MOVING IN AND MOVING UP:

LABOR MARKETS DYNAMICS OF WOMEN AND MEN IN EL SALVADOR

January, 2015

Margarita Béneke de Sanfeliú (FUSADES, El Salvador)

T. H. Gindling (University of Maryland Baltimore County)

José Andrés Oliva Cepeda (FUSADES, El Salvador)

Susana Maria Delgado Vides (FUSADES, El Salvador)

ABSTRACT

We use individual and household level panel data to study labor market dynamics with a focus on what

factors help men and women to achieve advantageous jobs in the labor market and whether those factors

differ between men and women. Specifically, we examine the influence of personal characteristics (such

as education), family characteristics (such as the number of children), and job characteristics (such as

the industry sector of employment) in determining whether a women (or man) moves up into an

advantageous labor market state from an unfavorable state.. We consider three labor market states to be

“advantageous” (“favorable” in Spanish): (1) formal salaried employees, (2) non- agricultural self-

employed workers and employers with a decent wage (defined as reported labor earnings above the legal

minimum wage) and (3) agricultural self-employed workers or employers with a decent wage. We

examine the transitions into and out of these advantageous labor market states and other labor market

and non-labor market states including informal salaried employment, unfavorable non- agricultural

self-employment, unfavorable agricultural self-employment, unemployment and out of the labor force

(distinguishing between those going to school, those engaged in unpaid domestic work, and all other non-

labor force activity).

Our work sheds light on the answers to two key questions: (1) what are the characteristics of the men and

women who move up to an advantageous labor market state from an unfavorable one? and (2) what are

the characteristics of the men and women who fall out of an advantageous labor market states into

unfavorable ones? The answers to these questions contribute to the appropriate design and targeting of

public policy interventions to promote success in the labor market. Our work also sheds light on whether

the characteristics correlated with success in the labor market differ between women and men, and

therefore whether the appropriate design and targeting of policies is different for men and women. Our

findings suggest that education is the most important personal characteristic promoting transitions into

non-agricultural advantageous labor market states and reducing transitions from advantageous labor

market states. In particular, a tertiary (post-secondary) education is a strong predictor of whether a man

or women is found in, and stays in, the most advantageous labor market state, formal salaried employment.

Along with a tertiary education, a secondary education also promotes advantageous non-agricultural self-

employment. For women, dependent children or elderly members in a household reduces the probability

of a transition into an advantageous labor market state. This is likely because unpaid domestic care is still

largely the responsibility of women in El Salvador (especially for a wife/spouse, grandmother or oldest

daughter). This suggests that a key policy intervention to promote advantageous employment for women

would be to make it easier and cheaper to care for children outside of the home while women are working.

1. Introduction

Economic empowerment of women, defined as the “capacity to participate in, contribute to and

benefit from growth processes in ways that recognize the value of their contributions, respect their

dignity and make it possible to negotiate a fairer distribution of the benefits of growth” (OECD,

2011). In the context of this paper, we consider economic empowerment occur when women are

employed in jobs which provide them with a good salary and/or the possibility of improving their

economic status. In our research we study the job, personal and institutional characteristics that

promote the entry of men and women into the labor force and the ability of women to obtain better

jobs once they enter. In particular, the focus of our research is an examination of the characteristics

of women who become empowered and successful in the labor market.

From the Fourth World Conference on Women held in China in September 1995, El Salvador

undertook an action platform for addressing efforts to combat discrimination against women. Thus,

in 1996, was issued the National Policy on Women the Institute for the Development of Women

(ISDEMU) was founded. The Salvadoran government in its Government Plan (2010-2014)

established that gender inequality is one of the most important challenges to overcome. Following

a qualitative evaluation, changes were introduced in the National Policy on Women, so that,

thereafter, its main goal is “to deconstruct the cultural patterns of gender inequality and to give

incentives to build the knowledge that helps create transformative interventions that have a

significant impact on the lives of women”. Thus the efforts of the National Institute for the

Development of Women (ISDEMU) were redirected and, in 2011, the Equality and Elimination of

Discrimination Against Women Act was passed in order to strengthen its institutional framework.

This law determines that ISDEMU is responsible for developing activities, such as training, technical

assistance, technology transfer, tax incentives, access to soft loans, marketing opportunities and to

promote competitiveness, in in order to empower women; it is also responsible for providing

incentives to credit institutions to create special entrepreneurship programs for women (art. 24).

For these programs to be effective at reducing poverty and improving the participation of women

in the process of economic growth, it is important that those designing and running these programs

know the characteristics of the women who could benefit from the programs, as well as the sectors

of the economy where self-employed women or women who are employers in small firms are most

likely to be successful. Many have argued that self-employment in developing countries is not an

indicator of economic empowerment, but rather that workers in developing countries become

self-employed because they are rationed out of formal sector jobs (Fields, 1975; Tokman, 2007, de

Mel, et al, 2010). On the other hand, some point to other self-employed who are innovative and,

successful entrepreneurs (de Soto, 1989; Bennet and Estrin, 2007). In our study of whether the

labor market has contributed to the inclusion of women in the benefits of growth, it is important

to distinguish “successful,” “favorable” or “advantageous” self-employment and small-scale

entrepreneurship, which can promote the inclusion of women in the benefits of development,

from self-employment that exists only because women cannot find the jobs they want in the formal

sector. We consider three labor market states to be “advantageous” (“favorable” in Spanish): (1)

formal salaried employees, (2) non-agricultural self-employed workers and employers with a

decent income, defined as reported labor earnings above the legal minimum wage or (in the

appendix) a household income above the poverty line, and (3) agricultural self-employed workers

or employer with a decent income. We believe that these states are advantageous because they

contribute to meeting basic needs of workers and their households, while providing them with the

ability to respond to adverse situations and in the case of women contribute to the process of

economic empowerment.

Our definition of “advantageous” labor market states is similar to the International Labor Office

(ILO) concept of “decent work.” Two key components of the ILO concept of “decent work” are

remunerative employment and social security (Ghai, 2003). “Social security serves to meet

people’s urgent subsistence needs and to provide protection against contingencies, and as such is

an important aspect of decent work” (Ghia, 2003, p.122). Our first advantageous labor market

state is formal salaried employment, defined as paid employment where workers are insured by

social security. Remunerative employment is work that pays sufficiently to allow a worker’s family

to live at an adequate level. “For developing countries, a good indicator of remunerative work is

provided by data on absolute poverty” (Ghia, 2003, p. 119). The ILO suggests that a good indicator

of whether workers do not have remunerative employment: the first defines advantageous self-

employment as those whose labor earnings are above the legal minimum wage, the second (in the

appendix) defines advantageous self- employment as those who live in a household with an income

above the poverty line. We do not use the term “decent work” in this paper because our measure

of advantageous labor market states does not take into account other components that the ILO

considers when defining decent work, such as basic worker rights and social dialogue (i.e. access

to collective bargaining).

Our study sheds light on the answers to two key questions: (1) what are the characteristics of the

men and women who move up to an advantageous labor market state from an unfavorable one?

and (2) what are the characteristics of the men and women who fall out of an advantageous labor

market states into unfavorable ones? The answers to these questions contribute to the

appropriate design and targeting of public policy interventions to promote success in the labor

market. Our work also sheds light on whether the characteristics correlated with success differ

between women and men, and therefore whether the appropriate design and targeting of policies

is different for men and women. Along with the qualitative study of female employees and self-

employed workers, this paper will inform the debate on the following questions: (1) What public

policies would support the ability of women to benefit from growth? (2) What is the role of formal

wage and salaried employment and small-scale entrepreneurship in supporting the ability of

women to benefit from economic growth? and, (3) What public policies would best support the

ability of women to become successful formal sector employees or successful small-scale

entrepreneurs?

In the empirical work presented in this paper we use panel data to study the determinants of job

mobility for women into and out of the labor force (distinguishing between those going to school,

those engaged in unpaid domestic work and those in other non-labor market states), into and out

of unemployment, and into and out of formal salaried employment,1 informal salaried

employment, self-employment, and employer/owner. We also conduct a similar analysis for men

and compare whether the determinants of job mobility differ between men and women in El

Salvador. We extend the existing literature on labor market mobility in Latin America by separating

self-employment into “advantageous” and “unfavorable”.

We focus our study on what factors help women to achieve advantageous jobs in the labor market.

Specifically, we examine the influence of personal characteristics (such as education), family

characteristics (such as the number of children), and job characteristics (such as the industry sector

of employment) in determining whether a women (or man) moves up into an advantageous labor

market state from an unfavorable state. Because we have panel data on individuals, we will also

be able to pay particular attention to the timing of the transition process. For example,

Cunningham and Bustos Sakvagno (2011) suggest that many successful self-employed workers and

entrepreneurs first spend a short time as informal sector employees acquiring job- relevant skills,

and then move on to formal jobs or return to school, and only then start their own small

businesses. Even then, they may spend short periods of time in informal wage paying jobs or more

education on their way to long-term successful employment states. This suggests that to

understand what makes a woman successful in the labor market and as a small-scale entrepreneur

it is important to simultaneously study the movement of women between formal employment,

informal employment, advantageous self-employment, education and other non-employment

states. For example, if informal or formal employment is an important stepping stone to becoming

a successful entrepreneur, then public policies to promote entrepreneurship might actually need

to promote wage and salaried employment for those workers (at least when they first enter the

labor force), rather than enact policies that encourage women to move directly from non-

employment to self-employment or ownership of small-scale firms. This type of study can only be

done with panel data that follows individuals over time. Firm level enterprise surveys, which follow

enterprises (and not people) over time, would not allow the researcher to examine these

transitions between informal employment, formal employment, schooling, unpaid domestic work

and self- employment.

We find that in El Salvador men are more mobile than women; over 22% of men change states

from one year to the next, compared to only 12% of women. We find substantial mobility of both

men and women between self-employment and the informal sector. For women (but not men)

there is also substantial mobility between the labor market and unpaid domestic care. There is

substantially less mobility out of (and especially into) formal salaried employment than in any other

sector, for both men and women.

For both men and women, the most important characteristic promoting transitions into

advantageous labor market states is education. Any additional education will increase the

probability of transitioning into the salaried formal sector, although tertiary education is the most

1 Originally we had separated part-time and full-time salaried formal employment. However, we found that the few

part-time salaried formal employees had earnings and other characteristics similar to full-time salaried employees

and so decided to combine part-time and full-time into one salaried formal sector.

advantageous for the salaried formal sector while a secondary education is most advantageous for

advantageous non-agricultural self-employment. Our results suggest that, compared to women,

education has a bigger impact on the ability of men to transition into advantageous labor market

states. This suggests that educated women may face a “penalty” in searching for advantageous

employment.

Very young workers are not likely to transition into advantageous non-agricultural self-

employment. It is most likely that a worker transitions into advantageous non-agricultural self-

employment in their 30s or 40s. This suggests that workers most likely to be successful in self-

employment obtain experience before becoming self-employed.

