77929


                                   Informal sector dynamics in times of
                                   fragile growth: the case of Madagascar

                                   Julia Vaillant,a Michael Grimm,b Jann Lay,c,d
                                   François Roubaud,e *

                                   September 2011


Abstract

This paper investigates the dynamics of the informal sector in Madagascar during a period of
fragile growth. Overall, the behavior of informal firms in terms of earnings, employment and
capital accumulation points to a degree of heterogeneity which goes beyond a simple dualistic
model and even a more refined model that would distinguish between an upper
entrepreneurial and a lower subsistence tier within the informal sector. However, in line with
the dualistic model, the informal sector indeed fulfils a labor absorbing function in times of
crisis. During the growth period we see capital accumulation in most of the sectors and lots of
evidence that households expand their activities. However, this happens mainly through the
creation of new firms instead of the expansion of existing ones, which is consistent with much
higher returns at very low levels of capital. More rapid expansion can be observed in sectors
that operate with lower capital intensity, which is also consistent with risk or credit constraints
as major deterrents to expansion. While there is some indication that total factor productivity
increased over time, returns to capital and labor where not higher at the end of the observation
period than at the beginning. Returns are also rather low at high levels of capital. These
findings point to a limited growth potential of the informal sector as a whole. The
heterogeneity in capital returns hints at large inefficiencies in allocating capital across
informal firms.


Keywords: Informal sector, microenterprise, firm growth, capital returns, Madagascar.

JEL codes: O12, O17, L26, D22
a
  DIAL, IRD, Université Paris-Dauphine, Paris, France
b
  International Institute of Social Studies, Erasmus University Rotterdam, The Hague, The Netherlands
c
  German Institute of Global and Area Studies (GIGA), Hamburg, Germany
d
  University of Göttingen, Germany
e
  DIAL, IRD, Hanoi, Vietnam



                                                      1
*
 Corresponding author: Julia Vaillant, DIAL, IRD, Université Paris-Dauphine, 4 rue d’Enghien, 75010
Paris, Phone : +33 1 53 24 14 76, Fax : +33 1 53 24 14 51, E-mail: vaillant@dial.prd.fr.

Acknowledgements

This research is part of a project entitled “Unlocking potential: Tackling economic,
institutional and social constraints of informal entrepreneurship in Sub-Saharan Africa�
(http://www.iss.nl/informality) funded by the Austrian, German, Norwegian, Korean
and Swiss Government through the World Bank’s Multi Donor Trust Fund Project:
“Labor Markets, Job Creation, and Economic Growth, Scaling up Research, Capacity
Building, and Action on the Ground�. The financial support is gratefully
acknowledged. The project is led by the International Institute of Social Studies of
Erasmus University Rotterdam, The Hague, The Netherlands. The other members of
the research consortium are: AFRISTAT, Bamako, Mali, DIAL-IRD, Paris, France, the
German Institute of Global and Area Studies, Hamburg, Germany and the Kiel Institute
for the World Economy, Kiel, Germany.

We thank participants at the CSAE 2011 Annual Conference in Oxford, at the
Development Economics Conference 2011 of the German Economic Association in
Berlin and at the DIAL Development Conference “Shocks in developing countries� in
Paris for useful comments and suggestions.


Disclaimer

This is work in progress. Its dissemination should encourage the exchange of ideas
about issues related to entrepreneurship and informality. The findings, interpretations
and conclusions expressed in this paper are entirely those of the authors. They do not
necessarily represent the views of the World Bank, the donors supporting the Trust
Fund or those of the institutions that are part of the research consortium.




                                                     2
1 Introduction

The informal sector is the main source of income and employment for most urban households
in the developing world. This sector was traditionally viewed as the low-productive,
subsistence sector of urban areas, bound to disappear with economic growth. However, this
process did not materialize and the sector grew in size in many of these countries. While the
dualistic model of informal and formal sectors was challenged as early as the seventies (Hart,
1973; Moser, 1978), studies of informal activities which explicitly stressed their
heterogeneous nature began to develop in the nineties (Fields, 2004; Bosch and Maloney,
2010). The informal sector comprises various forms of production and employment typically
in micro and small enterprises (MSEs). While some of these informal firms may indeed
represent a form of urban subsistence production, the informal sector is usually also the host
of a significant number of successful entrepreneurs. Indeed recent research on MSEs in
developing countries consistently finds rather high returns to capital in microenterprises
(Fajnzylber, et al., 2006; McKenzie and Woodruff, 2006; De Mel, et al., 2008; Kremer, et al.,
2008; Grimm, et al., 2011) suggesting that this sector has potential and is not just catering for
subsistence production. Heterogeneity contradicts the assertion that sustained economic
growth shrinks the informal sector until it disappears. Little is known about the way informal
firms react to (or shape) overall macroeconomic dynamics, in particular over long periods.
Yet, there is a small literature that examines the cyclical behavior of the informal sector
(Maloney, 2004; Bosch and Maloney, 2010). Because of the heterogeneity of informal firms,
there are no straightforward predictions on how size, structure and performance of the
informal sector and the firms constituting it change during long periods of economic growth
or spells of crisis. This paper is an attempt to provide some evidence on such changes for the
Malagasy economy between 1995 and 2004.

Between 1995 and 2001, Madagascar experienced an exceptional period of economic
expansion. The informal sector – defined as employment in unincorporated enterprises that
are not registered or do not keep any written accounts 1 – accounted then for 50-60% of
employment in Madagascar’s capital Antananarivo and growth appeared to be associated to a
decline in this share. In the first half of 2002, the major political crisis following the
presidential elections of December 2001 reversed this trend, as general strikes, roadblocks and
the vacancy of power caused a temporary collapse of GDP growth and a massive departure of
foreign industries in particular those located in special Export Processing Zones (EPZ). The
recent internal political and global financial crises have caused an even greater
informalization of the economy, concentrating 65% of total employment in Antananarivo in
2010.


1
    This definition follows the ILO recommendation (1993).


                                                       3
These first figures suggest that the informal sector in Madagascar – a country without
substantial unemployment insurance schemes or social safety nets – fulfills a labor absorbing
function in times of economic downturns. Yet, this is only one of the facets of the informal
sector. While the heterogeneity of informal activities is well-established in the literature, the
dynamic implications of this heterogeneity are less acknowledged and understood. Based on
the simplifying assumption of an upper (or entrepreneurial) tier and a lower (subsistence) tier
of informal activities, the literature suggest pro-cyclical behavior – that is moving with the
aggregate production - in the upper tier and counter-cyclical behavior in the lower tier (Bosch
and Maloney, 2010). Hence, both tiers are expected to differ in their response to cyclical
movements. They are typically also expected to behave differently during periods of growth.
Capital accumulation would only be expected in the upper tier, while subsistence activities
would be expected to be caught in a poverty trap and stagnate at low levels of capital.
Similarly, total factor productivity would be expected to be low and stagnant in the lower tier,
with productivity gains being limited to upper tier firms.

Based on the recent evidence mentioned above, we think that this distinction of an upper and
lower tier is too simplistic and hides the high potential of many firms at the lower end of the
capital distribution. Taking this as a background, this paper attempts to answer the following
questions: which transformations occurred in the informal sector in the decade under review?
In particular, which activities expanded and contracted in terms of employment and capital?
Do returns to capital and labor increase or decrease with economic expansion? How do all
these changes relate to the macroeconomic dynamics? To answer these questions, we rely on
a quite unique dataset of four representative and perfectly comparable cross-sectional surveys
of informal enterprises in Antananarivo in 1995, 1998, 2001 and 2004, i.e. before and after
the 2002 crisis. This dataset allows a detailed description of the characteristics of the informal
sector and its firms during a relatively long period of sustained growth, followed by a crisis
and two years of recovery.

The remainder of the paper is organized as follows. In Section 2, we briefly present the
economic and political context in Madagascar over the 1995 to 2005 period. In Section 3 we
discuss the theoretical framework and the related empirical literature. In Section 4 we
describe the data and main characteristics of the informal firms we analyze. In Section 5 we
present our results organized in five sub-sections. We start by analyzing the patterns of firm
growth and capital accumulation over time taking into account the heterogeneity of informal
firms by systematically disaggregating by sector. Then, building on the recent empirical
literature on returns to capital in small-scale activities in developing countries, we estimate
profit functions to evaluate capital and labor returns in different segments of the capital
distribution. By adding year interactions we study changes in these returns over time. In
addition, specific sector returns are estimated. A more aggregate level of analysis is adopted
in the last section that presents the results of the estimation of a sectoral profit function, using
a pseudo-panel approach. Section 6 concludes.




                                                 4
2 The economic and political context in Madagascar (1995-2005)

After a long period of economic recession which started with the country’s independence in
1960 and interrupted only by very short periods of growth, Madagascar experienced an
exceptional period of economic expansion between 1997 and 2001. Several factors, both
economic and political, drove this favorable development. Firstly, the political stability since
the election of Didier Ratsiraka in 1996 and agreements with the Bretton Woods institutions
to reduce debt created a favorable environment for investment. Secondly, the development of
EPZs attracted foreign industry, in particular textile, which stimulated exports and
employment. The rise of tourism also contributed to economic growth.

The presidential elections of December 2001 triggered a serious political crisis that lasted six
months and had catastrophic economic effects (Razafindrakoto and Roubaud, 2002). The
candidate Marc Ravalomanana challenged the first round results that he claimed were
fraudulent. He maintained to have won the elections and refused the holding of a second
round, against the incumbent president, Didier Ratsiraka. After huge demonstrations and
general strikes, the conflict intensified as roadblocks around Antananarivo were set up by
followers of Didier Ratsiraka in an attempt to paralyze the economy of the capital city. Finally,
Marc Ravalomana was proclaimed president in May, and Didier Ratsiraka left the country in
July 2002.

As shown in Figure 1, the political crisis had disastrous effects on the economy: GDP
declined by 12.7% and inflation was close to 16% in 2002 (Gubert and Robilliard, 2010).
Exports and foreign direct investments fell sharply, unemployment rose by 71% between mid-
2001 and the end of 2002 (Cling, et al., 2005). The share of employment in the informal
sector grew again, as workers were laid off from the private sector, in particular in the EPZs,
which were hit hard by the crisis. Despite the severity of the economic downturn, recovery
was quick, with GDP growth of 9.8% in 2003 and around 5% in the two following years.
However, unemployment doubled between 2001 and 2005 and in the main urban areas
increased from 4.4% to 12%. Income per capita in 2004 also remained under its pre-crisis
level (Gubert and Robilliard, 2010).

                                            Figure 1 about here

The growth process registered at the national level until 2001 is confirmed by the survey data
that will be used in this paper (see Section 4). Urban households benefited most from the
situation. In Antananarivo, the real average labor income increased by 53% between 1995 and
2001, which corresponds to a huge 8% annual growth rate, an unprecedented pace in
Madagascar’s history (Razafindrakoto and Roubaud, 2002, 2010). Consequently, the poverty
incidence decreased from 39% to 19% while income inequality was also reduced. The 2002
crisis stopped this positive trend: the unemployment rate nearly doubled along with a massive
increase in time-related underemployment 2 and child labor. Real incomes dropped by 5%.
Thereafter, despite the quick macroeconomic recovery, household living conditions stagnated:
2
 A person is in a situation of time-related underemployment if he works less than 35 hours a week and wishes to
work more.


