5 dummies for capital city, city with population over 1 million, city with population over 250,000 to 1 million, city with population
50,000 to 250,000, and city with less than 50,000 population. Legal Organization is one of three values that reflects whether the firm is organized as a publicly listed company,
privately held limited liability company or one of the following - sole proprietorship, partnership, or another alternative form. Ownership is one of four values reflecting whether the
owner (>=50% ownership) is the state government; private domestic individuals, companies or organizations; private foreign individuals, companies or organizations; or another
ownership form. Size at birth is the log of firm size when the firm started operations. We rescale Ethnic Fractionalization, Latitude, and Settler Mortality on a four-point scale using
quantiles.
                                                                                        1                  2              3              4            5               6
                                                                                     Young            Mid-Age            Old         Young        Mid-Age            Old
   Age Groups
                                                                   All                (<5)              (5-19)         (20-39)         (<5)        (5-19)          (20-39)
   Country characteristics
   Country Dummies                                                0.023              0.012              0.023           0.029         0.018         0.021           0.021
   Legal Origin Dummies                                           0.001              0.002              0.004           0.001            0          0.001           0.001
   Ethnic Fractionalization Dummies                               0.006               0.001             0.006           0.004          0.002        0.012           0.006
   Latitude Dummies                                               0.002              0.004              0.006           0.004         -0.001        0.001           0.001
   Settler Mortality Dummies                                                                                                          0.003         0.012           0.013

   Legal Origin, Latitude, Ethnic Fractionalization
                                                                0.008               0.004            0.009          0.009
   Dummies

   Legal Origin, Latitude, Ethnic Fractionalization,
                                                                                                                                   0.004         0.014          0.013
   Settler Mortality Dummies

   Firm-level characteristics
   Age                                                          0.018
   Sector Dummies                                               0.002               0.004            0.004          0.002          0.001         0.003          0.004
   Location (City Size) Dummies                                 0.001                 0              0.001             0             0           0.002             0
   Ownership Dummies                                              0                   0              0.001             0           0.001         0.001             0
   Legal Organization Dummies                                   0.008               0.003            0.007          0.013          0.005         0.012          0.019
   Log (Size at birth)                                          0.034               0.032            0.038          0.048          0.031         0.030          0.049
   All firm-level characteristics together                     0.075                0.046            0.059          0.075          0.056         0.058          0.086
   N                                                           33982                6119             20144          5724           2234          8352           3896




                                                                                                                                                                                     48
                                             Table 7: Productivity and Lifecycle – Analysis of Variance
The table documents how firm level variables and country level variables contribute to the adjusted R-square of the following regression model when they are entered one at a time:
Labor Productivity = α + β1 Firm Characteristic Dummies + β2 Country Characteristic Dummies + e. Labor Productivity (Log) is the logarithm of the ratio of total annual sales (in
USD) to number of permanent, full time employees. The country characteristic is captured by using one of the following variables at the country level: Country dummies, Income
Dummies, Legal Origin, Ethnic Fractionalization, or Latitude. Country Income Dummies consists of 4 dummies for low income, lower middle income, upper middle income, and high
income countries. Legal Origin takes the value one if Common law, two if French civil law, and three if Socialist law countries. Ethnic Fractionalization is the probability that two
randomly selected individuals in a country are not from the same ethnic group. Latitude is the absolute value of the latitude of the country scaled between zero and one. The firm effect
is captured by one of the following variables: Sector, Location (City Size), Legal Organization, Ownership, and Size at birth. Sector is 14 industrial sector dummies. Location is 5
dummies for capital city, city with population over 1 million, city with population over 250,000 to 1 million, city with population 50,000 to 250,000, and city with less than 50,000
population. Legal Organization is one of three values that reflects whether the firm is organized as a publicly listed company, privately held limited liability company or one of the
following - sole proprietorship, partnership, or another alternative form. Ownership is one of four values reflecting whether the owner (>=50% ownership) is the state government;
private domestic individuals, companies or organizations; private foreign individuals, companies or organizations; or another ownership form. Size at birth is the firm size when the
firm started operations. We rescale Size at birth, Ethnic Fractionalization, Latitude, and Settler Mortality on a four-point scale using quantiles. We use dummy variables for all the
country and firm variables.