Remittances and other non-labor income promote advantageous non- agricultural self-

employment. Receiving more remittances and other non-labor income increases the probability of

transitioning into advantageous non-agricultural self- employment and decreases the probability

of transitioning out of this advantageous state. This is true for both men and women. For men,

receiving more remittances and other non-labor income decreases the probability of transitioning

into formal sector employment. In general, our results suggest that receiving more remittances

decreases the probability of transitioning into formal sector employment but increases the

probability of transitioning into advantageous self-employment. For women, an increase in

remittances and other non-labor income also increases the probability of leaving the labor force

for unpaid domestic care or other types of non-labor activity.

Access to electricity and other public services increases the probability that both men and women

transition from an unfavorable state into advantageous non-agricultural self-employment. This

suggests that policies of providing access to public services to poor families can also be thought of

as policies to promote advantageous self-employment.

Female spouses are less likely to be salaried formal employees compared to other sectors, while

male spouses are more likely to be in salaried formal employment. A higher number of dependent

members in the household (i.e. too young or too old to work) make it less likely that women (but

not men) are found in the salaried formal sector. A higher number of dependent children (7-19

years old) also reduce the probability that women (but not men) transition into formal salaried

employment. The difference between men and women is likely a consequence of the traditional

expectation that wives will provide unpaid domestic care to children and other dependents, while

husbands are expected to have full-time jobs outside of the home. Surprisingly, the number of

dependent family members does not have a significant influence on the probability that women

will transition into advantageous non- agricultural self-employment. This may be because women

are attracted to self-employment for the flexibility it offers.

The next section of this paper describes the panel data used in this paper. Section 3 describes and

compares the labor market characteristics of men and women in the Salvadoran labor market and

how these characteristics differ between men and women. In this section we use multinomial logit

analysis to identify the personal and job characteristics of men and women in each of the 11 labor

market states, and how those characteristics differ between men and women. Section 4 uses the

panel nature of our data to measure the degree to which men and women move from unfavorable

states to advantageous labor market states, and vice-versa, from one year to the next. Finally,

section 5 identifies the characteristics that determine whether or not men and women transition

into or out of successful labor market states. Section 6 concludes and presents policy

recommendations.

2. Data

To study the labor markets dynamics of women and men in El Salvador, we created an annual panel

data set using the Multipurpose Household Surveys (EHPM) from the years 2008-2012. These

survey data have been collected by the General Directorate of Statistics and Census (DIGESTYC)

since 1975, although we only use the surveys from 2008 to 2012 because it is only in those years

that the necessary variables are available to allow us to match individuals across surveys.

The EHPM includes information related to social, economic and demographic aspects of

households, aimed for diagnosis, planning and evaluation purposes of the country. In addition, the

survey contains the most comprehensive household’s information for both the rural and urban

areas of the country. This data set allows us to follow women and men as they change jobs or as

the job characteristics change.

The EHPM is based on a census mapping technique, developed by DIGESTYC. Every five years,

DIGESTYC updates and renews the sampling frame of households. During a five year period, the

sample is created in a single rotating base from the same sampling framework. The base census

map divides the territory into basic units called “segments”, consisting of one or more blocks (each

block with a group of 12-16 households) and there is a specific unique number for each segment.

Some segments are kept and others are partially rotated every year to reduce the non-response

bias over time.

The division of segments allowed us to create the panel data for the years 2008-2012. The total

sample contains 407,737 observations – 194,508 males (47.7 percent) and 213,229 females (52.3

percent). However, our analysis is restricted to the working age population, therefore, 266,546

observations – 122,403 males (45.9 percent) and 144,143 females (54 percent). From this sample,

165,360 belong to the economically active population – 101,089 males (61.1 percent) and 64,271

females (38.9 percent).

To create the panel data we matched households, then household heads, and finally all the

members of the household. This methodology considers key variables such as the segment

number, geographical location, year of birth, gender and age of each individual to create a unique

identifier that allows a matching process of the head of household and the members living in the

household for each home. We are able to match the same individuals across two consecutive years,

but cannot follow the same individuals for more than two years. Thus, our data effectively consists

of four panel data sets, each of which follows households and individuals for two years (2008-2009,

2009-2010, 2010-2011 and 2011-2012).

On average, 23 percent of the EHPM observations were repeated the following year. Between

2008 and 2009, observations were repeated for 22.6 percent of the 2008 survey sample; from 2009

and 2010, 21.6 percent; for the 2010 and 2011 period, 23.9 percent; and finally, between 2011 and

2012, 23.9 percent as well. The percentages for women are as follows, 22.4 percent from the

period 2008 and 2009; 21.6 percent, between 2009 and 2010; 23.7 between 2010 and 2011, and

finally 24.1 percent between 2011 and 2012. These results are shown in the appendix A (Table A1).

In order to check the representativeness of the panel data sample, we compared some basic

characteristics of the panel data with the full cross-sectional data set (Table 1). The distribution of

workers by gender, region of the country, economic activity and the distribution by employment

status are all similar in the panel and full data.

3. Women in the El Salvador labor market

3.1. Background

The labor market in El Salvador is characterized by a high overall labor force participation

rate for men, an unemployment rate of 6.6 percent on average for the period of 2008-2012

and low level of formal sector employment in general (Table 1). According to the World Bank

statistics, El Salvador had the lowest economic growth among Central American countries

during the period that we study (2008-2012). Remittances from relatives living abroad represent

one sixth of the country’s GDP, and they continue to show increasing trend overtime.

Economic growth plays a pivotal role in the creation of more jobs in El Salvador; as economic

growth increases, unemployment declines. In El Salvador, economic growth shows a high

correlation with job creation rates. Figure 1 shows the trends of unemployment vs. economic

growth for the years being examined. Figure 1 is consistent with the hypothesis that low economic

growth provokes higher unemployment rates.

From table 1 it can be observed that for the years that we study about 47 percent of women

participate in the labor market compared to about 81 percent of men. From 2008 to 2012 the

labor force participation rate of women has been steady and remains substantially below that of

men. Also, the distribution of the labor force by employment status is different for men and

women; 10 percent of women focus on paid domestic service, while almost no men belong to

this status. Men are more likely than women to be permanent or temporary wage and salaried

employees. On the other hand, self- employment is higher for women than men (Appendix, Table

A3).

Table 1. Descriptive statistics in the Panel data and full EHPM samples (2008-2012)

Source: EHPM (2008-2012)

Figure 1 Occupation vs. Economic Growth – 1995-2013

All Panel Data All Panel Data All Panel Data

All observations 100.0 100.0 52.6 52.3 47.4 47.7

By Labour Market Concept

Economically active population 62.8 62.2 47.4 46.7 81.2 80.4

Urban 65.6 65.2 67.0 66.5 64.0 63.5

Rural 34.4 34.9 33.1 33.5 36.0 36.5

Unemployed 6.6 6.6 4.5 4.5 8.1 8.0

By Education

No Education 12.4 11.6 12.4 11.8 12.4 11.5

Incomplete primary 23.2 23.0 22.8 22.3 23.5 23.5

Complete primary 14.5 14.5 13.7 13.6 15.1 15.1

Incomplete secondary 18.8 19.2 16.6 17.2 20.4 20.6

Complete secondary 17.9 17.7 18.8 18.3 17.1 17.2

Somekind of higher level 13.3 14.0 15.8 16.8 11.5 12.1

By Sector

Agriculture and mining 20.7 21.1 5.3 5.4 31.8 32.1

Manufacturing/Construction 16.2 16.9 11.1 12.3 19.9 20.2

Textile manufacturing 4.8 4.7 7.6 7.6 2.7 2.7

Commerce 28.5 27.7 41.5 41.6 19.1 18.0

High complexity services 18.2 18.7 16.2 16.8 19.6 20.0

Low complexity services 11.7 10.9 18.3 16.3 7.0 7.1

By Categories

Formal salaried 16.3 16.2 12.3 11.5 21.1 21.7

Informal salaried 17.9 15.6 11.1 9.3 26.1 23.0

Advantageous self-employed not agriculture 5.8 6.2 4.9 5.0 6.9 7.6

Unfavorable self-employed not agriculture 9.5 10.3 12.9 14.1 5.5 5.8

Agricultural advantageous self-employed 1.5 1.9 0.2 0.2 3.1 3.8

Agriculture unfavorable self-employed 3.3 3.8 0.5 0.7 6.7 7.4

Paid domestic workers 4.2 4.0 3.5 3.4 5.2 4.7

Unemployed 4.1 4.2 2.1 2.2 6.5 6.5

Non-labor / studing 8.3 8.7 8.2 8.5 8.4 9.0

Non-labor force at domestic work 20.2 20.4 36.4 37.2 0.9 0.7

Non-labor force / others 8.7 8.8 8.0 7.8 9.5 9.9

All observations Women Men

Notes: 2007 data is not being used due to change in sampling frame.

See appendix A, table A2

Source: own elaboration with data from EHPM y BCR

The EHPM includes a direct question for the reasons of not seeking a paid job. As noted above,

there are more women out of the labor force than men. 69 percent of women state that they

are out of the labor force because of family obligations or unpaid domestic work responsibilities

– 14.8 percent state that they are studying and 11.7 percent report disability or illness. In

contrast, the main reason for men not seeking a job is because they are studying (40 percent,

appendix A, Table A4).

The labor market dynamics of men and women are conditioned by the traditional division of labor

by gender. For instance, the industry sectors where people are likely to work differ between men

and women. Women have a higher participation in commerce, 42 percent vs. 18.8 percent for

men in 2012; low complexity services2 (18.6 percent vs. 7.6 percent for men); and textile

manufacturing (6.9 percent vs. 2.5 percent for men). In the case of men, most work in the

agriculture sector; (34.3 percent of men work in agriculture compared to 5.3 women in 2012),

followed by manufacturing and construction (19.1 percent vs. 11.2 percent of women); and

high complexity services (17.9 percent vs. 15.9 percent for women, Table 1).

Regarding education, women achieve higher education levels than men. For example, in 2012 20.1

percent of women had completed secondary education vs. only 18.5 percent of men; 15.7 percent

of women had some kind on tertiary education vs. 11.9 percent of men (Table 1).

Figure 2 portrays an interesting insight about the income gap between men and women by

education level for wage and salaried employees and the self-employed. At most educational

levels, the income distribution of men is to the right of that for women, indicating that men

earn more than women at these education levels. For employees, the gap between the earnings

of men and women narrows with education, so that for wage and salaried employees with a

tertiary education the earning gap between men and women is almost zero. On the other hand,

2 Low complexity services include: communitarian, social and personal services, private domesticservices. High complexity services include: utilities, transport, storage, telecommunications, financial services, public administration.

the earnings gap between self-employed men and self-employed women increases as education

level increases.

As education increases individuals opt to enter into the service sector; data shows that 79.8

percent of women who belong to the service sector complete upper level education,

meanwhile 44.2 percent of women working in the commerce section either have no education

or incomplete primary level. In the case of men, the least educated remain in agriculture, and

the most educated tend to be employed in the service sector (Appendix A, Table A7).

Another important fact to consider is that the percentage of workers in formal salaried jobs is

similar for both men and women, 26 percent for both (see Table 2). However, due to the lower

rate of job-market participation, women´s participation in formal jobs is also lower. According to

data from the “Instituto Salvadoreño del Seguro Social” (ISSS), 41.8 percent of covered individuals

are women and 58.2 percent are men (Appendix, Table A8).