                                                      5
in 2004, earnings were as low as in 2002, and in 2006, they were only 2% higher than during
the crisis.

3 The dynamics of the informal sector: Theoretical aspects and
  empirical evidence

Standard macro-economic labor market theories viewed the informal sector essentially as the
residual of a segmented labor market, caused by rigidities in the formal sector. The informal
sector was therefore considered to be disguised unemployment in the absence of
unemployment benefits. In contrast, others have argued that informal and formal labor
markets are in fact competitive, and workers are, at the margin, indifferent between the two
(Maloney, 2004). The sector in which they choose to work is optimal given their preferences
and the constraints they face. They may switch from one to the other to take advantage of job
opportunities. However empirical observation suggests that this is only likely to be true for a
subset of workers (Fields, 2004). In fact empirical evidence shows that the informal sector is
in many dimensions very heterogeneous, in terms of the motivation to be in this sector or of
the employed capital and labor. It includes both involuntary salaried labor, queuing for formal
jobs, and voluntary self-employment, the latter being similar in many respects to the
entrepreneurial, small firm sector in developed countries. According to these theories, the
informal sector thus comprises an upper tier, with modes of production and jobs similar to
formal firms, and a lower tier, corresponding to the residual or subsistence sector in the
dualistic view (Fields, 2004; Bosch and Maloney, 2010). However, Cunningham and Maloney
(2001) show for Mexico that only a small fraction of firms actually corresponds to the
dualistic view in which self-employment is the disadvantaged sector of a segmented market.
Instead, their findings argue in favor of strong heterogeneity among small firms, of the same
nature as in developed countries, where small firms that have reached their optimal long-run
size co-exist with profitable starting firms and start-up firms that will not survive.

Non-convex production technologies combined with imperfect credit markets have been put
forward by microeconomic theory to explain why many firms are stuck with low levels of
capital and earnings, so called ‘poverty traps’ (Banerjee and Newman, 1993; Galor and Zeira,
1993). Indivisibilities in the production technology would create entry barriers to high return
activities. Below a certain threshold of capital stock, entrepreneurs are trapped in activities
yielding very low or zero returns. If in addition capital markets do not function for them, they
are unable to invest enough to reach the necessary capital threshold and to converge to the
optimal level of capital.

However, quite in line with Cunningham and Maloney (2001), a number of recent studies
from diverse settings including Sri Lanka, Ghana, Kenya, West-Africa and Mexico have
found very high returns to capital at low levels of capital in small-scale activities, thus
contradicting the assumption of low subsistence returns at low levels of capital (McKenzie
and Woodruff, 2006; Udry and Anagol, 2006; De Mel, et al., 2008; Kremer, et al., 2008;
Fafchamps, et al., 2011; Grimm, et al., 2011). In addition, there is mixed evidence on the
existence of high entry barriers (McKenzie and Woodruff, 2006; Udry and Anagol, 2006;



                                               6
Grimm, et al., 2011).3 These empirical findings are at odds with the idea of poverty traps and
more generally with the idea that the informal sector relates to a large extent to subsistence
activities. High returns at low levels of capital rather suggest that these firms have the
potential to accumulate capital and possibly create jobs. To the best of our knowledge, capital
accumulation in informal firms has been almost completely ignored in the empirical literature.
This may be due to the – possibly wrong – assumption that these firms do not accumulate.
Another important reason is certainly the limited availability of data to assess this question.

Yet there are a number of studies, mostly rather recent, that look at the determinants of firm
growth more broadly. Based on a number of country case studies, primarily from Sub-Saharan
Africa, Mead and Liedholm (1998), for instance, find that at start-up, firms that are smaller
grow more rapidly than those that are larger. They also find that younger informal firms grow
faster than older firms. Based on their analysis the authors derive the following hypotheses
regarding the relationship between economic expansion and firm growth. In times of
macroeconomic growth, according to the authors, informal sector job creation is likely to be
channeled mainly through firm expansion, while jobs would be destroyed as workers move to
better economic opportunities, such as formal jobs. On the other hand, stagnation would lead
to downsizing and more new firm creation.4

The results by Mead and Liedholm (1998) are largely in line with those obtained by
Faijnzilber et al. (2006) for Mexican microenterprises. Faijnzilber et al. (2006) find an
insignificant relationship between time in business and firm size growth, and a significant,
negative effect of firm size on growth. This indicates that larger firms are more likely to have
reached their optimal size. They conclude that microenterprises in Mexico show a dynamic
pattern consistent with a number of standard results from the theoretical literature on firm
dynamics. 5 They further substantiate their findings by comparing Mexican and US
microenterprises. This comparison shows remarkable similarities with respect to the dynamics
of self-employment in both countries.

Firm performance and dynamics have also recently been studied by Byiers & Iacovone
(2011), using cross-country data and firm-level data collected in 2004 and 2007 in
Madagascar. They first run a cross-country regression to investigate firm growth determinants
using a pooled firm level dataset from 11 developing countries including India, Vietnam,
Mauritius and several sub-Saharan African countries. They find a positive relationship
between firm size growth and firm size. When using a panel of 135 Malagasy formal firms,
they find a negative effect of firm age on growth. In the same study they also compare labor
productivity in small informal firms with labor productivity in small formal firms. They
systematically find lower labor productivity in formal firms which might be due, according to
the authors, to the high costs of formalization. In line with what we will show below, they find
considerable heterogeneity in performance indicators of informal MSE across sectors.


3
  Though in some cases returns were rather small for female entrepreneurs. For a detailed review of the
literature, see Grimm et al. (2011).
4
  See also Nichter & Goldmark (2009) for similar statements.
5
  See, for example, the model developed by Jovanovic (1982).

                                                  7
Apart from these studies, the literature on informal sector dynamics has rather focused on the
reaction of the labor market to business cycles. The simple dualistic view implies that the
informal sector follows counter-cyclical dynamics as positive shocks to the economy should
raise formal employment and shrink the size of the informal sector. In contrast, taking the
position that informal sector employment is voluntary as workers are indifferent between
sectors at the margin, wages and employment in both sectors should move together, in a pro-
cyclical way. With an upper and a lower tier of the informal sector, we can expect pro-
cyclical movements in the upper tier and counter-cyclical in the lower tier (Bosch and
Maloney, 2010). Maloney (2004) shows that the respective sizes (in terms of employment)
and wages of the formal and informal sectors can in the short run move in opposite directions,
as in Colombia after the 1995 recession or together as in Mexico during the recovery period
of 1987-1992.

The above contributions on returns and MSE dynamics have considerably improved our
understanding of informal activities. Yet, informal firms have seldom been studied in the
context of longer cycles and sustained economic growth, especially in the African context. In
particular, our analysis will be concerned with the dynamics of capital in the informal sector.
If the informal sector primarily hosts individuals queuing for formal jobs in subsistence
activities, economic growth will be accompanied by decumulation of informal capital, as
informal activities are globally abandoned for formal jobs. This may be particularly the case if
workers with both relatively high human capital and high wealth are at the top of the “waiting
list� for formal jobs. However, indifference at the margin between informality and formality
makes such a scenario less plausible: potentially successful entrepreneurs may have little
incentive to change sectors. Rather, they may capture the business opportunities offered by
economic growth and possibly accumulate capital.

These considerations illustrate that capital accumulation patterns in the informal sector, or
more generally, firm behavior during growth (as well as crisis) eventually depends on the
nature and characteristics of informal activities. Reactions may differ across sectors, firm size
and age. This is why the empirical analysis below will first provide a detailed picture of the
characteristics of informal firms, including profits and capital stocks between 1995 and 2004.
Then we will look at labor market dynamics and cyclicality before turning to an analysis of
structural changes, growth patterns and returns to capital in the informal sector.

4 Data presentation

4.1 1-2-3 Surveys

We use data from the 1-2-3 Surveys, collected in Antananarivo between 1995 and 2004
(Rakotomanana, et al., 2003; Razafindrakoto, et al., 2009). These surveys were specifically
designed to collect information among representative samples of informal firms. Phase 1 is a
labor force survey, conducted every year since 1995, among 3,000 households. Principal and
secondary activities of every member aged 10 years and over are recorded, including the type
and status of the firm (formal/informal). This allows establishing a list of all informal firms
headed by any member of a household (whether it is the main or the secondary activity). This
serves as the sampling frame for phase 2, in which around 1,000 firms are surveyed. These
                                               8
have been randomly selected from the phase 1 listing. The stratification scheme – by
industries (nine) and type of firms (with or without wage workers) – as well as an
oversampling of the most atypical kind of firms (e.g. big manufacturing enterprises) make
sure that the full heterogeneity of the informal sector is captured. In this paper we mainly use
data from phase 2, i.e. the part that collects information on the characteristics of firms, such as
the number and characteristics of hired paid and unpaid workers, investment, expenditures for
intermediate inputs, fees and taxes, sales and profits. Finally, phase 3 is a household
expenditure survey which interviews another representative sub-sample of households drawn
from phase 1. Hence, for a sub-sample of informal firms, information from all three phases is
available. Phase 2 has been conducted every three years since the start of the 1-2-3 survey.
Hence, we can use data on a representative sample of approximately 4,000 informal firms
surveyed in 1995, 1998, 2001 or 2004.

Four key advantages of the 1-2-3 survey over other alternative datasets need to be stressed
here. First, the mixed household-enterprise survey frame is the only one to ensure the full
representativeness of the informal sector. Previous studies, especially in Africa, are typically
based on enterprise surveys (like ‘Regional Program on Enterprise Development’ surveys)
which cover only a part of all informal firms; in general the upper tier of the informal sector
in some specific industries, mainly manufacturing. Second, our four rounds of surveys are
fully comparable, as sampling methodology and questionnaire have been maintained constant
over time. To our knowledge, this is the first time ever such a series of repeated cross sections
on informal firms on a representative basis is available in Sub-Saharan Africa. Third, as
informal firms do not usually keep books (and in many cases do not even have any written
records), the survey questionnaire has been designed in order to assist the owners of the firms
in establishing, product by product, all its sales and expenses, on a flexible period of time,
adapted to each individual case. This detailed and comprehensive procedure to collect
information is the only way to get reliable data, and to avoid the usual underestimation biases
caused by more aggregate questions. The same extensive process is used to reconstitute the
capital stock evaluated at replacement costs to take into account depreciation. Given that for
each capital item, the date of purchase is collected, the full sequence of capital accumulation
can be obtained, year by year. The only drawback is that we do not have information on
capital goods that were fully depreciated or were sold in one of the previous periods. Fourth,
the survey comprises a special module designed to capture monthly seasonality. Given the
usually high intra-year fluctuations in the informal sector, this is again a significant advantage.
In sum, our dataset offers a unique opportunity to get insights on the evolution and dynamics
of the informal sector over a decade.

4.2 Main characteristics of the informal sector

Table 1 provides a summary of key characteristics of informal firms and their owners, both
overall and disaggregating the sample by quartiles of capital stock. While the average age of
the firms is around 8.7 years, the median age is 3.7 years lower, indicating that there are a lot
of very young firms in the sample, and a few (much) older ones. Half of the informal firms
are owned by women. Two thirds of informal firms are operated by the owner only.