                                                       All             Young           Mid-Age             Old            Young           Mid-Age             Old
 Age Groups
                                                                        (<5)            (5-19)           (20-39)            (<5)            (5-19)          (20-39)
 Country characteristics
 Country Dummies                                      0.524            0.559             0.519            0.520            0.628            0.569            0.501
 Legal Origin Dummies                                 0.048            0.086             0.051            0.032            0.186            0.102            0.032
 Ethnic Fractionalization Dummies                     0.011            0.016             0.013            0.019            -0.000           0.010            0.021
 Latitude Dummies                                     0.002            0.012             0.005            0.013            0.022            0.042            0.037
 Settler Mortality Dummies                                                                                                 0.007            0.058            0.063

 Legal Origin, Latitude, Ethnic
 Fractionalization Dummies                            0.069            0.118             0.076            0.076

 Legal Origin, Latitude, Ethnic
 Fractionalization, Settler Mortality
 Dummies                                                                                                                   0.251            0.206            0.124

 Firm-level characteristics
 Age                                                  0.003
 Sector Dummies                                       0.026            0.031             0.027            0.022            0.024            0.018            0.014
 Location (City Size) Dummies                         0.009            0.036             0.009            0.008            0.106            0.043            0.009
 Ownership Dummies                                    0.004            0.006             0.003            0.004            0.013            0.002            0.004
 Legal Organization Dummies                           0.023            0.045             0.016            0.027            0.022            0.005            0.014
 Log (Size at birth)                                    0              0.003             0.000            -0.000           0.014            0.001              0
 All firm-level characteristics together              0.063            0.105             0.055            0.061            0.149            0.070            0.046
 N                                                                      4218            13848             4035              1521             5521            2657




                                                                                                                                                                                     49
                Table 8: Productivity Dispersion and Olley-Pakes Covariance Term
This table presents the standard deviation in (log) labor productivity and the covariance term from the Olley-Pakes
decomposition in each country-age bin. Labor Productivity (Log) is the logarithm of the ratio of total annual sales (in USD) to
number of permanent, full time employees. The Olley-Pakes covariance term is the difference between the weighted average of
firm-level (log) labor productivity and the un-weighted average firm-level (log) labor productivity in each country-age bin. The
weights used are the employment share of each firm in that country-age bin.


                                   STD in Labor Productivity (Log)                            Olley-Pakes Covariance
 Nation                            <5                5-19                20-39            <5           5-19         20-39
 N                                  92                 92                 89              92            92           89
 min                              0.212              0.504               0.515          -6.734        -5.509       -7.082
 Mean                             1.392              1.331               1.305          -1.464        -0.857       -0.840
 Median                           1.342              1.274               1.166          -0.858        -0.606       -0.263
 max                              3.170              3.192               3.216           1.548         1.299        1.154




                                                                                                                               50
Appendix Table A1: Summary Statistics on Firm Employment by Age

Panel A: Across income groups

                   Index of Employment = Mean employment for firms in that age bin/Mean
  Age                                     employment for firms
                                                 Lower-         Upper-
                    Full Sample     Low Inc     Middle Inc     Middle Inc    High Inc
  <5                      1            1            1              1            1
  5-19                 1.482         1.578        1.343          1.570        1.413
  20-24                1.703         1.921        1.735          1.561        1.535
  15-19                1.895         1.769        1.998          1.907        1.818
  20-24                2.330         2.427        2.061          2.508        2.474
  25-29                3.349         2.890        3.828          3.545        2.048
  30-34                3.371         3.825        3.473          3.251        1.662
  35-39                4.370         4.194        3.904          4.109        7.490
  40+                  5.304         4.823        4.676          5.502        7.684
  # of countries        118            31          38             38            11