Figure 2: Monthly earnings male-female gap between wage employees and self- employed

workers by education level

Source: Data base of Multipurpose Household Survey (EHPM), DIGESTYC, 2008 – 2012.

3.2. Characteristics of men and women in each state

0.2

.4.6

.81

Pro

port

ion

0 100 200 300 400 500Wage

Woman Men

Primary

0.2

.4.6

.81

Pro

por

ción

0 200 400 600 800 1000Wage

Woman Men

Secundary

Cuenta Propia

0.2

.4.6

.81

Pro

por

ción

0 500 1000 1500Wage

Woman Men

Terciary

0.2

.4.6

.81

Pro

port

ion

0 100 200 300 400 500Net income

Woman Men

Primary

0.2

.4.6

.81

Pro

por

ción

0 200 400 600 800 1000Net income

Woman Men

Secundary

0.2

.4.6

.81

Pro

por

ción

0 500 1000 1500Net income

Woman Men

Terciary

Wage earners

Self-employed

To study the movement into and out of advantageous labor market states in El Salvador we

classified men and women into the following states: not in the labor force (divided between

those in school, those engaged in unpaid domestic work, and others not in the labor force),

unemployed, unpaid family work, employed in the salaried informal sector, employed in salaried

formal sector, non-agricultural self- employed/employer, agricultural self-employed/employer.

The last two states are also further divided into advantageous and unfavorable self-employment.

The category "salaried formal sector" includes employees who are benefiting from social security,

either employed full time or part-time. All wage and salaried employees who are not benefiting

from social security are classified as "informal".

The category "Advantageous Non-Agricultural Self-employed/Employer" includes self- employed

workers who are not engaged in agriculture and whose labor earnings are greater than the

legal minimum wage.3

All self-employed workers and employers who are not engaged in

agriculture and that do not meet the above conditions are classified as "Unfavorable Non

Agricultural Self-employed/Employer".

The category "Advantageous Agricultural Self-employed/Employer" includes self- employed

workers who are engaged in agriculture and whose labor earnings are greater than the legal

minimum wage. All self-employed workers and employers who are engaged in agriculture and

that do not meet the above conditions are classified as "Unfavorable Agricultural Self-

employed/Employer".

The "Unpaid family worker" category includes any person who works without remuneration in a

business, firm or family farm. The category "Unemployed" includes people who do not have a

job but who are actively seeking one.

The category "Not in the Labor Force - Student" includes persons 16 years of age or older who

are not part of the labor force and who report to be exclusively devoted to studying.

We distinguished between two further categories of non-students what are not in the labor

force. The category "Not in the Labor Force – Unpaid Domestic Work" includes persons 16 years

of age or older who are not part of the labor force who report to be exclusively devoted to

domestic work. “Other inactive” are those who report to not be in the labor force for any other

reason. This last category includes those who are retired and young people who are neither

3 Multiple minimum wages are set by industry sector in El Salvador during different periods of time. Before May 2008 urban monthly minimum wage was $183 compared to rural minimum wage, $85.8. From May 2008 to December 2008, the urban minimum wage was $192.3 against $90 for the rural area. Finally from 2009 the minimum wage for the urban area was $207.9 and $97.2 for the rural area. For comparisons to results of a companion project in Nicaragua we also constructed a separate set of results that define advantageous” as having a household income per capita above the poverty line; the corresponding results w h i c h are presented and discussed in the appendix B, are similar for both definitions of “advantageous”.

working nor in school (sometimes called “NENE”-Not Employed and Not in Education), and the

disabled.

Table 2 shows the proportion of men and women in each of the eleven labor market states.

Men are more likely to be salaried formal employees than women (21.2 percent vs. 12.3 percent).

The same is true for informal sector employees where 26.1 percent are men vs. 11.1 percent

females. Men are also more likely than women to be in advantageous non-agricultural self-

employment (6.9 percent vs. 4.9 percent), while women are more likely to be in unfavorable

non-agricultural self-employment (12.9 percent for women vs. 5.5 percent for men). Men are

more likely to be agricultural self-employed (either advantageous or unfavorable). In total, men

are more likely than women to be found in advantageous labor market states in El Salvador.

Men are also more likely than women to be unpaid family employees or unemployed. Men

and women are equally likely to be full-time students. Finally, women are more much likely to be

engaged in unpaid domestic work (0.9 percent of men vs. 36.4 percent of women).

Table 2

Distribution of the working age population according to their labor market state, comparison by

gender (2008-2012)

Source: EHPM 2008-2012

We examine the characteristics of men and women in each sector by estimating an equation

where the probability that a man or women is found in status j can be captured using the

multinomial logit technique:

Pij = exp(Σk Bjk Xik) / Σj exp(Σk Bjk Xik) (1)

In this equation, Pij is a multinomial variable that indicates if the individual i is found in status j.

The Xik is a vector of k personal and job characteristics. From the estimated coefficients, Bjk,

Sector

Men Women Men Women

Formal salaried 21.10% 12.30% 26.00% 25.90%

Informal salaried 26.10% 11.10% 32.10% 23.40%

Advantageous self-employed not agriculture 6.90% 4.90% 8.60% 10.30%

Unfavorable self-employed not agriculture 5.50% 12.90% 6.80% 27.20%

Agricultural advantageous self-employed 3.10% 0.20% 3.80% 0.40%

Agriculture unfavorable self-employed 6.70% 0.50% 8.30% 1.10%

Un-paid family work employee 5.20% 3.50% 6.40% 7.30%

Unemployed 6.50% 2.10% 8.10% 4.50%

Non-labor/students 8.40% 8.20%

Non-labor force/unpaid domestic work 0.90% 36.40%

Non-labor force/other reasons 9.50% 8.00%

Total Working Age

Population

Active in the

workforce

we can calculate the marginal impact of each explanatory variable k on the probability of a being

in each state, controlling other personal and job characteristics. Tables 3 and 4 present the

estimates of the marginal impact of each explanatory variable k on the probability of being in each

state, calculated from the multinomial logit coefficient estimates.

For both men and women, older workers are more likely to be in an employed state (both

advantageous and unfavorable) compared to a non-employed state. That is, younger workers

are more likely to be unpaid family workers, students, in unpaid domestic work, or not in the labor

force.

For men and women, more education significantly increases the probability of formal salaried

employment and advantageous non-agricultural self-employment. On the other hand, education

does not increase the probability that either men or women are employed in advantageous

agricultural self-employment (in fact it decreases the probability slightly).

Table 3.

Marginal effects on the probability of being in each labor market state, women

Table 4.

Marginal effects on the probability of being in each labor market state, men

For both men and women, household heads are more likely to be in an advantageous labor

market state (salaried formal employee and advantageous self-employed--both agricultural and

Variables

AGE 0.008 *** 0.009 *** 0.007 *** 0.017 *** 0.000 *** 0.000 *** -0.002 *** 0.000 -0.001 *** -0.034 *** -0.004 ***AGE_squared 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 ***Education level 0.010 *** 0.001 *** 0.002 *** -0.002 *** 0.000 *** 0.000 *** 0.002 *** 0.002 *** 0.000 *** -0.016 *** 0.001 ***Head of household 0.004 *** -0.024 *** 0.053 *** 0.082 *** 0.002 *** 0.007 *** -0.061 *** -0.005 *** -0.001 *** -0.030 *** -0.028 ***Spouse -0.017 *** -0.113 *** 0.016 *** 0.020 *** -0.001 *** 0.000 -0.022 *** -0.018 *** -0.002 *** 0.195 *** -0.059 ***High population density area 0.024 *** 0.031 *** 0.008 *** 0.017 *** -0.001 *** -0.003 *** -0.002 0.006 *** 0.001 *** -0.090 *** 0.010 ***Age 6 or younger -0.005 *** -0.020 *** -0.001 0.000 0.000 0.000 ** -0.007 *** -0.003 *** 0.000 *** 0.042 *** -0.008 ***Age 7 to 18 -0.002 *** 0.004 *** 0.001 *** -0.001 0.000 *** 0.001 *** 0.002 *** 0.000 0.000 *** -0.003 *** -0.002 ***Age 19 to 65 0.002 *** -0.002 ** 0.000 -0.017 *** 0.000 ** -0.001 *** -0.001 ** 0.000 0.000 * 0.019 *** 0.001 **Age 65+ -0.003 *** -0.011 *** 0.001 0.000 0.000 0.000 -0.009 *** -0.002 *** 0.000 0.019 *** 0.005 ***Tubed w ater inside dw elling 0.030 *** 0.044 *** 0.020 *** 0.038 *** -0.001 ** -0.003 *** 0.001 -0.004 ** 0.000 *** -0.122 *** -0.004Electricity 0.000 -0.048 *** 0.023 *** 0.016 *** 0.000 0.000 0.010 *** -0.004 *** 0.000 *** 0.001 0.000Water 0.007 *** 0.000 0.010 *** 0.001 0.000 ** -0.001 *** 0.005 *** 0.001 *** 0.000 *** -0.023 *** 0.001d09 -0.004 *** -0.002 0.001 0.003 0.000 0.000 0.001 0.003 *** 0.000 0.003 -0.005 ***d10 -0.005 *** 0.002 0.002 -0.008 * 0.000 0.001 ** 0.000 0.005 *** 0.000 0.007 -0.004 **d11 -0.007 *** 0.006 * 0.001 -0.005 0.000 0.002 *** 0.002 0.002 ** 0.000 0.006 -0.007 ***d12 -0.007 *** 0.011 *** 0.001 -0.004 0.000 0.001 0.003 * 0.002 * 0.000 0.000 -0.007 ***

Unpaid Worker

UnemployedNLF -

StudentNLF -

DomesticInactive

Formal (salaried)

Informal (salaried)