                                       Table 1 about here
                                                9
Table 1 also shows the mean and median of performance indicators such as value added,
capital stock and capital stock at start-up. Value added is calculated as the difference between
sales and intermediary consumption. Intermediary consumption includes raw material and
inventory purchases, rent and utilities, and other expenses. Value added thus includes capital
income, all labor income and entrepreneurial profits. Capital is the total stock of capital, again
measured at the actual replacement value. To obtain real values, the Consumer Price Index
(CPI) is used as a deflator. This is of course, at least from a conceptual point of view, not very
satisfactory. Capital goods should be deflated over time using a specific price index, since the
price of equipment, real estate or vehicles is likely to evolve differently from consumer prices,
in particular from prices for basic food- and non-foodstuffs. Unfortunately, specific price
indices for physical capital goods are not available in Madagascar (as in most others
developing countries), which only provides a CPI and its components by large categories. To
test the sensitivity of our results with respect to the deflator used, we alternatively deflated
capital by the equipment, furniture and housing component of the CPI. This led to a very
similar result in terms of the development of the capital stocks over time. Hence, in what
follows we always apply the CPI.

Mean monthly value added is 491,000 Fmg (US$49) while median value added is only
193,000 Fmg (US$20), indicating substantial heterogeneity across firms.6 This is even more
visible in the value of the capital stock, whose distribution is strongly skewed to the right. The
mean is over 2,600,000 Fmg (US$260) but the median is only 341,000 Fmg (US$34). There
are a few firms operating with a very high capital stock, while the large majority owns very
little or no capital at all. In fact, 10% of the firms operate with no physical capital whatsoever.

The most striking result is that apart from education and direct performance or size indicators
such as employment and value added, there is not much variation in firm characteristics
across quartiles. The age of the owner is slightly increasing, but the difference is relatively
small with 2.7 years separating the average firm in the first quartile from the average firm in
the fourth quartile. Even the gender of the owner does not vary much, only the top quartile has
significantly more male owners than the other three quartiles. Average experience is slightly
decreasing with capital, maybe surprisingly. However, education of the owner increases
strongly with capital.

High capital firms are larger in terms of staff. In fact the large majority of firms operating
with very little or no capital just employ the owner. While value added increases in a broadly
linear manner across capital quartiles, the value of the capital stock increases ten-fold between
the first and the second, and between the third and the fourth quartiles respectively. This
suggests, at this stage, decreasing returns to capital.

Disaggregating the sample by sector of activity can provide additional insights on the
different types of firms that operate in the informal sector. Eight sectors are defined: food
processing, clothing and textile manufacturing, other industries, construction, trade, services
to households and enterprises, catering and transport (Table 2). We also included in our table

6
    The amounts in US Dollars are calculated using the current US$-Malagasy Franc exchange rate.


                                                       10
a variable measuring investment at start-up in order to assess the existence and extent of entry
barriers in the informal sector. It is defined as the amount of physical capital purchased during
the first year of activity, and is shown for the sub-sample of firms that are four years old or
younger; the latter to minimize the recall bias with respect to the investments undertaken
when the business started. This is of course only an imperfect approach to the problem of
entry barriers, but we think that this can at least give a rough idea with respect to the
minimum of capital needed to start in a specific sector.

                                       Table 2 about here

There are some striking differences across sectors. Firstly, some activities are clearly gender-
specific: textile and catering are feminine activities while construction and transport are
almost entirely run by men. Older firms are concentrated in textile and other industries,
construction and service activities. These sectors exhibit an average time in business of over
nine years. On the other hand, catering, food processing and trade firms have been running for
less than seven years on average. Interestingly, the size of the firm in terms of employment is
not associated with its performance in terms of profit. Construction and transport have both
high rates of pure self-employment and the highest mean value added of the sample. These
two sectors stand out as extremes in terms of their capital stock. Capital profitability is very
high in construction firms as the mean capital stock is the lowest of all sectors. At the other
end of the distribution, 80% of transport firms are situated in the highest quartile of capital
stock. This is due to the fact that vehicles represent the main type of investment in that sector.
The two other sectors, in which firms are largely owner-operated, are services and textile
manufacturing and they have the lowest average value added. Half of all service firms have a
capital stock below the 25th percentile of the distribution. The catering sector appears rather
well-performing from these descriptive statistics, with a larger size, high value added and
capital stock.

5 Results

5.1 Informal and formal labor market dynamics

As mentioned in Section 3, the dynamics of the informal sector have mainly been studied
from a labor market perspective. We first adopt this perspective too and then move on to
issues of structural changes, growth patterns and returns to capital of firms.

As in other cities in SSA, the informal sector represents the largest share of employment in
Antananarivo. Even if its share is 10 to 15 percentage points lower than in other West African
urban centers, it remains the first job provider, totalizing more than one out of two jobs (Table
3). Taken as a whole and as can be seen in Table 3, the informal sector in Antananarivo seems
to follow anti-cyclical dynamics. Between 1997 and 2001, its share steadily decreased, from
60% to 53% of total employment, reversing the previous process of informalization of the
economy. This drop occurred in a context of public administration and state enterprise
downsizing, as part of the structural adjustment program. In terms of employment, this
process mainly benefited the private formal sector. This structural change was mainly due to
the rapid development of EPZ and at least to some extent to the expansion of formal domestic

                                               11
enterprises. The average annual growth rate of employment over the period was 27% in EPZ
but only 3% in the informal sector. As shown by Table 3, in Antananarivo this led to a tripling
of the share of EPZs in total employment between 1995 and 2001, from 3% to more than 10%,
while the share of private formal sector jobs remained stagnant at 25% (Cling, et al., 2005).

                                       Table 3 about here

The 2002 crisis totally reversed this trend. In less than one year, the informal sector ‘re-
colonized’ the labor market, absorbing the laid-off workers from closing formal enterprises
and the new entrants, deprived from any alternative source of jobs. All the progress made
during the five previous years was erased, and the informal sector grew up to its highest level
since 1995 (59.5 % of total employment). While both dependent and independent informal
employment increased, the growth in the number of informal entrepreneurs was much faster
than the overall increase in the number of workers. This is a sign that informal sector
employment growth is extensive rather than intensive, as it happens mainly through the
creation of new firms rather than the expansion of employment in existing firms. Interestingly,
in the period of growth (1998-2001), although dependent informal labor was absorbed in
formal enterprises, the absolute number of firms continued to increase, even faster than the
overall growth of the employed labor force. This suggests that the informal sector consists of
both workers queuing for a formal job and voluntary entrepreneurs. Conversely, in the period
of crisis and the following recovery, the decrease in formal employment seems to have been
mainly compensated by an increase in informal independent labor (the share in total
employment increases from 35 percent to 38.6%), rather than informal hired or family labor,
suggesting that existing firms were not able to absorb the surplus labor released by the formal
sector, and most of these workers started an informal activity. Additionally, an important
fraction of the fast growth in the number of informal firms is explained by new entries on the
labor market.

The EPZ paid the highest tribute to the crisis, employment being divided by nearly three.
From 2002 onwards, the EPZ recovers its pre-crisis number of jobs. Yet, recovery of domestic
formal enterprises seemed to be limited (Cling, et al., 2009). Finally, as a second political
turmoil occurred in 2009 combined with the international financial crisis, which resulted in a
new drastic shock, the informal sector employment absorbed nearly two thirds of the labor
force in 2010 (Rakotomanana, et al., 2010).

In terms of labor income, the informal sector is, as expected, the lowest paying segment of the
urban labor market, with jobs in the public sector at the top of the earnings ladder (first row of
Table 4). Interestingly, although it is significant, the earning gap with EPZs jobs is quite low,
stressing the potential trade-offs in choosing one sector or the other for low skill workers,
especially women (Glick and Roubaud, 2006). The decline in informal sector employment in
the second half of the nineties was accompanied by large income gains from informal
activities. Between 1995 and 2001, real average informal earnings increased by 66%, this is
more than the 53% registered over all sectors together. Given that the informal sector is less
exposed to international competition than the formal tradable sector, informal firms have been
able to benefit from the increase in domestic demand. In spite of the lower income elasticity
of their products and of a decreasing market share for consumption goods (-6 percentage
                                               12
points), informal goods still satisfied nearly three quarters of household consumption in 2001.
If only food is considered the share catered by the informal sector was even 95%
(Razafindrakoto and Roubaud, 2010).

                                      Table 4 about here

Conversely, in 2002, the average income in the informal sector was reduced by 11%, while
the decline for the whole labor market was ‘only’ 5%. Shrinking aggregate demand combined
with the absorption of labor quitting the formal sector are likely to be the main drivers of this
sharp contraction. The shift from formal to informal consumption goods following the
impoverishment of the population was not sufficient to counterbalance the two former effects
(Razafindrakoto and Roubaud, 2010). On the contrary, the formal sector was able to maintain
real wages, but at the expense of a massive reduction in jobs. These figures are consistent
with the common belief that the formal sector would adjust during downturns through
quantity, while price adjustment would be the main mechanism at work in the informal sector.
Subsequently, informal sector incomes progressively recover part of their purchasing power,
at least up to 2009, before a new drastic drop occurred.

Up to now, we analyzed informal sector dynamics through labor force survey data, which
does not allow focusing on the productive aspects of employment. We also considered each
sector as an aggregate without taking into account its intrinsic heterogeneity. Taking
advantage of the survey data presented in Section 4 will now allow us to overcome these
limitations.

5.2 Main structural changes

In this section, we hence describe the main changes in the structure and characteristics of
informal firms over time, using the data from Phase 2 of the 1-2-3 survey collected in 1995,
1998, 2001 and 2004. The first remarkable result is, as can be seen in Table 5, that there is no
substantial change in the distribution of the main socio-demographic characteristics of firm
owners (e.g. in terms of gender or education) over time.

                                      Table 5 about here

Value added of informal firms increased between 1995 and 2001, i.e. during the period of
economic growth and fell afterwards as a result of the sharp economic contraction starting in
2001. This is in line with the observation of increasing informal incomes highlighted in the
previous section. The average time in business of a firm increased steadily, starting with 8
years in 1995 and reaching an average of 9.5 years in the most recent survey. In fact, in 1995
over a fifth of all firms had started their activity a year before the survey. This proportion
decreased to 16% in 2004, while the share of older firms (4 years or more in business)
increased by 10 percentage points. If we disaggregate this further by the size of the capital
stock, we can see that this effect is particularly and surprisingly strong among low capital
firms (below the median of the capital stock distribution). In the two lowest quartiles of




                                               13
capital stock the average age of firms has in fact increased by 2.5 years.7 This development is
particularly pronounced between 2001 and 2004. Obviously it is impossible to explain these
shifts in detail, given that we do not have information on exits and entries of firms,8 but, these
findings suggest that new firms operating with very little capital did probably not survive the
crisis. Put differently, a young age combined with low capital is likely to make firms very
vulnerable.