Panel B: Across Regions

                   Index of Employment = Mean employment for firms in that age bin/Mean
  Age                                     employment for firms
                       SAR         MNA        LAC        ECA         EAP          AFR
  <5                     1           1         1           1           1           1
  5-19                1.019        1.261     1.483       1.377      1.597        1.587
  20-24               1.438        2.153     1.827       1.645      1.529        1.708
  15-19               2.010        1.418     2.140       1.736      1.733        1.859
  20-24               2.044        1.194     2.640       2.129      1.796        2.407
  25-29               1.872        1.957     2.875       4.465      2.038        3.592
  30-34               2.236        1.533     3.452       2.877      2.601        3.944
  35-39               1.851        2.232     3.386       6.094      2.873        4.979
  40+                 4.476        1.555     4.618       7.338      3.078        4.982
  # of countries         4           2         31          30         10           41




                                                                                          51
                                Appendix Table A2: (Log) Firm Size and Age
This table shows results from the following regression: Firm Size = α + β1 Age Dummies + β2 Industry Dummies + β3Country
Dummies + β4Year Dummies + e. Firm Size is the logarithm of the number of permanent, full-time employees in the firm where
permanent, full-time employees are defined as all paid employees that are contracted for a term of one or more fiscal years and/or
have a guaranteed renewal of their employment contract and that work 8 or more hours per day. Age Dummies consist of 9 age
dummies for the following age bins: <5(reference category), 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, and 40+.
Manufacturing is a dummy variable that takes the value 1 for firms in the manufacturing sector and 0 otherwise. Year Dummies
consist of dummies for each of the years in which the Enterprise Survey is conducted. All regressions are estimated by ordinary
least squares. Robust standard errors are reported in parentheses. The square brackets in col. 1 is the % change in firm size given
by 100 x (exp(b)-1) where b is the respective age coefficient.


                             (1)                 (2)                 (3)                       (4)
                          Log (Firm           Log (Firm           Log (Firm
                            Size)               Size)               Size)               Log (Firm Size)
                                                               Manufacturing          Non-Manufacturing
  5-9                      0.284***           0.242***          0.260***                  0.197***
                            (0.016)            (0.015)           (0.023)                   (0.020)
                          [32.843%]
  10-14                    0.536***           0.413***            0.429***                 0.369***
                            (0.017)            (0.017)             (0.025)                  (0.023)
                           70.916%
  15-19                    0.654***           0.527***            0.498***                 0.542***
                            (0.020)            (0.020)             (0.028)                  (0.028)
                          [94.527%]
  20-24                    0.661***           0.573***            0.603***                 0.516***
                            (0.026)            (0.026)             (0.033)                  (0.042)
                          [93.673%]
  25-29                    0.798***           0.722***            0.724***                 0.695***
                            (0.030)            (0.030)             (0.038)                  (0.051)
                         [122.109%]
  30-34                    0.947***           0.812***            0.782***                 0.898***
                            (0.037)            (0.036)             (0.045)                  (0.060)
                         [157.796%]
  35-39                    1.015***           0.913***            0.962***                 0.824***
                            (0.042)            (0.041)             (0.053)                  (0.067)
                         [175.936%]
  40+                      1.574***           1.371***            1.440***                 1.267***
                            (0.027)            (0.027)             (0.035)                  (0.047)
                         [382.591%]
  Constant                 2.566***          2.495***             2.901***                 1.814***
                            (0.011)           (0.073)              (0.157)                  (0.078)
                                              Country,
  Fixed Effects              None                                     |------- Country, Year --------|
                                           Industry, Year
  N                         44696               44676               24564                    20112
  adj. R-sq                 0.101               0.218               0.191                    0.182



                                                                                                                               52