Advantegous NASE

Unfavorable NASE

Advantegous ASE

Unfavorable ASE

Variables

AGE 0.019 *** -0.011 *** 0.008 *** 0.004 *** 0.001 *** 0.000 -0.007 *** -0.007 *** 0.000 *** -0.001 *** -0.005 ***AGE_squared 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 ***Education level 0.025 *** -0.016 *** 0.005 *** 0.001 *** -0.003 *** -0.011 *** -0.001 *** 0.001 * 0.000 *** 0.000 -0.001 ***Head of household 0.087 *** 0.028 *** 0.058 *** 0.008 *** 0.034 *** 0.044 *** -0.100 *** -0.042 *** 0.000 *** -0.010 *** -0.106 ***Spouse 0.077 *** -0.003 0.048 *** -0.002 0.007 ** 0.018 *** -0.024 *** -0.055 *** 0.000 *** -0.007 *** -0.059 ***High population density area 0.084 *** -0.034 *** 0.014 *** 0.022 *** -0.030 *** -0.067 *** -0.021 *** 0.015 *** 0.000 *** 0.001 0.015 ***Age 6 or younger 0.005 *** 0.010 *** 0.004 *** 0.001 0.002 *** 0.005 *** -0.006 *** -0.010 *** 0.000 *** -0.002 *** -0.011 ***Age 7 to 18 -0.003 *** 0.002 ** 0.000 -0.008 *** 0.004 *** 0.005 *** 0.005 *** 0.000 0.000 *** -0.001 * -0.004 ***Age 19 to 65 0.015 *** -0.014 *** -0.002 *** -0.009 *** 0.001 *** 0.000 0.005 *** 0.004 *** 0.000 *** 0.000 0.000Age 65+ -0.010 *** -0.011 *** 0.002 -0.007 *** 0.002 *** 0.010 *** 0.006 *** 0.004 ** 0.000 * 0.000 0.004 ***Tubed w ater inside dw elling 0.078 *** -0.031 *** 0.029 *** 0.017 *** -0.014 *** -0.053 *** -0.028 *** -0.001 0.000 *** 0.000 0.005Electricity 0.014 *** -0.098 *** 0.030 *** 0.007 *** 0.011 *** 0.001 0.007 *** 0.001 0.000 *** 0.002 ** 0.026 ***Water 0.020 *** -0.045 *** 0.013 *** 0.006 *** -0.003 *** -0.006 *** -0.002 0.007 *** 0.000 *** 0.002 * 0.008 ***d09 -0.013 *** -0.032 *** 0.002 0.001 0.002 0.007 *** 0.004 *** 0.025 *** 0.000 ** 0.002 * 0.001d10 -0.016 *** -0.002 0.002 0.003 0.001 0.002 0.002 0.007 * 0.000 *** 0.003 * -0.001d11 -0.015 *** -0.004 0.004 ** -0.001 0.005 *** 0.006 ** 0.006 *** 0.003 0.000 * -0.001 -0.002d12 -0.018 *** 0.018 *** 0.001 -0.005 ** 0.005 *** 0.009 *** 0.007 *** -0.006 * 0.000 * -0.003 *** -0.007 ***

Unpaid Worker

UnemployedNLF -

StudentNLF -

DomesticInactive

Formal (salaried)

Informal (salaried)

Advantegous NASE

Unfavorable NASE

Advantegous ASE

Unfavorable ASE

non-agricultural).

Female spouses are less likely to be in salaried formal employees, while male spouses are more

likely to be in salaried formal employment. The difference between men and women spouses is

likely a consequence of the traditional expectation that wives will provide unpaid domestic

care to children and other dependents, while husbands are expected to have full-time jobs

outside of the home. Both male and female spouses are also more likely to be in advantageous

non-agricultural and ag r i cu l t ur a l self-employment. Wives may be more likely to be found in

self-employment than in salaried formal employment because self-employment can provide

flexibility in terms of hours worked that allows women to provide domestic care and obtain a salary

in the labor market. Our qualitative analysis suggests that the main disadvantage of formal sector

work for women is the inflexibility in terms of number of hours worked and when work is possible.

The women interviewed in the qualitative analysis suggest that the informal sector is more flexible

in terms of hours or work, location, and other factors.

A higher number of dependent members in the household (i.e. too young or too old to work)

make it less likely that women are found in the salaried formal sector. This is not true for other

advantageous labor market states. That is, a higher number of dependent members in the

household (i.e. too young or too old to work) is not significantly correlated with a smaller

probability that women are found in advantageous self- employment. Again, this may be because

part-time work and non-standard work hours in the salaried formal sector are not common,

making it difficult for women to provide unpaid domestic care and work full-time in the formal

salaried sector.

The negative impact of more dependent household members on participation in the salaried

formal sector is clear for women but not for men. For men, an increase in the number of

children in the household six years old and younger increases the probability that he is found in

the salaried formal sector, while an increase in the number of children ages 7-18 or age 65 and

older decreases the probability of being found in the salaried formal sector. Having more young

children in the household also increases the probability that men are found in other advantageous

labor market states. This may reflect social pressure for fathers to financially support children

(and no pressure for fathers to provide unpaid domestic care). For both men and women, having

more working age (19-64) household members increases the probability of being in the

salaried formal sector.

For both men and women, access to public services (tubed water, electricity and water) increases

the probability that a person is found in the salaried formal sector and advantageous non-

agricultural self-employment.

We include two explanatory variables that measure the impact of income from other sources

than own work; remittances and non-labor income (“non-labor” includes income from capital,

land and the earnings of other family members). For both men and women, receiving a higher

level of remittances and other non-labor income decreases the probability of being a formal

or informal sector employee. Quantitatively, the negative impact of remittances on the

probability of formal sector employment is larger for women than men. For men, receiving more

remittances increases the probability of being in advantageous self- employment (both non-

agricultural and agricultural) and decreases the probability of being in unfavorable non-

agricultural self-employment. For both men and women a higher level of non-labor income

increases the probability of being in advantageous non- agricultural self-employment and

decreases the probability of being in unfavorable self- employment. However, for women

receiving more remittances decreases the probability of being in advantageous non-agricultural

self-employment. For both men and women, receiving a higher level of remittances or other non-

labor income increases the probability that a person is not in the labor force (either a

student, in unpaid domestic work or other inactive). Taken all together, the evidence from

El Salvador suggests that remittances and other non-labor income reduce labor force

participation but promote advantageous self-employment (and reduce the probability of being

in unfavorable self-employment). This is true for both men and women.

The negative impact of remittances and other non-labor income on employment in the formal and

informal sectors is larger for women than men. It may be that women are more likely to leave

formal and informal employment for unpaid domestic care when non-labor income increases

because women are primarily responsible for the care of children and other unpaid domestic

responsibilities. Consistent with this hypothesis, remittances and other labor income have a bigger

positive impact on the probability that men will be found in advantageous non-agricultural self-

employment. To simplify, our evidence suggests that an increase in remittances and other non-

labor income “allows” workers to avoid or leave formal or informal employment, with men instead

becoming advantageous self-employed workers and women becoming full-time unpaid domestic

care providers.

4. Probabilities of transition

The major focus of this study is on the factors that influence women and men to work and the

places where they do, and in particular on what factors influence whether women

dynamically transition into advantageous labor market states. Using panel data we calculate the

probabilities of finding a person i in status j at time t+k, conditional on being in status z at time

t:

Pizj = Pr(Si,t+k = j | Sit = z) (2)

where Sit = the labor market state of individual i in time t.

We estimated matrices of probabilities of transitions into and out of the 11 labor market and

non-labor market states defined earlier. These 11 X 11 matrices were estimated separately

for men and women. The results are shown in F igure 3 , Table 5 and Table 6 , where the

rows represent the origin or the state in the previous period, and the columns represent the

destination status on the next period. The numbers presented in the main diagonal show the

number of people who did not move from any category (had no mobility); numbers off the

diagonal show the amount transitioned to the indicated state.

4.1. Women

Quantitatively, the most mobility among women is mobility to and from unpaid domestic

work; with 31 percent of women in unpaid domestic work in one year in a different state

the next year, and 30 percent of those in unpaid domestic work in one year who were not in

unpaid domestic work the year before. As noted below, there is very little mobility of men to

and from unpaid domestic work. This is an important difference between the labor market

mobility of men and women, and suggests that there policies that provide skills to women

currently in unpaid domestic work may be effective and efficient. There is also substantial

mobility of women to and from unfavorable non-agricultural self-employment and informal

salaried employment.

Fig. 3: The origin of workers found in advantageous labor market states

Women move into the three categories considered to be “advantageous” labor market

categories at smaller rates than do men; in our data 22.7 percent of men move into an

advantageous labor market state each year, while only 12 percent of women do. This suggests

that women do not fully participating in the benefits of economic growth in El Salvador.

The results in table 5 and figure 3 also indicate that women who work in formal salaried

employment have a high probability, 85.4 percent, of remaining in this category. Of those

women who enter formal salaried employment, most come from informal salaried employment

or from unpaid domestic work (see figure 3). Of those women who leave the salaried formal

sector most go to the informal sector.

There is much less stability in non-agricultural advantageous self-employment than there is in

salaried formal employment. Fewer than 50% of women who are in advantageous non-agricultural

self-employment in one year are also found in that sector in the next year. Men are significantly

more likely to remain in advantageous non- agricultural self-employment from one year to the

next compared to women. Of those women who transition into advantageous non-agricultural self-

employment, most come from unfavorable non-agricultural self-employment (see Figure 3).

Unpaid domestic work is the next largest source of women who transition into advantageous non-

agricultural self-employment. Of women who transition out of advantageous non- agricultural self-

employment, most go into unfavorable non-agricultural self- employment or into unpaid domestic

work.

Very few women transition into advantageous agricultural self-employment. This state is

dominated by men.

Most women do not remain unemployed for long; 85% of women who are unemployed in one year

are not unemployed the next. Unemployed women are most likely to leave the labor force and

become unpaid domestic workers or informal sector employees.

4.2. Men

Overall, men exhibit higher mobility than women. As with women, the most significant mobility

among men is into and out of informal salaried employment, with 24.1 percent of men in that

sector in one year leaving it the next year, and 25 percent of those found in that sector in one

year coming from a different sector the year before. As with women, the least mobile men are

salaried formal employees; 83 percent of male salaried formal employees remained in that sector

from one year to the next.

Table 5

Probabilities of transition matrix for women ( Sum of transitions for the period 2009-2012)

Source: For each stata, the first row show the number of observation un each transition, the second row

should the probabilities that in status j at time t+k conditional on being in status Z at time t, the third row

shows the probabilities that an individual in status j at time t+k was in status z a time t.