Table 5 shows a notable change in the firm size distribution. It confirms that growth in the
informal sector was mainly extensive. There is much more self-employment in 2004 than in
1995, and there are less firms in the range of two to four workers. Marginally, existing firms,
successful enough to be relatively large, could have absorbed some of the excess labor laid off
from formal firms. It is however mainly through the start-up of new activities that
unemployed workers responded or had to respond to the crisis.

The increase in self-employment can also be related to changes in the sectoral distribution. As
shown in Table 5, the biggest change occurred in fact at the sector level. First of all, while
trade activities represented 40% of all informal firms in Antananarivo in 1995, this share was
only 31% nine years later. Overall, the share of industrial activities, which comprise food,
textile and other manufacturing, also decreased over time. While this can partly be explained
by a reallocation of activities within the informal sector, formalization combined with the
general decline in employment in industrial activities between 2001 and 2004 certainly also
contributed to this reduction (see Table 6). For textile producing firms in particular, there was
a clear shift of labor from informal to formal firms before 2001. This shift was partly reversed
after 2001. This was due to competition in the labor market rather than the product market.
Indeed, textile goods produced in EPZs are exported while those produced by informal firms
are for the domestic market and workers seem to prefer employment in EPZs over informal
employment. That trend had been interrupted between 2001 and 2004, as the crisis of 2002
resulted in the closing of many clothing factories in the EPZs. Surplus labor reallocated to the
informal sector, in fact not only in textile activities. The share of transport and catering
activities was stable. Services have gained the most over the decade, from 19.6% of all firms
to 30%. This trend does not seem to have been affected by the crisis, as it increased by 2.5 to
5 percentage points over each three-year period. The increase in the share of the construction
sector is also quite important, starting from less than 4% in 1995 and representing 8% of
informal firms in 2004. During the growth episodes the demand for housing might have
increased and triggered an increase of supply in response.

                                             Table 6 about here

In order to better understand the underlying drivers of these sectoral shifts, we relate these
changes to the amount of capital necessary to start an activity in these various sectors. We
find that sectors which gained the most — services and construction — show rather low start-
up investments, they usually operate on little physical capital and recover their initial
7
 Results are not shown but available on request.
8
 More precisely, the mean firm size over time is driven by three forces: changes in size of surviving firms, exit
of firms of a given size and entry of firms of a given size.


                                                       14
investment very fast. This is documented in Table 2. This suggests that indeed low entry
barriers encourage the start-up of less capital intensive activities such as petty reparations, car
mechanics, laundering or hairdressing. In addition, industrial activities and transport exhibit
high entry barriers and shrank or did at least not expand much over the observation period.
Trade did also shrink despite low entry barriers, but this is probably due to entry costs in
terms of required inventories not necessarily start-up investments. Surprisingly there also
seems to be no straightforward link between the profitability of an activity (in terms of value
added per capital stock) and its development over time, except for construction.

5.3 Growth patterns

In this section we examine the patterns of firm growth, in terms of both labor and capital. We
start by looking at the development of the distribution of the average capital stock per firm by
sector and year (Figure 2) and the total sectoral capital stock by year (Figure 3). Since we use
sampling weights to compute the total capital stock, Figure 3 provides an estimation of the
total capital stock of the informal sector by sector and year in Antananarivo. Overall, Figure 2
suggests a decumulation or stagnation of capital over time, except in the 1995 to 1998 period
during which we observe an increase at least in the median capital stock. Overall, the total
capital stock in the informal sector strongly increased over the first three years of the period
considered, after which it seems to have rather stagnated. However, as mentioned above, the
total number of informal firms increased over the entire decade. Again, extensive growth
seems to be one driver of this development. Around 30% of all households own more than
one informal firm. The survey includes a question on the projected use of a loan if the
entrepreneur could obtain one for his or her activity. While a little less than half declared that
they would invest it in the existing firm, about 40% declared that they would rather start
another firm. Investing in a new firm rather than expanding an existing one is hence a
common strategy. This is particularly true in the service sector where 50% of firm owners
would opt for extensive rather than intensive growth.

                                       Figure 2 about here

Textile manufacturing firms show a significant decline in both the mean capital stock per firm
and the total capital stock in that sector. As the share of this activity decreased, this indicates
that, in addition to textile activities being less frequent, the remaining ones also have a lower
capital stock on which they operate. Firms in the construction sector increased in number, but
they operate on less and less capital as well. On the other hand, food processing businesses
saw their median capital stock rise, while the total capital stock clearly decreased. As there are
fewer firms in this sector in 2004, this suggests that only the food processing businesses that
had a rather high capital stock are still active in 2004, unlike the pattern observed for textile
industries. Other sectors do not exhibit such clear patterns of capital accumulation or
decumulation.

We now turn to the development of firm size in terms of employed staff. Table 8 shows the
proportion of firms which expanded, contracted or stagnated in terms of their staff. The



                                                15
survey asked respondents about the number of workers when the firm was set up9 and the
number of workers at the time of the interview. This allows us to calculate employment
growth rates at the firm level. However, to reduce recall bias problems we restrict our sample
to firms that are ten years old or younger.

                                               Table 8 about here

First we can note that the large majority, that is 83% of firms neither expanded nor contracted
since their set-up. This is again consistent with our hypothesis on extensive firm growth.
Second, firms surveyed in 2001 had expanded but also contracted more than firms surveyed
before or after that date. This may be due to the overall dynamism of the economy, of which
firms surveyed in 2001 benefited for the longest period of time. This would have generated
both more employment growth in surviving firms and more exits of very successful firms that
have joined the formal sector. This is consistent with the stylized facts with respect to the
relationship between firm growth and macroeconomic expansion which we discussed in
Section 3.

Employment dynamics by sector yield further interesting insights. Textile manufacturing,
transport and service firms in most cases stagnated, and show little expansion or contraction.
On the other hand, some of the other sectors are very dynamic. For instance over 30% of the
firms in the food processing and other industries, catering and construction businesses grew in
size since their start-up. Food processing in particular exhibits close to no reduction in size at
all. This is the only dynamic sector both in terms of employment growth and capital
accumulation, while clothing and apparel businesses stagnated and decumulated capital. The
food processing sector’s decline in employment seems to be associated with a concentration
process that resulted in fewer firms with a higher average size.

To further analyze the determinants of firm growth, we now run regressions of firm
employment growth on a set of explanatory variables:

         Growthi  �0  X 'i �1  F 'i �2  S 'i �3  T 'i �4  �i .                          (1)

The vector X i is a set of owner characteristics (age when started firm, female, marital status,
education, number of firms owned by the household), Fi is a vector of firm characteristics
(initial size, time in business), Si is a vector of sector and Ti is a vector of year dummies
indicating the start-up year of the business. Firm employment growth is measured in three
different ways. In the first model growth is the absolute difference between the actual and the
initial labor force size, divided by the number of years in activity. In the second model, the
dependent variable is the compound annual growth rate of employment. Finally an ordered
probit is estimated, where the dependent variable takes three values, indicating contraction,
stagnation and expansion. Results are shown in Table 9.



9
    Except in the 1995 survey in which this question was not asked.


                                                        16
First we see that at least the models in columns (1) and (2) suggest that households owning a
second firm have lower employment growth (per firm). This confirms again that many
households engage in extensive growth rather than intensive growth. We do not find an
equally negative effect for households with more than two firms, but this may have various
reasons: there are only few households with three and more firms, which makes it difficult to
get a reliable estimate. In addition, households with three and more firms may again have
different motives and characteristics and thus engage in both expansion of existing firms and
the set up of new firms.

Second, we note a convergence effect. Firms that started with two, three, four and more
workers grow less than pure self-employment firms. This is in line with other findings on
microenterprise growth in developing countries (Mead and Liedholm, 1998). However, in our
case the same disclaimer as above applies. We lose of course firms that formalized. Hence, if
these firms showed higher employment growth than firms that remained informal, our model
would underestimate the extent of firm growth and the convergence effect would be biased.

Third, the regressions seem to suggest that the start year is rather irrelevant for firm growth.
Fourth and finally, the sectoral dummies confirm what we saw already in our descriptive
statistics. Catering and industrial activities experienced employment growth while service and
transport activities contracted in terms of their staff. All these results are also confirmed by
the ordered probit model.

In sum, we must conclude from this exercise that the informal sector as a whole showed only
little dynamics, both in terms of capital accumulation and employment growth. However, the
heterogeneity across sectors is substantial. There are more small firms at the end of the period
than in 1995. In addition, employment growth regressions show a convergence effect, which
seems to indicate decreasing returns to factors, contrary to what is predicted by poverty trap
models. Hence, after having analyzed the changes in the structure of the informal sector over
time and the evolution of capital and labor, we now analyze how the returns to these factors,
in particular capital, evolved over time. To this end, we estimate production functions by
sector and different levels of capital stock.

5.4 Returns to capital and labor

We start by estimating marginal returns to capital on the pooled sample of firms. We estimate
alternatively a logarithmic and a polynomial specification. The logarithmic specification reads

         lnVAi  �0  �1 ln Ki  �2 ln Li  Xi�3  S 'i �4  T 'i �5  �i ,                (2)

where VAi is the value added of firm i, Ki is the stock of capital of the firm, Li the total
number of hours worked in the firm both by paid and unpaid workers (such as family
members). X i is a vector of owner characteristics such as gender, education and experience.
The regression also includes year Ti and sector Si indicators. The polynomial specification
reads:


                                                        17
         VAi  �0  �1Ki  �2 Ki 2  �3Li  �4 Li 2  Xi�5  S 'i �6  T 'i �7  �i .          (3)

The polynomial specification has the advantage of explicitly allowing for non-linearities in
the marginal returns to capital and labor, while the logarithmic specification is less influenced
by outliers. However, both specifications are of course potentially affected by an unobserved
heterogeneity bias, which we can with the data at hand not appropriately tackle. For instance,
in a context of imperfect capital markets, firms with a higher performance find it less difficult
to invest and should end up with a higher capital stock. Another source of bias is of course
measurement error.

We run these regressions on the entire sample as well as on three sub-samples defined by the
level of capital of the firm. We define the median and the 80th percentile of the capital
distribution as thresholds for low, medium and high levels of capital. 10 These thresholds are
calculated separately for each year. Firms operating with no capital are excluded from the
regression as well as firms with zero profit. In addition, we drop influential outliers from the
regression, identified by the DFITS-statistic (Belsley, et al., 2004).11

Results are shown in Table 10. We also report the implied marginal returns to capital (MRK)
calculated at the mean levels of value added and capital. More precisely, to get the MRK
based on the logarithmic specification, we calculate:

      ˆ VA ,
MRK  �                                                                                         (4)
       1
           K

where VA and K are the means of value added and capital respectively. The MRK based on
the polynomial specification is simply given by:

           ˆK
      ˆ  2�
MRK  �1    2 .                                                                                 (5)

Before we discuss the coefficients of main interest, we briefly discuss some of the effects
associated with the control variables. We notice a strong negative gender effect. On average,
value added is significantly lower for women, even after controlling for sector and the level of
capital stock. The time dummies indicate that the increase in value added over the entire
period is significant, however value added of high-capital firms is lower in 2004 than in 2001.
This suggests that capital-rich firms benefit more in times of growth but also suffer more in
times of crisis. Such a result is consistent with Gubert & Roubaud (2011), who assessed the
impact of microfinance loans on informal firms in Madagascar between 2001 and 2004, based
on a quasi-experimental approach and panel data analysis. As suggested by the descriptive
analysis the textile manufacturing sector appears to be the least profitable sector while
transport and construction are the most profitable sectors.