When men do leave formal sector employment, they are most likely to transition into informal

sector employment or unemployment. Unlike for women, very few men leave formal sector

employment each year to enter unpaid domestic work. Of those men who enter formal

salaried employment, most come from informal salaried employment (see Figure 3). Unlike

women, very few men enter salaried formal employment from unpaid domestic work

Table 6

Probabilities of transition matrix for men (Sum of transitions for the period 2009-2012)

Status1 2 3 4 5 6 7 8 9 10 11 Total in t

Total in

t - Diagonal

%

Mobility

1-Formal salaried employment 1976 112 18 39 0 0 12 34 11 93 18 2313 337 3.6

85.4 4.8 0.8 1.7 0.0 0.0 0.5 1.5 0.5 4.0 0.8 100.0

83.1 4.6 1.5 1.0 0.0 0.0 1.3 6.3 0.7 0.9 0.9 9.0

2-Informal salaried employment 166 1192 49 243 2 10 40 75 40 434 41 2292 1100 11.8

7.2 52.0 2.1 10.6 0.1 0.4 1.7 3.3 1.7 18.9 1.8 100.0

7.0 48.6 4.1 6.5 3.9 3.8 4.5 13.8 2.6 4.1 2.1 9.0

3. Advantageous self-employed not

agriculture 14 52 507 383 1 2 46 7 3 69 14 1098 591 6.3

1.3 4.7 46.2 34.9 0.1 0.2 4.2 0.6 0.3 6.3 1.3 100.0

0.6 2.1 42.7 10.2 2.0 0.8 5.2 1.3 0.2 0.7 0.7 4.3

4. Unfavorable self-employed not

agriculture 28 264 449 1913 2 25 76 23 15 826 89 3710 1797 19.3

0.8 7.1 12.1 51.6 0.1 0.7 2.0 0.6 0.4 22.3 2.4 100.0

1.2 10.8 37.8 51.0 3.9 9.6 8.5 4.2 1.0 7.8 4.6 14.5

0 1 4 9 11 12 2 0 0 28 3 70 59 0.6

0.0 1.4 5.7 12.9 15.7 17.1 2.9 0.0 0.0 40.0 4.3 100.0

0.0 0.0 0.3 0.2 21.6 4.6 0.2 0.0 0.0 0.3 0.2 0.3

0 10 2 26 9 68 6 5 1 110 12 249 181 1.9

0.0 4.0 0.8 10.4 3.6 27.3 2.4 2.0 0.4 44.2 4.8 100.0

0.0 0.4 0.2 0.7 17.6 26.2 0.7 0.9 0.1 1.0 0.6 1.0

7-Unpaid family workers 10 58 48 85 3 3 322 14 78 253 17 891 569 6.1

1.1 6.5 5.4 9.5 0.3 0.3 36.1 1.6 8.8 28.4 1.9 100.0

0.4 2.4 4.0 2.3 5.9 1.2 36.2 2.6 5.0 2.4 0.9 3.5

8-Unemployed 54 90 9 35 1 2 20 76 29 151 10 477 401 4.3

11.3 18.9 1.9 7.3 0.2 0.4 4.2 15.9 6.1 31.7 2.1 100.0

2.3 3.7 0.8 0.9 2.0 0.8 2.2 14.0 1.9 1.4 0.5 1.9

9. Non-labor / studing 47 124 4 35 0 3 92 107 1263 236 33 1944 681 7.3

2.4 6.4 0.2 1.8 0.0 0.2 4.7 5.5 65.0 12.1 1.7 100.0

2.0 5.1 0.3 0.9 0.0 1.2 10.3 19.7 81.3 2.2 1.7 7.6

10. Non-labor force at domestic work 75 503 89 913 20 124 259 191 95 7847 607 10723 2876 30.8

0.7 4.7 0.8 8.5 0.2 1.2 2.4 1.8 0.9 73.2 5.7 100.0

3.2 20.5 7.5 24.3 39.2 47.7 29.1 35.1 6.1 74.1 31.7 41.9

9 46 8 72 2 11 15 12 19 544 1072 1810 738 7.9

0.5 2.5 0.4 4.0 0.1 0.6 0.8 0.7 1.0 30.1 59.2 100.0

0.4 1.9 0.7 1.9 3.9 4.2 1.7 2.2 1.2 5.1 55.9 7.1

Total in T+1 2379 2452 1187 3753 51 260 890 544 1554 10591 1916 25577 9330 100.0

9.3 9.6 4.6 14.7 0.2 1.0 3.5 2.1 6.1 41.4 7.5 100.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 36.5

Total in T+ 1 - Diagonal 403 1260 680 1840 40 192 568 468 291 2744 844 9330

% Mobility 4.3 13.5 7.3 19.7 0.4 2.1 6.1 5.0 3.1 29.4 9.0

5. Agricultural advantageous self-

employed

6. Agriculture unfavorable self-

employed

11-Not in the labor force-any other

state

Source: For each stata, the first row show the number of observation un each transition, the second row

should the probabilities that in status j at time t+k conditional on being in status Z at time t, the third row

shows the probabilities that an individual in status j at time t+k was in status z a time t.

Men who transition into advantageous non-agricultural employment are most likely to come

from informal employment or unfavorable non-agricultural employment, and not from formal

sector employment (see Figure 3). Unlike women, men are not likely to transition into non-

agricultural self-employment from unpaid domestic work.

Compared to other advantageous labor market states there is less stability in advantageous

agricultural self-employment. Table 6 shows that there is a lot of year-to- year movement

between advantageous and unfavorable agricultural self-employment. Almost 30% of

advantageous agricultural self-employed men were unfavorable agricultural self-employed the

year before, and over 26% of men in advantageous agricultural self-employment this year will

be in unfavorable agricultural self- employment next year. This likely reflects the volatility of

STATUS

1 2 3 4 5 6 7 8 9 10 11 Total in tTotal in

t - Diagonal

%

Mobility

1-Formal salaried employment 3349 327 63 38 12 37 20 123 16 5 41 4031 682 7.5

83.1 8.1 1.6 0.9 0.3 0.9 0.5 3.1 0.4 0.1 1.0 100.0

82.2 5.8 4.2 3.2 1.1 1.7 1.8 8.7 1.3 3.7 1.9 18.6

2-Informal salaried employment 379 3340 213 213 126 406 202 406 68 24 142 5519 2179 24.1

6.9 60.5 3.9 3.9 2.3 7.4 3.7 7.4 1.2 0.4 2.6 100.0

9.3 59.5 14.2 17.9 11.3 18.6 17.8 28.9 5.3 17.9 6.7 25.4

3. Advantageous self-employed not

agriculture 51 172 782 242 33 38 23 33 2 8 36 1420 638 7.1

3.6 12.1 55.1 17.0 2.3 2.7 1.6 2.3 0.1 0.6 2.5 100.0

1.3 3.1 52.2 20.3 3.0 1.7 2.0 2.3 0.2 6.0 1.7 6.5

4. Unfavorable self-employed not

agriculture 33 197 250 431 16 61 32 53 8 11 76 1168 737 8.1

2.8 16.9 21.4 36.9 1.4 5.2 2.7 4.5 0.7 0.9 6.5 100.0

0.8 3.5 16.7 36.2 1.4 2.8 2.8 3.8 0.6 8.2 3.6 5.4

9 140 23 26 502 294 21 48 1 2 41 1107 605 6.7

0.8 12.6 2.1 2.3 45.3 26.6 1.9 4.3 0.1 0.2 3.7 100.0

0.2 2.5 1.5 2.2 45.0 13.5 1.8 3.4 0.1 1.5 1.9 5.1

21 413 36 51 325 1009 52 95 5 5 146 2158 1149 12.7

1.0 19.1 1.7 2.4 15.1 46.8 2.4 4.4 0.2 0.2 6.8 100.0

0.5 7.4 2.4 4.3 29.1 46.3 4.6 6.8 0.4 3.7 6.9 9.9

7-Unpaid family workers 24 250 33 19 27 63 531 58 111 10 43 1169 638 7.1

2.1 21.4 2.8 1.6 2.3 5.4 45.4 5.0 9.5 0.9 3.7 100.0

0.6 4.5 2.2 1.6 2.4 2.9 46.7 4.1 8.7 7.5 2.0 5.4

8-Unemployed 109 429 46 74 32 124 66 347 31 24 113 1395 1048 11.6

7.8 30.8 3.3 5.3 2.3 8.9 4.7 24.9 2.2 1.7 8.1 100.0

2.7 7.6 3.1 6.2 2.9 5.7 5.8 24.7 2.4 17.9 5.4 6.4

9. Non-labor / studing 51 165 7 20 3 14 141 116 1006 10 58 1591 585 6.5

3.2 10.4 0.4 1.3 0.2 0.9 8.9 7.3 63.2 0.6 3.6 100.0

1.3 2.9 0.5 1.7 0.3 0.6 12.4 8.3 78.9 7.5 2.8 7.3

10. Non-labor force at domestic work 6 29 8 9 3 7 15 30 4 8 38 157 149 1.6

3.8 18.5 5.1 5.7 1.9 4.5 9.6 19.1 2.5 5.1 24.2 100.0

0.1 0.5 0.5 0.8 0.3 0.3 1.3 2.1 0.3 6.0 1.8 0.7

42 148 37 69 36 124 34 97 23 27 1375 2012 637 7.0

2.1 7.4 1.8 3.4 1.8 6.2 1.7 4.8 1.1 1.3 68.3 100.0

1.0 2.6 2.5 5.8 3.2 5.7 3.0 6.9 1.8 20.1 65.2 9.3

Total in T+1 4074 5610 1498 1192 1115 2177 1137 1406 1275 134 2109 21727 9047 100.0

18.8 25.8 6.9 5.5 5.1 10.0 5.2 6.5 5.9 0.6 9.7 100.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 41.6

Total in T+ 1 - Diagonal 725 2270 716 761 613 1168 606 1059 269 126 734 9047

% Mobility 8.0 25.1 7.9 8.4 6.8 12.9 6.7 11.7 3.0 1.4 8.1

5. Agricultural advantageous self-

employed

6. Agriculture unfavorable self-

employed

11-Not in the labor force-any other

state

agricultural prices and therefore of earnings from agricultural self-employment.

In contrast with women, men tend to remain longer as unemployed, 24.9 percent of

unemployed men in one year are still unemployed in the next year. Of the men who transition

out of unemployment, most go into informal salaried employment. Unlike for women, very few

men transition from unemployment into unpaid domestic work or into any other non-labor

market state.

5. Determinants of the transitions into and out of a successful state

According to the literature, variables that measure human capital, family characteristics and the

characteristics of the job are important determinants of labor market transitions.

Those with more human capital (i.e. education and experience) are more likely to be in the labor

force, and if they work are more likely to be full-time formal sector employees (Duryea, Marqéz,

Pagés and Sarpetta, 2006; Bosch and Maloney, 2010; Cunningham and Bustos-Salvagno, 2011).

Workers with more experience (especially in the formal sector) are more likely to be successful

entrepreneurs (Cunningham and Bustos-Salvagno, 2011). Some studies suggest that human

capital is a more important factor in explaining the success in the case of women entrepreneurs

compared to men (Bardasi, Sabarwal and Terrell, 2010). Age is also important in determining

whether or where a person works. For example, younger workers are more likely to be

informal sector workers while owners/employers are more likely to be older (Bosch and Maloney,

2010).

It has been argued that the reasons for becoming self-employed are different for men and

women; specifically it has been argued that women become self-employed because they seek

more flexible work schedules (Delman and Davidson, 2000). To examine this possibility, other

explanatory variables will include some that describe the structure of the family, including the

number of infant children and marital status.

It has been found that the selection of economic activity differs between men and women

entrepreneurs. Women entrepreneurs are predominantly concentrated in service activities,

while men tend to be owners of companies engaged in manufacturing and construction activities

(Bardasi, Sabarwal and Terrell, 2010). It has also been shown that in developing economies,

women are less likely to operate in high-technology activities (Anna, et al. 1999). To examine

the role of the selection of economic activity as a determinant of advancement, the regressions

will include economic activity dummies as explanatory variables.

Next, we will examine the determinants of the transitions into and out of each advantageous

state by estimating probit models.

5.1. Entering an advantageous state

Using a sample of workers in unfavorable states in time t, we estimate a probit equation of the form: T

Prob(INADVANTit = 1) = αo + X ′itβ + Σ γtYRt + µit , (3) t=1

In equation (3), INADVANTit equals one if the individual is in unfavorable state at time t but is

in an advantageous state at time t+1, and zero if the individual i is in an unfavorable state at

time t and stays in a unfavorable state at time t+1.

Xit is the explanatory variables vector which includes; individual specific human capital variables

(years of education and whether the individual had received job training), whether the individual

lives in a high population density area, the relationship to the household head, change in

the marital status, industry sector, household characteristics (number of young children, number

of school age children, number of working age household members, number of household

members older than 65 years of age), whether the individual has access to public services

(tube water and electricity).

Explanatory variables also include the amount of remittances and “non-labor income” (equal

to household income minus remittances and minus the labor earnings of individual i). We expect

that because of substitution of leisure for labor, a higher non-labor income will reduce the

probability that a worker is in the labor force. At the same time, a higher non-labor income

may make it more likely that a worker is a successful self-employed worker because non-labor

income may provide a source of financing for capital. We separate remittances from other non-

labor income to examine whether these two sources of non-labor income have different impacts

on the labor market choices of men and women. Finally, to control for year-specific factors

such as aggregate supply and aggregate demand changes or design changes in the household

surveys, we include a dummy variable for each year, YRt. From the estimated coefficients, βit,

we can calculate the marginal impact of each explanatory variable on the probability of a

transition from a not advantageous state to each advantageous labor market state.

Our estimates of these effects are reported on table 7; positive number indicates that an increase

in the corresponding explanatory variable increases the probability of transition from an

unfavorable state to each advantageous state indicated by the column of the table.

5.1.a. Characteristics of those who transition from unfavorable states into advantageous non-

agricultural self-employment

• Older: For both men and women, the probability that someone moves up from an

unfavorable state to advantageous non-agricultural self-employed increases

with age, but at a decreasing rate. The probability of transitioning into advantageous

non-agricultural self-employment increases until men and women are about 48 years

old. After 48 years old, the probability of this transition decreases with age.

• More Education: For both men and women, a secondary education increases the

probability that someone moves up from an unfavorable state to advantageous non-

agricultural self-employed. For women, a completed primary education also helps. On

the other hand, obtaining additional tertiary (university) education

does not increase the probability of moving up into advantageous non- agricultural

self-employment beyond the benefit of a secondary education.

• Household heads: are more likely to transition into advantageous non- agricultural

self-employment; this is true for both men and women. Being a spouse of the

household head also increases the likelihood of a transition into non-agricultural self-

employment. This is also true for both men and women.

• Access to electricity: access to electricity increases the probability that both men and

women transition from an unfavorable state to advantageous non- agricultural self-

employed. Access to water or living in a high-density population area are not significant

determinants of transitioning into advantageous non- agricultural self-employment.

• Sector of employment: For both men and women, those who transition into

advantageous non-agricultural self-employment work in the industry sectors of

manufacturing and construction, commerce, and high complexity services. The

magnitude of the impacts of sector of employment are similar for men and

women except in high complexity services. The results presented in table 8 imply

that advantageous self-employed men are more likely than women to come from

high complexity services. This is consistent with Anna, et al. (1999), who presents

evidence that in developing economies, entrepreneurial women are less likely to operate

in high-technology activities.

Remittances and non-labor income: Higher non-labor income and remittances

increases the probability of a transition into advantageous non-agricultural self-

employment (although only the coefficient on non-labor income is statistically

significant). These results suggests that, in El Salvador, access to non-labor income to

finance capital for self-employment makes it more likely that for someone will become

an advantageous self-employed/employer.

Table 7.

Marginal effects on the probability of entering an advantageous state, comparison by gender

Notes: Table reports marginal effects evaluated at the means of all variables, from estimates of β in Eq. (3) using

probit regressions f or samples identified by column. A positive coefficient means that an increase in the related

variable, increases the probability that a person enters a particular successful state.

n.i./Not included in this model; d.c/Dropped because of collinearity

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1

Log Likelihood -2405.811 -1546.937 -2095.508 -2304.597 -2152.891 -248.047Number of obs. 15,282 22,187 15,282 22,187 15,282 22,187Pseudo R2 0.122 0.210 0.202 0.222 0.113 0.134

AGE 0.00324 *** 0.00032 ** 0.00354 *** 0.00213 *** 0.00247 *** 0.00009 **-0.00045 -0.00016 -0.00035 -0.00023 -0.00038 -0.00004

AGE_squared -0.00004 *** -0.00001 *** -0.00004 *** -0.00002 *** -0.00003 *** 0.00000 *(5.17e-06) (1.87e-06) (3.80e-06) (2.54e-06) (3.82e-06) (4.07e-07)

Primary completed 0.01870 *** 0.00515 ** 0.00440 0.00608 *** -0.00708 *** -0.00049 **(0.00500) (0.00208) (0.00273) (0.00207) (0.00243) (0.000235)

Some secondary completed 0.03080 *** 0.00680 *** 0.00724 ** 0.00443 ** -0.01430 *** -0.00042 *(0.00520) (0.00208) (0.00292) (0.00203) (0.00250) (0.000249)

Secondary completed 0.07300 *** 0.02990 *** 0.02650 *** 0.00667 ** -0.01350 *** -0.000258(0.0103) (0.00570) (0.00585) (0.00263) (0.00284) (0.000263)

Higher level studies 0.09330 *** 0.03680 *** 0.02070 *** 0.01960 *** -0.01880 *** n.i.(0.0148) (0.00819) (0.00753) (0.00614) (0.00279)

Head of household 0.00194 0.000222 0.01770 *** 0.01620 *** 0.02520 *** 0.00132(0.00383) (0.00132) (0.00303) (0.00381) (0.00381) (0.00114)

Spouse 0.00410 -0.00207 ** 0.02750 ** 0.01090 *** 0.02210 * 0.000388(0.00862) (0.000995) (0.0114) (0.00229) (0.0124) (0.000502)

Tow ards married status 0.00387 0.000107 0.0148 0.000190 0.00218 n.i.(0.00813) (0.00269) (0.00966) (0.00461) (0.00999)

Tow ards single status -0.00483 0.00349 0.01890 * 0.00986 * -0.01430 *** 0.00471(0.00871) (0.00332) (0.0109) (0.00532) (0.00485) (0.00314)

High population density area 0.02280 *** 0.00833 *** 0.00143 0.000525 -0.02220 *** n.i.(0.00443) (0.00175) (0.00237) (0.00142) (0.00208)

Manufacturing & construction 0.00487 0.00662 ** 0.05150 *** 0.07770 *** n.i. n.i.(0.00377) (0.00314) (0.00630) (0.0102)

Commerce 0.01840 *** 0.00334 ** 0.06470 *** 0.08700 *** n.i. n.i.(0.00512) (0.00135) (0.00783) (0.00606)

Textile & clothing industry 0.0208 0.02060 *** 0.0328 0.03680 *** n.i. n.i.(0.0179) (0.00719) (0.0207) (0.0105)

High complexity services 0.03520 *** 0.01040 *** 0.04860 *** 0.01780 *** n.i. n.i.(0.00965) (0.00299) (0.0104) (0.00475)

Low complexity services 0.00477 *** 0.00276 *** 0.00577 *** 0.00330 *** n.i. n.i.(0.000818) (0.000471) (0.000636) (0.000809)

Age 6 or younger 0.00119 -0.000623 0.00137 0.00180 *** 0.00200 5.51e-05(0.00154) (0.000449) (0.00113) (0.000696) (0.00132) (0.000117)

Age 07_18 -0.000128 -0.00060 ** -0.00189 *** 0.000127 0.00262 *** 4.56e-05(0.000815) (0.000269) (0.000639) (0.000393) (0.000709) (5.96e-05)

Age 19_65 0.00302 *** 0.000220 -0.00164 * -0.00088 * 0.00114 -0.000125(0.00106) (0.000313) (0.000851) (0.000528) (0.000970) (9.53e-05)

Age 65+ -0.00165 -0.000744 -0.00679 *** 8.91e-05 0.00345 -0.000230(0.00252) (0.000729) (0.00216) (0.00125) (0.00224) (0.000237)

Tubed w ater inside dw elling 0.02050 *** 0.00245 0.00556 0.000888 -0.01250 ** -0.000324(0.00313) (0.00159) (0.00377) (0.00304) (0.00536) (0.000489)

Electricity -0.00394 -0.000621 0.00980 *** 0.00677 *** 0.00546 ** -0.000362(0.00332) (0.00112) (0.00182) (0.00120) (0.00233) (0.000334)

Water 0.00168 0.000754 0.00259 0.00142 -0.00504 * -8.77e-06(0.00280) (0.000854) (0.00200) (0.00131) (0.00266) (0.000202)

Remittances -0.02340 * 0.000619 0.00960 0.00591 0.04420 *** 0.000316(0.0125) (0.00198) (0.00895) (0.00390) (0.00878) (0.000577)

Non labor income -0.00593 * 0.000114 0.00951 *** 0.00343 *** -0.00122 6.74e-05(0.00343) (0.000716) (0.00205) (0.00121) (0.00406) (0.000301)

d09 -0.00405 6.21e-05 3.67e-05 0.00403 ** -0.000758 -8.55e-05(0.00312) (0.000972) (0.00255) (0.00198) (0.00330) (0.000270)

d10 -0.00222 2.51e-05 0.00280 0.00373 ** 0.00430 0.000516(0.00312) (0.000957) (0.00259) (0.00188) (0.00342) (0.000438)

d11 -0.00523 * 0.000454 0.00295 0.00847 *** 0.00341 9.42e-05(0.00300) (0.000967) (0.00260) (0.00216) (0.00341) (0.000301)

d12 d.c. d.c. d.c. d.c. d.c. d.c.

Men Women Men Women Men Women

VariableSalaried formal sector Advantageous non-agricultural SE Advantageous agricultural SE

• Dependent family members: Surprisingly, table 8 presents no evidence that an increase

in the number of dependent family members reduces the probability that women

become advantageous self-employed workers or employers. On the other hand, there is

some evidence that an increase in the number of children 7-18 years and the number

of household members 65 years and older does significantly reduce the probability

that men transition into advantageous self- employment. This pattern may exist

because women with dependent family members are attracted by the flexibility of self-

employment, which can allow women to provide child care and other unpaid domestic

care while still earning an income. Consistent with this, an increase in the number of

children between 7 and 19 years old reduces the probability that women transition into

formal sector salaried employment, where there is little flexibility in terms of hours

worked. On the other hand, men in El Salvador are generally not expected to provide

unpaid domestic care, but are expected to financially support the family.

5.1.b. Characteristics of those who transition from unfavorable states into salaried

formal employment:

• Older: For both men and women, age increases the probability of transitioning into

salaried formal employment (although at a decreasing rate

with age). The probability of transitioning into salaried formal employment

increases with age until the late 30s for men and the 20s for women. After that,

the probability of transitioning into salaried formal employment decreases with

age.

• More education: For both men and women, more education increases the

probability of transitioning into salaried formal employment. The probability of

transitioning into salaried formal employment increases for each additional

education level. The magnitude of the coefficients on the education variables are

all higher for men than for women, suggesting that education is a more important

determinant promoting the transition into salaried formal employment for men

than women. This may be because educated women suffer a “penalty” compared to

men when searching for formal sector employment.