                                               Table 10 about here



10
     These thresholds were chosen rather arbitrarily but results are robust to minor changes.
11
     Applying this procedure, we lose about 5% of the observations.

                                                          18
As shown in the first two columns of Table 10, the implied marginal returns to capital are
2.2%. However, at low levels of capital, returns are much higher. An entrepreneur operating
at the average level of capital of 194,000 Fmg and realizing the average monthly value added
of 354,000 Fmg would increase his or her monthly value added by 18% if the equivalent of
10,000 Fmg (approximately $1) is added to the capital stock. Although the estimated
production elasticities are higher in the medium and high range of capital, implied marginal
returns calculated at average value added and capital are with 0.5% to about 5% substantially
lower than in the lower segment of the capital distribution. Decreasing returns to capital
contradict the standard theory of poverty traps, but are in line with findings from other studies
on Africa, Latin America and Asia (see Section 3). This finding may also explain why we
observe such little growth in firm size and capital stock. Lower returns at high levels of
capital suggest that firms reach their optimal size at a relatively low level of capital and labor.
Yet, it is of course important to note that looking just at informal firms means that we ignore
all those firms that graduated from the informal to the formal sector. It is of course possible
that those enterprises that grow in terms of capital and staff and that manage to realize
increasing returns become formal in large numbers. However, the employment distribution
over time in Table 3 suggests that the number of formal firms did not evolve much and rather
decreased than increased. Moreover, there is not much evidence that formalization would
boost productivity of rather small firms. Byiers & Iacovone (2011), for instance, find a lower
productivity in formal micro firms compared to informal micro firms. They speculate that this
may at least partly be linked to the relatively high costs of formalization and the rather low
quality of public services that entrepreneurs get in return.

Next, we analyze the change in marginal returns over time by adding year interaction terms to
the list of regressors. Results are shown in Table 11. We note that the MRKs are remarkably
stable over time (around 2%). However, looking again at the MRKs estimated for the
different segments of the capital distribution we get a more contrasting story. At low levels of
capital, the increase in the MRK is quite important, leaping from 8% to 25%. The negative
coefficients of the year indicators show that this segment did not globally benefit from
economic growth during the studied decade. In contrast, at medium and high levels of capital,
marginal returns to capital were globally stable or decreasing.

                                                 Table 11 about here

Given the sectoral heterogeneity discussed earlier in this paper, we now run the same
regressions on sub-samples by sector. Table 12 shows again the implied MRKs. 12 Food
processing and other industries are aggregated in a single sector to ensure that the sample size
is large enough to get a reliable estimate. Yet, we refrain from estimating MRKs for the
transport and construction sectors given the small sample sizes. We find for all sectors the
same pattern: decreasing returns to capital. Catering exhibits the highest returns among the
five sectors we consider overall and for low and medium levels of capital. Trade exhibits very
rapidly decreasing returns. Returns are lowest in the textile sector and weakly decreasing but
only at very low levels of capital. Marginal returns are in fact nil in capital-intensive trade
firms. Hence, what we concluded above for the informal sector as a whole is also confirmed

12
     The full set of regressions is available on request.

                                                            19
at the sectoral level: there are only very modest growth prospects, which again may explain
extensive growth rather than intensive growth.

5.5 Returns to capital and labor at the sector level

In this final section, we again estimate returns to capital and labor. However, now we estimate
these production functions at the sector level and not at the level of individual firms. This
analysis should provide us with an estimate of the ‘macroeconomic’ or ‘social’ marginal
returns to capital. More precisely, we construct a pseudo-panel of cohorts based on the age of
the firm and the sector of activity.13 Pseudo-panels rely on the assumption that in each cross-
section, the firms surveyed are a random, representative sample of the cohort to which they
belong. This assumption is fulfilled in the case of the Malagasy 1-2-3 survey. So, we take the
mean value added of each cohort in each period and regress it on the mean of the various
covariates, in particular physical capital, and control for time-invariant cohort fixed effects
(Deaton, 1985). Cohorts are defined by the year of birth of the firm. To ensure that each
cohort is large enough, we group firms in three-year periods (1987-1989; 1990-1992; 1993-
1995; 1996-1998; 1999-2001; 2002-2004). Given the low number of very old firms, we
aggregate all firms born before 1987 in a single cohort. We also have to aggregate again some
of the sectors. Hence, we consider four aggregated sectors: food and other manufacturing
industries, textile and clothing, trade of primary and transformed goods and services and
catering. Again, the transport and construction sectors are dropped because the number of
observations is too small and they do not fit very well in any of the other categories. The
analysis is therefore carried out using a sample of 28 cohort-sector cells, each observed in the
four rounds of the survey.14

We estimate two models, without and with cohort-sector fixed effects. The first model is:

ln Yijt  �  �1 ln Kijt  �2 ln Lijt  �3 X ijt  Tt ' �4  S j ' �5  (Tt  S j )' �6  � ijt ,          (6)

where subscript i refers to the cohort, j to the sector, and t to the survey year. Capital Kijt ,
labor Lijt and owner characteristics Xijt are cohort means in each year. Tt are year indicators,
and S j are sector dummies. Ignoring fixed effects may have an advantage in cases where
there is only a very modest variation in capital within cohorts. The model with cohort-sector
fixed effects reads:

ln Yijt  �ij  �1 ln Kijt  �2 ln Lijt  �3 X ijt  Tt ' �4  (Tt  S j )' �5  �ijt ,                    (7)

where ï?¬ij is a cohort and sector fixed effect. Results of the pseudo-panel estimation are shown
in Table 13.


13
   A difference with a standard cross-sector panel is that here we stick to cohorts and thus stick to firms that
survive over time. Firms that drop out are likely to be different from the survivors; hence there are potential
selection effects involved.
14
   Cohorts of firms born after 1995 obviously have fewer observations than those born before.

                                                              20
The model without cohort fixed effects uses the variation between and within sector cohorts
while controlling for time-varying sector effects. The results of this estimate are surprisingly
similar to the above estimates that use the disaggregated information at the firm level. Both
the capital and labor coefficients are significant and with about 0.2 for capital and 0.5 for
labor of a very similar magnitude to the previous estimates. The estimated production
elasticity of capital is also very much in line with macro-level estimates in low-income
countries.15 At first sight, the estimated production elasticity of capital suggests that the sector
aggregates of informal firms behave according to standard economic theory, i.e. according to
standard neo-classical production technologies. With its low levels of capital stock –
compared to larger firms – we should see capital accumulation in the informal sector. Why
this does not happen in the aggregate becomes apparent when we consider capital productivity
and the implied MRK reported in Table 13. Average capital productivity is so low that
monthly marginal returns are only about 3 %, which explains why so little capital flows into
these activities – in particular if investors are not able to distinguish high and low return firms
in the informal sector. This result is of course due to the heterogeneity of returns in informal
firms that we have shown just above and driven by the low capital productivity in relatively
large informal firms in terms of capital stock. The heterogeneity demonstrated by the
disaggregated return estimates in the previous sub-section in conjunction with the finding of
low aggregate capital returns can be taken as a sign that the informal sector is extremely
inefficient in allocating capital across firms.

The second set of results using a cohort fixed effect does not lead to a significant coefficient
for capital, but the labor-output elasticity is again in the same range as the micro-level
estimates. The insignificant production elasticity of capital indicates that there is only a very
limited variation in capital within single cohorts and that the between-cohort within-sector
variation is likely to be the main driver of the first set of results – yet taking into account that
we control for sector, year, and sector-year effects. The finding that the cohort fixed-effect
estimate provides a labor output elasticity of very similar magnitude suggests that informal
firms – as expected – adjust to changing macroeconomic conditions mainly by adjusting labor
inputs.

6 Conclusions

This paper is a first attempt to analyze in detail structural change in the informal sector in a
poor sub-Saharan African country over a long period. The context we have chosen is
particularly interesting because it refers to a period of fragile growth, characterizing many
African countries today. We started by relating overall dynamics of this sector to the
macroeconomic changes that have occurred during the observation period. We then compared
the informal sector in terms of aggregated employment dynamics over time to those in other
sectors. We then looked at sectoral changes within the informal sector over time and analyzed
the dynamics of the two most important production factors: capital and labor. We completed
this analysis by estimating the returns to these factors and their changes over time. Overall,
the observed behavior of informal firms in terms of earnings, employment and capital


15
     See e.g. Caselli (2005).

                                                21
accumulation supports a theoretical framework that goes beyond the simple dualistic model
and even beyond a more refined version of it that would distinguish between an upper
entrepreneurial and a lower subsistence tier.

However, as usually assumed by these simple dualistic models, the informal sector indeed
fulfills a labor absorbing function in times of crisis. In the Malagasy case, this seems to
happen through creating new informal firms rather than absorbing labor into existing ones.
The fact that labor formerly employed in informal firms is being absorbed by formal activities
in the Export Processing Zones is also in line with the dualistic view. Employment in informal
clothing and apparel firms decreases considerably in the period under consideration. This is
due to competition on the labor rather than the product market, as workers choose to work in
EPZs rather than in informal firms. In contrast, a number of other findings point to parts of the
informal sector and its firms as a very dynamic segment of the Malagasy economy. During
the growth period we see capital accumulation in most of the sectors and lots of evidence that
households expand their activities. However, in contrast to labor absorption into informal
firms, this happens mainly in the form of extensive informal sector growth, i.e. through the
creation of new firms instead of the expansion of existing ones. This behavior is consistent
with much higher returns at very low levels of capital, as accumulation at the extensive
margin may then imply that the household is able to reap those marginal returns. In addition,
risk diversification strategies may at least partly explain this observation. More rapid
expansion can be observed in sectors that operate with lower capital intensity, which is also
consistent with risk or credit constraints as major deterrents to expansion.

While there is some indication that total factor productivity increased over time, returns to
capital and labor where not higher at the end of the observation period than at the beginning.
Returns are also rather low at high levels of capital. These findings point to a limited growth
potential of the informal sector as a whole. Yet, there seems to be a sub-segment of informal
firms that is able to earn high marginal returns albeit at very low levels of capital. The
heterogeneity in capital returns hints at large inefficiencies in allocating capital across
informal firms. Although this paper does not explicitly address the causes of the observed
pattern, we hypothesize that these inefficiencies could be linked to the macroeconomic and
political context which seems to be not very conducive for entrepreneurial activities.
Furthermore, constraints that arise directly at the level of the firm, such as capital market
imperfections and high risks in conjunction with risk aversion are likely to be important
causes of this inefficient capital allocation. Finally, an important drawback of our study is that
we cannot trace successful informal enterprises that eventually become formal. Although we
do not expect this to happen for a large number of firms, including them would probably
allow drawing a slightly more optimistic picture.

7 References

Banerjee, A, and A Newman. (1993). Occupational choice and the process of development.
       Journal of Political Economy, 101(2), 274-298.
Belsley, D.A., E. Kuh, and R.E. Welsch. (2004). Regression diagnostics: Identifying
       influential data and sources of collinearity (Vol. 546): Wiley-Interscience.