• Fewer school age children in households: For women, a larger number of

household members between 7 and 19 years old reduces the probability of

transitioning into salaried formal employment. As noted, this may reflect the

inflexibility of hours worked in salaried formal employment. Aside from this one

variable, the number of dependents in the household does not affect the probability

of transitioning into salaried formal employment.

• Remittances and other non-labor income: For men, but not women, remittances

and other non-labor income decreases the probability of transitioning into formal

salaried employment.

• Sector of Employment: Both men and women who transition into salaried formal

employment are very likely to have worked in high complexity services (compared

to agriculture, manufacturing and construction, commerce and low complexity

services). Women, but not men, who transition into salaried formal employment

are also more likely to have come from textile and clothing manufacture (possibly

because the large export and maquila sector in El Salvador disproportionately

employs women).

5.1.c. Characteristics of those who transition from unsuccessful states into

advantageous agricultural self-employment:

• Older: For both men and women, the probability that someone moves up from an

unsuccessful state to successful agricultural self-employed increases with age, but at a

decreasing rate .Fewer school-age children: For men only, an increase in the number of

school- age children makes the transition into successful agricultural self-employment less

likely. The number of dependent household members has no significant impact on the

probability that women transition into this sector

• Receiving remittances: increases the probability that men transition into

advantageous agricultural self-employment.

• Unlike transitions into other advantageous states, education does not promote the

transition into advantageous agricultural self-employment.

5.2 Leaving an advantageous state

Using a sample of workers in advantageous states in time t, we estimate a probit equation

of the form:

T

Prob(OUTADVANTit = 1) = αo + X ′itβ + Σ γtYRt + µit , (4) t =1

In equation 4, OUTADVANTit equals one if the individual i is in an advantageous state at time t

but is not in an advantageous state at time t+1, and zero if the individual i is in an advantageous

state at time t and stays in advantageous state at time t+1. Xit is the explanatory variables

vector which includes the same variables as equation 1. Also, to control for year-specific

factors such as aggregate supply and aggregate demand changes or design changes in the

household surveys, we include a dummy variable for each year, YRt. From the estimated

coefficients, βit, we can calculate the marginal impact of each explanatory variable on the

probability of a transition from each advantageous state to a not advantageous labor state.

Our estimates of these effects are reported on table 8.

A positive number in table 8 indicates that an increase in the corresponding explanatory variable,

increases the probability of transition from the advantageous states indicated in each column of

the table to a not advantageous state

5.2.a. Characteristics of those who transition out of advantageous non-agricultural self-

employment into an unfavorable state:

• Younger workers: especially the youngest workers are more likely to fall out of

advantageous non-agricultural self-employment.

• Less Education: for both men and women, having at least a secondary complete

education reduces the probability of transitioning out of advantageous non-

agricultural self-employment.

• More school-age children: for men, but not for women, an increase in the number of

young children (0-6 years old) makes it more likely to transition out of advantageous

non-agricultural self-employment. Surprisingly, the number of children or elderly

members of the household does not affect whether or not a women transitions out of

advantageous non-agricultural self-employment.

• Secondary family workers: Being a household head reduces the probability that an

advantageous non-agricultural self-employed worker will move out of that state,

although this is statistically significant only for men.

• Non-labor income: For both men and women, more non-labor income decreases the

probability that a worker transitions out of advantageous non-agricultural self-

employment. This suggests that non-labor income can help to finance the capital

necessary to maintain advantageous non-agricultural self-employment.

5.2.b. Characteristics of those who transition out of salaried formal employment into an

unfavorable state:

• Younger workers: especially the youngest workers are more likely to fall out of salaried

formal employment.

• Less Education: for both men and women a tertiary education reduces the probability of

a transition out of salaried formal employment. For men, but not women, having a

secondary education also significantly reduces the probability of transitioning out of

salaried formal employment.

• Household Head: Being a household head reduces the probability that male salaried

formal employee will move out of that sector. This variable is not

significant for women.

• Remittances and other non-labor income: For men, but not women, remittances and

other non-labor income increases the probability of transitioning out of formal

salaried employment. This suggests that higher remittances induce men to substitute

leisure for salaried formal employment.

• Sector of Employment: For both men and women, workers in low complexity services

and the textile and clothing industries are less likely to leave salaried formal

employment than are workers in agriculture.

Table 8

Marginal effects on the probability of leaving an advantageous state, comparison by gender

Notes: Table reports marginal effects evaluated at the means of all variables, from estimates of β in Eq. (4) using

probit regressions f or samples identified by column. A positive coefficient means that an increase in the related

variable, increases the probability that a person enters a particular successful state.

n.i./Not included in this model; d.c/Dropped because of colinearity

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1

Log Likelihood -1548.915 -827.803 -845.788 -718.058 -708.093 -16.152Number of obs. 3,956 2,295 1,336 1,083 1,075 66Pseudo R2 0.087 0.105 0.067 0.041 0.047 0.457

AGE -0.01670 *** -0.02490 *** -0.03290 *** -0.02070 ** -0.01320 * -0.0113(0.00299) (0.00415) (0.00792) (0.00887) (0.00694) (0.0167)

AGE_squared 0.00020 *** 0.00027 *** 0.00038 *** 0.00020 ** 0.00012 * 0.000105(3.54e-05) (5.19e-05) (8.52e-05) (9.20e-05) (7.00e-05) (0.000159)

Primary completed 0.0261 0.0442 0.00181 -0.0387 -0.0793 0.0152(0.0219) (0.0408) (0.0418) (0.0453) (0.0506) (0.0275)

Some secondary completed -0.0222 -0.0153 -0.0444 -0.0674 -0.11000 * n.i.(0.0172) (0.0287) (0.0390) (0.0456) (0.0590)

Secondary completed -0.06430 *** -0.0380 -0.09120 ** -0.13000 ** -0.141 n.i.(0.0161) (0.0276) (0.0430) (0.0542) (0.0947)

Higher level studies -0.08120 *** -0.06690 ** -0.18700 *** -0.0896 -0.237 d.c.(0.0163) (0.0301) (0.0452) (0.0750) (0.158)

Head of household -0.11900 *** 0.0324 -0.21700 **** -0.14300 ** -0.16000 ** 0.0150(0.0239) (0.0279) (0.0673) (0.0687) (0.0721) (0.0833)

Spouse -0.0342 0.0280 -0.25300 *** -0.0265 -0.131 n.i.(0.0283) (0.0216) (0.0686) (0.0648) (0.128)

Tow ards married status -0.00114 0.0158 -0.0831 0.0246 -0.0825 n.i.(0.0409) (0.0443) (0.125) (0.135) (0.133)

Tow ards single status 0.11400 * 0.0681 0.0365 -0.22600 ** 0.124 n.i.(0.0619) (0.0557) (0.104) (0.0946) (0.159)

High population density area -0.02870 ** -0.03450 ** 0.0160 -0.00186 0.121 n.i.(0.0118) (0.0142) (0.0344) (0.0390) (0.106)

Manufacturing & construction -0.0234 -0.06890 ** -0.00530 -0.27900 *** n.i. n.i.(0.0282) (0.0351) (0.0660) (0.102)

Commerce -0.0185 -0.0486 -0.11300 * -0.26500 *** n.i. n.i.(0.0291) (0.0435) (0.0635) (0.0935)

Textile & clothing industry -0.07580 *** -0.09470 *** -0.0791 d.c. n.i. n.i.(0.0216) (0.0317) (0.119)

High complexity services -0.0391 0.0217 d.c. -0.191 n.i. n.i.(0.0270) (0.0648) (0.125)

Low complexity services -0.02170 *** -0.03080 *** -0.02370 * -0.05880 ** n.i. n.i.(0.00599) (0.0111) (0.0139) (0.0284)

Age 6 or younger -0.00261 -0.00257 -0.04570 ** -0.00336 -0.0139 -0.00778(0.00778) (0.00988) (0.0219) (0.0230) (0.0194) (0.0209)

Age 07_18 0.00828 * 0.00459 -0.00958 -0.0190 0.00647 -0.0228(0.00452) (0.00612) (0.0119) (0.0131) (0.00996) (0.0247)

Age 19_65 -0.01350 ** -0.00671 0.02710 * 0.0162 0.00709 0.0481(0.00592) (0.00698) (0.0156) (0.0166) (0.0147) (0.0454)

Age 65+ -0.00168 0.0239 -0.0127 -0.00566 0.0327 0.0387(0.0142) (0.0150) (0.0386) (0.0387) (0.0355) (0.0451)

Tubed w ater inside dw elling -0.18800 ** d.c. -0.0684 0.28400 * -0.09950 * -0.0255(0.0834) (0.142) (0.146) (0.0538) (0.0356)

Electricity -0.03990 ** -0.0237 -0.15600 *** -0.14300 ** -0.0419 0.359(0.0195) (0.0304) (0.0575) (0.0636) (0.0430) (0.285)

Water -0.00313 -0.00995 -0.09900 ** -0.0523 0.06240 * -0.0441(0.0163) (0.0247) (0.0421) (0.0463) (0.0354) (0.0492)

Remittances 0.0954 0.0125 -0.00372 -0.00821 0.0572 0.0463(0.0828) (0.0724) (0.190) (0.130) (0.134) (0.142)

Non labor income 0.03310 *** 0.02800 ** -0.05660 * -0.06380 * -0.20900 *** 0.0375(0.0122) (0.0123) (0.0310) (0.0342) (0.0457) (0.0487)

d09 -0.03740 ** 0.00500 -0.0249 0.0504 0.0220 n.i.(0.0146) (0.0198) (0.0418) (0.0484) (0.0529)

d10 -0.02870 ** -0.0295 -0.000813 0.0716 0.0234 -0.280(0.0145) (0.0183) (0.0419) (0.0458) (0.0527) (0.242)

d11 -0.04530 *** -0.0122 0.0404 0.09260 ** -0.0119 -0.57700 **(0.0141) (0.0188) (0.0415) (0.0462) (0.0517) (0.264)

d12 d.c. d.c. d.c. d.c. d.c. d.c.

Men

Advantageous agricultural SE

Women

Salaried formal sectorAdvantageous non-agricultural

SE

Men Women Men Women

Variable

6. Conclusions and Policy Implications

Our work sheds light on the answers to two key questions: (1) what are the

characteristics of the men and women who move up to an advantageous labor market state from

an unfavorable one? and (2) what are the characteristics of the men and women who fall out

of advantageous labor market states into unfavorable ones? The answers to these questions can

contribute to the appropriate design and targeting of public policies interventions to promote

success in the labor market. Our work also sheds light on whether the characteristics correlated

with success in the labor market differ between women and men, and therefore whether the

appropriate design and targeting of policies is different for men and women

6.1 Summary of results

In general, in El Salvador men are more mobile than are women, over 22% of men change

states from one year to the next, compared to only 12% of women. There is substantial mobility

of both men and women into and out of advantageous non- agricultural self-employment. The

fact that it is possible to transition from a non-advantageous labor market state into advantageous

non-agricultural self-employment shows that there is a role for policies that promote the ability

of workers in non-advantageous states to transition into advantageous non-agricultural self-

employment.

On the other hand, there is very little mobility out of or into salaried formal sector employment;

85% of women and 83% of men stay in this sector from one year to the next. However, there is an

important number of women who move from formal salaried to unpaid domestic work.