                                               22
Bosch, M., and W.F. Maloney. (2010). Comparative analysis of labor market dynamics using
        markov processes: An application to informality. Labour Economics, 17(4), 621-631.
Byiers, B, and L Iacovone. (2011). An Analysis of Pre-Crisis Madagascar Firm Performance:
        Firm Growth and Productivity. Paper presented at the CSAE Annual Conference 2011,
        Oxford, United Kingdom.
Caselli, F. (2005). Accounting for cross-country income differences. In Aghion & Durlauf
        (Eds.), Handbook of economic growth (Vol. 1, pp. 679-741): Elsevier.
Cling, J.P., M. Razafindrakoto, and F. Roubaud. (2005). Export Processing Zones in
        Madagascar: a Success Story under Threat? World Development, 33(5), 785-803.
Cling, J.P., M. Razafindrakoto, and F. Roubaud. (2009). Export Processing Zones in
        Madagascar: The Impact of the Dismantling of Clothing Quotas on Employment and
        Labor Standards. In Robertson, Brown, Pierre & Sanchez-Puerta (Eds.), Globalization,
        Wages, and the Quality of Jobs (pp. 237-264). Washington D.C.: The World Bank.
Cunningham, W.V., and W.F. Maloney. (2001). Heterogeneity among Mexico's
        microenterprises: an application of factor and cluster analysis. Economic Development
        and Cultural Change, 50(1), 131-156.
De Mel, S, D McKenzie, and C Woodruff. (2008). Returns to Capital in Microenterprises:
        Evidence from a Field Experiment. Quarterly Journal of Economics, 123(4), 1329-
        1372.
Deaton, A. (1985). Panel data from time series of cross-sections. Journal of Econometrics,
        30(1-2), 109-126.
Fafchamps, M, D McKenzie, S Quinn, and C Woodruff. (2011). When is capital enough to
        get female microenterprises growing? Evidence from a randomized experiment in
        Ghana.Unpublished manuscript.
Fajnzylber, P., W. Maloney, and G.M. Rojas. (2006). Microenterprise dynamics in developing
        countries: How similar are they to those in the industrialized world? Evidence from
        Mexico. The World Bank Economic Review, 20(3), 389.
Fields, G.S. (2004). A Guide to Multisector Labor Market Models. Paper prepared for the
        World Bank Labor Market Conference, 2005.
Galor, O, and J Zeira. (1993). Income distribution and macroeconomics. The Review of
        Economic Studies, 60(1), 35-52.
Glick, P., and F. Roubaud. (2006). Export Processing Zone Expansion in Madagascar: What
        are the Labour Market and Gender Impacts? Journal of African Economies, 15(4), 722.
Grimm, M, J Lay, and Jens Krüger. (2011). Barriers of entry and capital returns in informal
        activities: Evidence from Sub-Saharan Africa.Unpublished manuscript.
Gubert, F., and Anne-Sophie Robilliard. (2010). Croissance et pauvreté à Madagascar: un
        aperçu de la dernière décennie (1997-2007). In Gastineau, Gubert, Robilliard &
        Roubaud (Eds.), Madagascar face au défi des Objectifs du millénaire pour le
        développement (pp. 25-52). Marseille: IRD.
Gubert, F., and F. Roubaud. (2011). The Impact of Microfinance Loans on Small Informal
        Enterprises in Madagascar. A Panel Data Analysis.Unpublished manuscript.
Jovanovic, B. (1982). Selection and the Evolution of Industry. Econometrica, 50(3), 649-670.
Kremer, M, JN Lee, and JM Robinson. (2008). The Return to Capital for Small Retailers in
        Kenya: Evidence from Inventories. Manuscript, Harvard University.
Maloney, WF. (2004). Informality revisited. World Development, 32(7), 1159-1178.
McKenzie, D., and C. Woodruff. (2006). Do Entry Costs Provide an Empirical Basis for
        Poverty Traps? Evidence from Mexican Microenterprises. Economic Development and
        Cultural Change, 55(1), 3-42.
Mead, D.C., and C. Liedholm. (1998). The dynamics of micro and small enterprises in
        developing countries. World Development, 26(1), 61-74.


                                            23
Rakotomanana, F., E. Ramilison, and F. Roubaud. (2003). The creation of an annual labour
       force survey in Madagascar. An example for sub-saharan Africa. InterStat(27), 35-58.
Rakotomanana, F., M. Razafindrakoto, F. Roubaud, and J.-M. Wachsberger. (2010). The
       economic impact of the political crisis on urban households in Madagascar. The
       labour market in Antananarivo in 2010:looking back over ten years. Paris: DIAL.
Razafindrakoto, M., and F. Roubaud. (2002). Madagascar à la croisée des chemins: la
       croissance durable est-elle possible? Afrique contemporaine, 75-92.
Razafindrakoto, M., and F. Roubaud. (2010). La pauvreté urbaine à Madagascar: dynamique,
       déterminants et politiques. In Gastineau, Gubert, Robilliard & Roubaud (Eds.),
       Madagascar face au défi des Objectifs du millénaire pour le développement (pp. 87-
       118). Marseille: IRD.
Razafindrakoto, M., F. Roubaud, and C. Torelli. (2009). Measuring the informal sector and
       informal employment: the experience drawn from 1-2-3 surveys in African countries. .
       African Statistical Journal, 9(Special Issue), 88-147.
Udry, C, and S Anagol. (2006). The return to capital in Ghana. The American Economic
       Review, 96(2), 388-393.




                                            24
8 Tables and figures
 60


 50


 40


 30


 20


 10


   0
        1995      1996      1997      1998      1999        2000     2001      2002      2003    2004     2005
 -10


 -20
                            Inflation         Interest rate           GDP           GDP per capita

                         Figure 1: GDP growth, inflation and interest rate
                          (Source: World Development Indicators, The World Bank.
                    Note: The interest rate is the nominal lending rate to private customers.)



Table 1: Basic characteristics of informal firms
                                                    Overall                   By quartile of capital stock
                                              Mean       Median         1st        2nd        3rd         4th
Age of the enterprise                           8.7         5.0         8.8        9.0       8.3           8.7
Age of owner                                   38.4        38.0        37.3       38.9       38.1         39.9
Owner's education (in years)                    7.0         6.0         5.6        6.4       7.8           8.9
Experience of owner (in years)                  8.2         5.0         8.5        8.4       7.8           7.9
Female owner (=1)                               0.5         1.0         0.5        0.5       0.5           0.4
Size of firm (total staff)                      1.5         1.0         1.2        1.5       1.6           1.8
Pure self-employment (=1)                       0.7         1.0         0.9        0.7       0.6           0.5
Monthly value added(1000 FMg)                 491.1       201.0       287.3      369.7      565.7        884.6
Capital (1000 FMg)                           2,681.6      341.0        21.5      254.7 1,347.1         11,627.3
Zero capital stock (=1)                         0.1         0.0         0.3        0.0       0.0           0.0
# Observations                                3,995       3,995       1,025       990       1,006         974
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Quartiles are calculated separately for each year. Means are weighted using sampling weights. Monthly
value added and capital stock are real values, deflated using the Consumer Price Index (source: INSTAT).




                                                       25
Table 2: Basic characteristics of informal firms by sector
                                  Food      Textile    Other       Construction     Trade    Services    Catering     Transport
                                                        ind.
Age of the enterprise               5.0     11.1        11.1          11.5             6.9       9.4       5.6              7.9
Age of owner                       36.6     40.1        39.5          39.5            37.7      37.4       38.1            40.4
Owner's education (years)           7.2      8.1         6.3           6.2             6.6       6.8       7.3              8.2
Experience of owner (years)         4.5     10.6        10.0          11.1             6.2       9.1       5.6              7.8
Female owner (=1)                   0.5      0.8         0.2           0.0             0.6       0.5       0.7              0.1
Size of firm (staff)                2.0      1.3         1.8           1.8             1.5       1.3       2.1              1.4
Pure self-employment (=1)           0.4      0.8         0.6           0.7             0.6       0.9       0.4              0.8
Value added                        465       390        545           800             471        362       690             853
Capital                           2,556     2,140      3,138          702            1,614      2,252     2,897           14,948
% in lowest capital quartile        5.6     16.9        19.2          44.2            35.8      50.9       4.1              5.4
% in highest capital quartile      20.2     23.8        25.8           3.9            13.1      15.5       24.2            79.7
Initial investment                2,814     2,814      2,662          594            1,242      2,189     2,449           13,039
Months to recover initial           4.4      8.1         3.0           0.2             0.6       1.3       2.5             16.0
investment
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Means are calculated using sampling weights. Value added, capital stock and initial investment are
expressed in 1,000 Fmg. Value added is monthly. The number of months necessary to recover initial investment
is calculated at the median monthly profit of the sector.




Table 3: Share of employment by institutional sector 1995-2010 (%)
                                    1995     1996     1997     1998    1999   2000       2001    2002    2004     2006      2010
Public sector                        14.2     14.3    13.0     13.2    13.1       10.6   10.7     11.2   10.4      8.8       7.8
Private formal sector                25.1     22.6    22.9     24.6    24.2       25.3   25.9     24.9   22.2     25.2      22.4
EPZs                                  3.1      4.4     4.6      5.5     6.7        8.9   10.2      4.1    8.9      8.0       4.8
Informal sector                      57.6     58.8    59.6     56.7     56        55.3   53.1     59.9   58.4      58       65.1
 dependent                           27.6     22.6    25.2     22.8    21.5       20.3   18.9     21.3   19.9     21.7       30.6
 independent                         30.0     36.2    34.4     33.9    34.5       35.0   34.2     38.6   38.5     36.3       34.5
Total                               100 100 100 100 100 100 100 100                                      100        100      100
Total (1,000 jobs)                  415 435 455 476 500 530 540 538                                      604        636      746
Source: 1-2-3 Surveys, Phase 1, 1995-2010, INSTAT/DIAL/MADIO; authors' calculations.
Note: Private formal sector figures do not include EPZs.




Table 4: Level and growth rates of earnings by institutional sector 1995-2010
                          Level                                Real growth rate (1995=100)
                          1995    1996      1997      1998     1999 2000 2001 2002      2004 2006                          2010
 Public sector             185     95.8     111.2     129.8    144.1 157.9 164.7 157.6  158.0 166.5                        155.4
 Private formal
                        136      91.7 123.3 137.8 133.5 143.2 143.8 148.3 149.8 151.7                                      126.0
 sector (excl. EPZs)
 EPZs                    79     131.4 137.8 152.1 155.4 167.7 168.5 169.9 171.8 176.2                                      176.8
 Informal sector         69     112.1 125.6 143.8 161.9 165.3 166.2 147.8 153.6 158.0                                      138.5
 Total                  103     100.1 117.8 136.3 145.4 150.1 153.1 144.9 145.4 148.6                                      128.6
Source: 1-2-3 Surveys, Phase 1, 1995-2010, INSTAT/DIAL/MADIO; authors' calculations.
Notes: The first column corresponds to monthly earnings in 1,000 Fmg; the other columns to the earnings
dynamics compared to 1995.




                                                              26
Table 5: Basic characteristics and sectoral distribution of informal firms by year
                                                1995              1998             2001             2004
Characteristic of firm owner
Age of owner                                    37.1              38.3             37.6              39.8
Owner's education (in years)                     7.0              6.8               7.1               7.0
Experience of owner (in years)                   7.4              8.0               7.7               9.1
Female owner (=1)                                0.5              0.5               0.5               0.5

Characteristics of firms
Age of the enterprise                            8.1              8.3               8.4               9.6
Time in business, 1 year or less                0.22              0.17             0.20              0.16
Time in business, 2 years                       0.10              0.09             0.10              0.09
Time in business, 3 years                       0.08              0.07             0.12              0.05
Time in business, 4 years or more               0.59              0.67             0.59              0.70

Size of firm (total staff)                       1.5              1.4               1.5               1.4
Pure self-employment (%)                        63.6              71.6             69.2              74.1
Two workers (%)                                 26.2              18.1             18.0              17.8
Three workers (%)                                6.1              7.1               7.3               3.9
Four workers (%)                                 2.9              1.8               3.4               1.8
Five or more workers (%)                         1.2              1.4               2.2               2.4

Monthly value added (1000 FMg)                  400.9            465.1             521.2            529.4
Capital (1000 FMg)                             2,554.9          3,149.2           2,824.3          2,306.3

Sectoral distribution of informal firms (%)
Food processing                              4.0                   3.2              2.6               1.4
Clothing and apparel                        20.3                  17.0             13.5              14.9
Other industry                               7.0                   7.3              9.0               8.4
Construction                                 3.6                   6.6              7.4               8.0
Trade                                       39.9                  36.9             36.8              31.4
Services                                    16.3                  20.1             22.8              25.4
Catering                                     4.2                   3.5              2.8               5.6
Transport                                    4.7                   5.5              5.0               5.0

Total (absolute #)                            122,546           161,838          203,215           237,704
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Means and percentages are calculated using frequency weights. The total is the absolute number of firms
estimated using frequency weights. Monthly value added and capital stock are real values, deflated using the
Consumer Price Index (source: INSTAT).




                                                     27
Table 6: Sectoral distribution of employment and share of institutional sectors within
each activity
                                                   1995              1998              2001              2004
Food processing                                     3.3               3.2               2.6               2.2
                               Formal sector               33.8             38.1              43.6              54.7
                   Informal sector, dependent              23.7             23.7              24.9              16.5
                 Informal sector, independent              38.9             31.9              29.8              21.7
Clothing and apparel                               17.4              15.3              18.3              16.3
                               Formal sector               24.3             17.2                 9               3.7
                                         EPZ               17.9              34               55.5              55.2
                   Informal sector, dependent              24.4             15.2               8.5               8.3
                 Informal sector, independent               33              33.3              26.8              32.7
Other industry                                      7.4               7.9               8.7                8
                                Public sector              11.3             14.2               8.3               8.6
                               Formal sector               41.1             41.4              37.1              33.3
                   Informal sector, dependent              24.8             12.5              17.5              22.5
                 Informal sector, independent              21.8             28.3              33.1              33.2
Construction                                        4.7               5.7               5.6               6.8
                               Formal sector               42.5             34.2              32.4              21.2
                   Informal sector, dependent               31              26.3              22.4              31.4
                 Informal sector, independent              21.4             33.9              43.8              44.8
Trade                                              22.6              22.7              22.4               21
                               Formal sector               29.7             26.8              29.4              23.1
                   Informal sector, dependent              19.2             16.2              16.6              15.2
                 Informal sector, independent              50.2             56.4              53.4              61.6
Services to households and firms                   36.5              36.3              33.8              36.5
                                Public sector              36.3             32.4              28.6              26.9
                               Formal sector                17              19.8              25.6              25.2
                   Informal sector, dependent              33.4             30.9              25.2              23.5
                 Informal sector, independent              13.4             16.8              20.6              24.4
Catering                                            4.1               3.5               2.8               4.3
                               Formal sector               35.1             34.3              38.1              22.4
                   Informal sector, dependent              28.8             28.2              18.7              28.5
                 Informal sector, independent              35.5             33.6              38.9              48.4
Transport                                            4                5.4               5.8                5
                                Public sector              12.2              9.7               6.7               4.1
                               Formal sector               44.1               44              53.2              48.7
                   Informal sector, dependent              20.5             14.5              10.4              12.1
                 Informal sector, independent              23.2             31.7              29.7                35

Total (%)                                           100               100               100              100
Total (1,000 jobs)                                  383               447               512              570
Source: 1-2-3 Surveys, Phase 1, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Figures in italics are the percentage of each institutional sector in the employment of a given sector. Total
does not add up to a hundred because institutional sectors with a share inferior to 5% are not shown, to lighten
the table. Percentages are calculated using frequency weights. The total is the absolute number of workers
estimated using frequency weights, excluding the agricultural sector.

Table 7: Actual employment of workers who previously owned an informal textile firm
                                                  Sector of employment at the time of the survey
                                       EPZ        Informal     Informal     Informal     Other              Total
Year left the previous job                         textile       trade      services
1990-1995                              12.2         20.0         16.1         11.1        40.6             100.0
1996-1998                              16.4          7.2         23.2          8.3        44.8             100.0
1999-2001                              14.3         15.0         24.2         15.8        30.7             100.0
2002-2004                              11.8         10.8         26.1         12.8        38.5             100.0
Total                                  13.6         14.5         21.1         11.8        38.9             100.0
Source: 1-2-3 Surveys, Phase 1, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Calculations use frequency weights and are done on the sample of workers who previously owned an
informal firm in the textile sector and changed jobs after 1990. This yields 315 observations.

                                                          28
             Food processing                                    Clothing and apparel                                    Other industry

1995                                                1995                                                1995

1998                                                1998                                                1998

2001                                                2001                                                2001

2004                                                2004                                                2004



       0    2,000        4,000     6,000    8,000           0   2,000     4,000     6,000     8,000             0    2,000     4,000    6,000   8,000


                   Construction                                           Trade                                Services to households and firms

1995                                                1995                                                1995

1998                                                1998                                                1998

2001                                                2001                                                2001

2004                                                2004                                                2004



       0           500           1,000      1,500           0   1,000      2,000     3,000     4,000            0       2,000           4,000      6,000


                     Catering                                            Transport                                              Total

1995                                                1995                                                1995

1998                                                1998                                                1998

2001                                                2001                                                2001

2004                                                2004                                                2004



       0   2,000    4,000 6,000      8,000 10,000           0   10,000     20,000    30,000    40,000           0   1,000    2,000   3,000 4,000   5,000

                                                           Capital (1000 FMg)

                                         Figure 2: Capital stock box plot by sector
                     (Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO.)




                                                                         29
                                    Total capital stock (billion FMg)
600
400
200
      0




                ing         rel        ustr
                                           y           on         de        firm
                                                                                s
                                                                                       erin
                                                                                              g           ort   Tota
                                                                                                                    l
            ess          ppa                      ucti        Tra                   Cat               nsp
      pr oc
                   a nd a
                              the r ind      onstr                   ds and                       Tra
    d            g                         C                       l
 Foo        thin            O
                                                             use
                                                                 ho
        Clo                                             to ho
                                                     s
                                                 vice
                                             Ser

                                                          1995                      1998
                                                          2001                      2004


                                    Figure 3: Total capital stock (billion Fmg)
                         (Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO.)




                                                                    30
Table 8: Share of firms that contracted, stagnated or expanded since start-up
                                                       Contracted        Stagnated              Expanded
                                                          %                    %                   %
Overall                                                   4.6                 83.2                12.3

By start year
1987-1989                                                   3.6                85.2                11.2
1990-1992                                                   4.3                82.6                13.1
1993-1995                                                   6.5                78.7                14.8
1996-1998                                                   4.0                81.8                14.2
1999-2001                                                   4.2                84.9                11.0
2002-2004                                                   3.9                91.4                4.7

By sector
Food processing                                             0.9                80.9                18.2
Clothing and apparel                                        5.4                87.8                6.9
Other industry                                              5.5                72.1                22.4
Construction                                                9.5                72.3                18.2
Trade                                                       4.0                82.5                13.5
Services to households and firms                            3.9                88.4                7.7
Catering                                                    6.1                76.8                17.1
Transport                                                   3.2                89.3                7.5

By initial size
Pure self-employment                                       0.0                 88.0                12.0
Two workers                                                17.5                71.2                11.3
Three workers                                              24.8                60.0                15.1
Four workers                                               29.1                51.2                19.6
Five or more workers                                       26.5                36.8                36.7

By survey year
1998                                                        3.8                 85                 11.2
2001                                                        5.3                80.3                14.4
2004                                                        4.5                84.3                11.1
Source: 1-2-3 Surveys, Phase 2, 1998-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Percentages are calculated using frequency weights. Firms aged 10 years and less included in the table.
The year 1995 is not included because the data on the number of workers at start-up is not available.




                                                     31
Table 9: Firm growth determinants, OLS and Ordered Probit regressions
                               (1)                          (2)                             (3)
                      Employment growth,             Employment growth,             Employment growth
                       absolute per annum            compound annual rate             (1=Contracted,
                                                                                       2= Stagnated,
                                                                                       3=Expanded)
                               OLS                            OLS                     Ordered probit

Characteristics of owner
Age when                  0.000     (0.001)         0.015     (0.042)                -0.000        (0.003)
started firm
Female (=1)              -0.011     (0.015)        -1.240     (1.045)              -0.193***       (0.073)
Married (=1)            0.031**     (0.015)      3.553***     (1.057)                0.081         (0.075)
Education (in          0.005***     (0.002)        0.228*     (0.121)                0.011         (0.008)
years)
Number of informal firms owned by the household (ref: one firm)
Two firms               -0.024*     (0.014)       -1.674*     (0.979)                -0.053        (0.069)
≥ Three firms            -0.011     (0.020)         0.484     (1.381)               -0.173*        (0.097)
Number of workers at start-up (ref: one worker)
2 workers             -0.049***     (0.017)     -5.921***     (1.178)              -0.480***       (0.082)
3 workers             -0.119***     (0.028)     -9.496***     (1.926)              -0.717***       (0.132)
≥ 4 workers            -0.085**     (0.037)     -7.330***     (2.575)              -0.486***       (0.165)
Age of the firm (ref: 1 year old)
2 years old              -0.024     (0.040)      -6.462**     (2.740)                0.009         (0.192)
3 years old               0.026     (0.042)        -4.083     (2.922)               0.409**        (0.204)
≥ 4 years old            -0.029     (0.027)     -7.706***     (1.885)                0.149         (0.132)
Sector of activity (ref: Trade)
Food                      0.012     (0.032)         0.498     (2.224)               0.335**        (0.155)
Textile                  -0.012     (0.023)        -1.090     (1.569)                -0.124        (0.111)
Other industry         0.062***     (0.024)       3.432**     (1.639)                 0.145        (0.112)
Construction              0.011     (0.027)        -0.883     (1.834)                 0.032        (0.124)
Services               -0.038**     (0.019)      -2.907**     (1.317)              -0.230**        (0.093)
Catering               0.086***     (0.027)      7.006***     (1.851)               0.301**        (0.127)
Transport              -0.058**     (0.026)      -3.841**     (1.796)              -0.257**        (0.127)
Year business was started (ref: before 1995)
1995                     -0.020     (0.029)        -0.958     (2.010)               -0.271*        (0.140)
1996                      0.003     (0.026)         0.199     (1.809)                 0.026        (0.125)
1997                     -0.004     (0.026)        -1.073     (1.810)                -0.124        (0.127)
1998                      0.009     (0.033)         0.847     (2.248)                 0.067        (0.156)
1999                      0.022     (0.030)         1.412     (2.043)                 0.079        (0.143)
2000                      0.045     (0.028)        3.362*     (1.916)                 0.038        (0.133)
2001                      0.003     (0.049)        -2.296     (3.411)                -0.326        (0.240)
2002                      0.000     (0.044)        -0.231     (3.008)                -0.078        (0.213)
2003                      0.007     (0.043)        -2.326     (3.002)                 0.031        (0.213)
Constant                  0.019     (0.042)       7.321**     (2.891)
cut1, Constant                                                                     -1.817***       (0.208)
cut2, Constant                                                                      1.047***       (0.204)
r2                        0.042                     0.067
N                         1,885                     1,885                            1,885
Source: 1-2-3 Surveys, Phase 2, 1998-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: The year 1995 is not included because the data on the number of workers at start-up is not available.
Regressions run on the sub-sample of firms aged 10 years and less. Standard errors in parentheses.




                                                    32
Table 10: OLS estimates of returns to capital
                           all              low levels of capital   medium levels of capital    high levels of capital
                log OLS      no-log OLS   log OLS      no-log OLS    log OLS       no-log      log OLS         no-log
                                                                                    OLS                         OLS
Log capital     0.14***                   0.10***                     0.12**                   0.43***
                 (0.01)                    (0.02)                     (0.05)                    (0.06)
Capital (1000                 0.02***                   0.22*                     0.15***                       -0.00
FMg)
                               (0.00)                   (0.12)                      (0.06)                    (0.01)
Capital                         0.00                    -0.00                      -0.00**                    0.00**
squared
                               (0.00)                   (0.00)                     (0.00)                     (0.00)
Log labor       0.56***                   0.54***                    0.57***                   0.48***
                 (0.02)                    (0.03)                     (0.04)                    (0.08)
Total labor                   0.81***                   0.32**                    0.97***                    1.46***
                               (0.11)                   (0.15)                     (0.16)                     (0.28)
Total labor                   0.00***                  0.00***                      0.00                       0.00
squared
                              (0.00)                    (0.00)                     (0.00)                     (0.00)
Owner's         0.06***      20.43***     0.06***      12.15***      0.05***      17.41***     0.05***       34.49***
education (in
years)
                 (0.00)       (2.37)      (0.01)        (2.44)        (0.01)       (4.52)       (0.01)        (8.29)
Experience of   0.01***       1.98**      0.01**         1.19          0.00         0.57         0.01          5.16
owner (in
years)
                  (0.00)       (0.84)       (0.00)      (0.78)        (0.00)        (1.83)      (0.00)        (3.42)
Female owner    -0.30***     -78.01***    -0.28***    -77.03***     -0.30***      -69.59**     -0.29**      -150.78**
(=1)
                  (0.04)      (19.52)       (0.06)       (18.33)      (0.07)       (34.43)      (0.11)        (75.64)
Year 1998        0.28***     79.59***      0.32***      53.79**      0.21**         52.94      0.41***        152.95*
                  (0.05)      (22.17)       (0.07)       (22.31)      (0.09)       (42.05)      (0.12)        (88.83)
Year 2001        0.48***    131.06***      0.60***     106.98***    0.34***        83.73*      0.48***         126.61
                  (0.06)      (26.38)       (0.07)       (25.24)      (0.10)       (43.08)      (0.13)        (98.04)
Year 2004        0.50***    155.16***      0.60***     128.03***    0.41***      113.80**      0.41***         150.07
                  (0.06)      (24.08)       (0.07)       (23.17)      (0.09)       (44.02)      (0.13)        (95.13)
Food ind.          -0.13       -53.77      -0.23**    -121.77***       0.15         -69.24       -0.09         36.65
                  (0.08)      (34.84)       (0.10)       (28.82)      (0.16)       (63.01)      (0.19)       (159.86)
Textile         -0.42***     -50.52**     -0.44***     -87.94***    -0.35***        -63.49       -0.27          -8.81
                  (0.06)      (24.72)       (0.08)       (22.97)      (0.11)       (47.07)      (0.17)       (117.26)
Other ind.      -0.17***       -12.20     -0.31***     -84.96***      -0.00         74.26        -0.00         72.06
                  (0.06)      (31.90)       (0.08)       (28.10)      (0.11)       (59.65)      (0.17)       (127.96)
Construction     0.34***     68.37**       0.33***      80.92**     0.47***      307.83**        -0.04          -3.67
                  (0.07)      (34.06)       (0.08)       (31.93)      (0.17)      (148.85)      (0.20)       (254.06)
Services        -0.16***      -42.44*     -0.22***     -83.43***      -0.17         -49.90        0.25         151.24
                  (0.06)      (25.31)       (0.07)       (21.46)      (0.11)       (49.23)      (0.15)       (114.53)
Catering           0.03         -5.41        -0.10       -46.49        0.17         18.91         0.27         -81.08
                  (0.07)      (34.62)       (0.09)       (31.35)      (0.12)       (65.22)      (0.20)       (166.62)
Transport          0.04         25.64        -0.10    -144.28***    -0.50***     -166.90**        0.12       238.07**
                  (0.08)      (46.94)       (0.14)       (39.51)      (0.15)       (72.23)      (0.14)       (106.59)
Constant         1.11***     -81.89**      1.35***        44.20     1.31***       -115.72       -1.07*        -219.74
                  (0.13)      (37.96)       (0.19)       (37.91)      (0.46)       (91.02)      (0.57)       (153.20)
R-squared         0.451         0.399       0.398         0.299       0.340         0.259        0.401         0.357
N                 3,456         3,446       1,731         1,761       1,026         1,036         680            676
Mean value          561          447          354          289         547           476         1141            909
added (VA)
Mean capital     3,611         2,880        194          192          1,931         1,928      15,083         14,424
(K)
Implied MRK      0.022         0.022       0.180        0.159         0.034         0.042       0.032          0.005
(at mean VA
and K)
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Note: Capital in thousand FMg. Robust standard errors corrected for clustering at the “segment" level in
parentheses. MRK: marginal monthly return to capital. * p<0.1 ** p<0.05 *** p<0.01.




                                                       33
Table 11: Test of parameter equality across years.
                                       All levels of        Low level of    Medium level       High level of
                                          capital              capital       of capital            capital
Log capital                             0.156***              0.060**          0.059             0.307***
                                         (0.019)              (0.028)         (0.155)             (0.096)
Log capital*year 1998                      0.011              0.083**          0.146              0.326**
                                         (0.026)              (0.042)         (0.177)             (0.140)
Log capital*year 2001                     -0.015                0.028          0.086               -0.008
                                         (0.025)              (0.038)         (0.193)             (0.143)
Log capital*year 2004                     -0.032                0.027         -0.083                0.021
                                         (0.025)              (0.040)         (0.180)             (0.141)
Log labor                               0.565***             0.439***        0.541***            0.899***
                                         (0.045)              (0.057)         (0.094)             (0.107)
Log labor*year 1998                        0.002                0.054          0.122            -0.527***
                                         (0.056)              (0.073)         (0.111)             (0.135)
Log labor*year 2001                       -0.015             0.236***          0.021             -0.349**
                                         (0.057)              (0.070)         (0.113)             (0.136)
Log labor*year 2004                       -0.009              0.157**          0.014              -0.268*
                                         (0.063)              (0.078)         (0.117)             (0.149)
Year 1998                                  0.120               -0.525         -1.703                0.432
                                         (0.338)              (0.407)         (1.562)             (1.337)
Year 2001                                0.633**             -1.013**         -0.429              3.058**
                                         (0.321)              (0.392)         (1.653)             (1.381)
Year 2004                                0.770**               -0.622          1.043                2.063
                                         (0.339)              (0.393)         (1.524)             (1.332)
R-squared                                  0.444                0.394          0.342                0.447
N                                          3,453                1,721          1,028                 6,75
F-Test: All log-capital                    0.275                0.264          0.274                0.064
interactions = 0
F-Test: All log-labor                      0.987               0.003             0.513             0.002
interactions = 0
F-Test: All owner's education              0.607               0.111             0.452             0.038
interactions = 0
Implied MRK 1995                           0.02                0.08               0.02              0.02
Implied MRK 1998                           0.02                0.18               0.06              0.04
Implied MRK 2001                           0.02                0.18               0.03              0.02
Implied MRK 2004                           0.02                0.25              -0.01              0.03
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Controls not shown: owner's education interacted with year, experience, gender, sector dummies. Robust
standard errors corrected for clustering at the “segment� level in parentheses. MRK: marginal monthly return to
capital. * p<0.1 ** p<0.05 *** p<0.01.


Table 12: Marginal returns to capital by sector
                          All levels of         Low level of           Medium level of          High level of
                             capital              capital                  capital                 capital
Food and other ind.          0.024                0.061                    0.024                    0.009
Textile                      0.017                0.048                    0.026                    0.026
Trade                        0.030                0.168                    0.031                   -0.002
Services                     0.021                0.127                    0.057                    0.020
Catering                     0.066                0.191                    0.080                    0.016
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: Implied marginal returns to capital by sector calculated at the mean of capital and value added in each
sector. The elasticities are estimated for each sector and level of capital separately. Full set of regressions
available on request.

                                                       34
Table 13: Pseudo-panel estimation of returns to capital and labor
                                             Log monthly value added           Log monthly value added
                                                       OLS                    Cohort-sector Fixed Effects
Log capital                                          0.195**                              0.089
                                                     (0.078)                            (0.095)
Log labor                                           0.468***                            0.531**
                                                     (0.170)                            (0.198)
Education of owner (years)                             0.028                             -0.022
                                                     (0.042)                            (0.049)
Experience of owner (years)                            0.005                              0.014
                                                     (0.006)                            (0.045)
Female owner (=1)                                  -0.892***                           -0.924**
                                                     (0.310)                            (0.377)
Year 1998                                            0.432**                            0.428**
                                                     (0.169)                            (0.212)
Year 2001                                             0.298*                              0.304
                                                     (0.166)                            (0.284)
Year 2004                                           0.546***                              0.542
                                                     (0.157)                            (0.390)
Textile sector                                        -0.017
                                                     (0.269)
Trade sector                                         0.542**
                                                     (0.224)
Services and catering sector                           0.173
                                                     (0.189)
Constant                                               1.445                            2.265**
                                                     (0.932)                            (1.104)
N                                                        88                                88
Implied marginal returns to capital                     0.04                              0.02
Source: 1-2-3 Surveys, Phase 2, 1995-2004, INSTAT/DIAL/MADIO; authors' calculations.
Notes: * p<0.1 ** p<0.05 *** p<0.01. Both regressions include sector*time interaction variables, not shown in
the table.




                                                     35