What did the advantageous non-agricultural self-employed do before they became successful? For

men, those who transition into advantageous non-agricultural self- employment tend to come

from unfavorable non-agricultural self-employment and from informal salaried employment. For

women, those who transition into advantageous non-agricultural self-employment tend to come

from unpaid domestic work as well as from unfavorable non-agricultural self-employment and

informal salaried employment. Advantageous non-agricultural self-employed do not generally

come from the salaried formal sector.

What did formal sector salaried employees do before they became successful? Of the small

number of people who transition into formal salaried employment, most come from informal

salaried employment. For women there is also a flow from unpaid domestic care.

For both men and women, the most important characteristic promoting transitions into the two

non-agricultural advantageous labor market states sector is education. Any additional

education will increase the probability of transitioning into the salaried formal sector, although

tertiary education is the most advantageous for the salaried formal sector while a secondary

education is most advantageous for advantageous non-agricultural self-employment. Our results

suggest that, compared to women, education has a bigger impact on the ability of men to transition

into advantageous labor market states. This suggests that educated women may face a “penalty”

in searching for advantageous employment.

Both men and women who transition into formal salaried employment are likely to have worked

in high complexity services. It is likely that education is necessary for employment in high

complexity services, and that this is another indication of the importance of formal education in

obtain advantageous employment. Women, but not men, who transition into formal salaried

employment are also likely to come from textile and clothing manufacture, possibly because

the large export sector in El Salvador disproportionately hires women. Those who transition into

advantageous non-agricultural self-employment are also likely to work in the industry sectors of

manufacturing and construction, commerce and high complexity services. This is true for both

women and men, although men are more likely than women to come from high complexity

services

Very young workers are not likely to transition into advantageous non-agricultural self-

employment. It is most likely that a worker transitions into advantageous self-employment in their

30s or 40s. This suggests that workers who transition into advantageous self-employment obtain

experience before becoming successful as self-employed.

Remittances and other non-labor income promote advantageous non- agricultural self-

employment. Receiving more remittances and other non-labor income increases the probability

of transitioning into advantageous non-agricultural self- employment and decreases the

probability of transitioning out of this advantageous state. This is true for both men and women.

For men, receiving more remittances and other non-labor income decreases the probability of

transitioning into formal sector employment. In general, our results suggest that receiving more

remittances decreases the probability of transitioning into formal sector employment but

increases the probability of transitioning into advantageous self-employment. For women, an

increase in remittances and other non-labor income also increases the probability of leaving the

labor force for unpaid domestic care or other types of non-labor activity.

Access to electricity and other public services increases the probability that both men and women

transition from an unfavorable state into advantageous non-agricultural self-employment.

Female spouses are less likely to be in salaried formal employees, while male spouses are more

likely to be in salaried formal employment. A higher number of dependent members in the

household (i.e. too young or too old to work) make it less likely that women (but not men) are

found in the salaried formal sector. A higher number of dependent children (7-19 years old)

also reduce the probability that women (but not men) transition into formal salaried

employment. The difference between men and women is likely a consequence of the

traditional expectation that wives will provide unpaid domestic care to children and other

dependents, while husbands are expected to have full-time jobs outside of the home.

Surprisingly, the number of dependent family members does not have a significant influence

on the probability that women will transition into advantageous non- agricultural self-

employment. This may be because women are attracted to self-employment for the flexibility it

offers.

6.2 Policy implications of the results

This paper informs the debate on the question: What public policies would best support the ability

of women and men to become successful formal sector employees or successful small-scale

entrepreneurs? Our results suggest that these policies include the following:

1. The most advantageous labor market state is formal sector salaried employment. Those in this

state are the highest paid, have pensions and have access to Social Security medical care. Most

people enter this sector soon after graduating from school and remain in this sector for

a long time; very few older workers transition from non-advantageous labor market states into

formal salaried employment. Our results suggest that skills (i.e. formal education and job

training) are the most important factor promoting salaried formal employment. A post-secondary

education is particularly useful for obtaining formal salaried employment. Tertiary education

substantially increases the probability of a transition into salaried formal employment and

significantly reduces the probability of transitioning out of salaried formal employment.

2. Education is also important in promoting success for both men and women in other labor

market states (at least those outside of agriculture). For advantageous non-agricultural self-

employment, a secondary education is most important; a secondary education (but not a tertiary

education) significantly increases the probability of a successful transition into advantageous

non- agricultural self-employment. There should be clear efforts to reduce the school drop-out

rates, especially among girls, and to promote high-school completion through alternative

programs for those already outside the school system.

The focus of any policy to increase education levels will be on the young. Our results suggest that

it is not likely that older people who are in informal salaried employment, self-employed or are

in unpaid family care will transition into formal salaried employment, even if they obtain more

education. This suggests that most progress towards expanding “advantageous” employment,

especially formal salaried employment, will be intergenerational. That is, those currently self-

employed will not become formal sector employees, but their children may. We know that

education is the key to obtaining formal sector employment. This is especially true for girls, who

need to obtain education before they have children or get married. Once they are married or

have children, it becomes very difficult to complete education or to obtain advantageous

employment.

Adults, especially the currently self-employed, do receive short-term benefits from educating

children. For example, educated children can provide financial support for self-employment or in

non-labor market states. This suggests that a policy of promoting the education of children can

also be thought of as a policy of promoting the welfare of their parents (who might be currently

self-employed), not only as a policy of promoting the welfare of the children. This suggests that

allocating public funds to educating children, and especially to making sure that girls stay in

school, may be the most effective use of public funds if the goal is to expand the number of adults

in “advantageous” labor market states. For example, a policy promoting the education of poor

children may be more effective than policies to “entrepreneurship” skills to adults. This also

suggests that it is counterproductive to promote “entrepreneurship” training to children in

primary or secondary schools, especially if these programs encourage children to leave school

early in order to start their own businesses (which are not likely to succeed). In order to expand

advantageous employment, it is more important to keep children in school (especially to obtain

some post-secondary education) than to train them to leave school to start their own businesses.

3. Education does not matter in agricultural self-employment. That is, it does not promote

advantageous agricultural self-employment. However, education children in rural areas will make

it more likely that these children find better employment as formal salaried employees or non-

agricultural self-employment.

4. There is substantial year-to-year movement up from less successful labor market states

into advantageous non-agricultural self-employment. The fact that it is possible to transition from

a non-advantageous labor market state into advantageous non-agricultural self-employment

suggests that there is a role for policies that promote the ability of workers in non-advantageous

states to transition into advantageous non-agricultural self- employment.

Below we discuss some policies suggested by our results:

6. Both men and women are more likely to transition into advantageous non-agricultural self-

employment when they are older and have more working experience. Our results suggest

that older workers are most likely transition into advantageous self-employment; young

workers are not likely to transition into advantageous self-employment. Our results suggest that

policies to promote advantageous self- employment should be targeted towards older workers

who already have some relevant work experience.

7. While a tertiary education is an important prerequisite for success as a salaried formal

employee and secondary education is an important prerequisite for advantageous non-

agricultural self-employed/employer, our results suggest that policy interventions targeted to

current students that are designed to promote advantageous self-employment are not likely to

be effective. Very few students move directly from school to self-employment, and younger

workers are not likely to be advantageous self- employed workers. O ur results suggest that

programs designed specifically to teach the skills needed to be successful entrepreneurs are

better targeted to older workers currently working as informal sector employees, non-

advantageous self-employed workers or as unpaid family workers.

7. For both men and women, policies to promote advantageous non-agricultural self-

employment/employer should target those currently working as informal sector employees

or unfavorable non-agricultural self-employed and not salaried formal sector employees.

8. For women, there is also scope for policy intervention to promote the transition from unpaid

domestic work (not in the labor force) into advantageous self-employment. This suggests the

policy interventions should target women who are currently not in the labor force who are

engaged in unpaid domestic work. For example, providing quality childcare services, and

expanding access to initial and preschool education.

9. Since only around 30 percent of workers in El Salvador have a formal salaried work, and that 85

percent of women and 83 percent of men stay in this state from one year to the next, this implies

that to allow more individuals to move to this advantageous state, new jobs need to be created.

Therefore, policies to promote higher economic growth rates need to be implemented; this is

particularly important for women.

10. Providing access to public services (electricity, and education) promotes the transition into

successful non-agricultural self-employment/employer and salaried formal employment. This is

true for both men and women.

11. Our results suggest that remittances and other non-labor income can be a source of

financing that promotes advantageous self- employment. Those with access to remittances and

other non-labor income have a higher probability of transitioning from unfavorable states into

advantageous self- employment. Similarly, access to remittances and other non-labor income

make it less likely that advantageous self-employed leave this advantageous labor market state

and fall into an unfavorable state. These results suggest that policies that increase access to

capital for those who do not have access to remittances or non-labor income could help to

promote advantageous self-employment.

References

Allen, I. Elaine, et al. Global Entrepreneurship Monitor: 2007 Report on Women and

Entrepreneurship, Babson Park, MA: The Center for Women’s Leadership at Babson

College, 2007.

Bardasi, Elena, Shwetlena Sabarwal and Katherine Terrell, 2010, “How do Female Entrepreneurs

Perform? Evidence from Three Developing Regions,” World Bank mimeo, December.

Bennett, John and Saul Estrin (2007) “Entrepreneurial Entry in Developing Economies: Modeling

Interactions Between the Formal and Informal Sector,” working paper, London

School of Economics

Bosch, Mariano and William Maloney, 2010, “Comparative Analysis of Labor Market Dynamics

Using Markov Processes: An Application to Informality,” Labour Economics, 17(4), 621-

631.

Cunningham, Wendy and Javier Bustos Salvagno, “Youth Employment Transitions in Latin

America,” World Bank Policy Research Working Paper 5521, January 2011.

Decreto No. 645, Ley de Igualdad, Equidad y Erradicación de la Discriminación contra las Mujeres.

Diario Oficial No. 70, El Salvador, 2011.

Delmar, F. and P. Davidson (2000), Where do they come from? Prevalence and characteristics

of nascent entrepreneurs, Entrepreneurship and regional development 12, 1-23.

de Mel, Suresh, David McKenzie and Christopher Woodruff (2010), “Who are the Microenterprise

Owners? Evidence from Sri Lanka on Tokman v. de Soto,” in International Differences in

Entrepreneurship, Lerner and Schoar, eds, University of Chicago Press.

De Soto, Hernan, 1989, The Other Path: The Economic Answer to Terrorism, Basic Books, New York.

Duryea, Suzanne, Gustavo Marquez, Carmen Pages and Stefano Scarpetta, 2006, “For Better or

Worse? Job and Earnings Mobility in Nine Middle- and Low-Income Countries,” in Susan M.

Collins and Carol Graham, Brookings Trade Forum 2006: Global Labor Markets?, Brookings

Institution Press, Washington, 1878-209.

Fields, Gary S., 1975, “Rural-Urban Migration, Urban Unemployment and Underemployment, and

Job Search Activities in LDC’s,” Journal of Development Economics, Vol. 2, pp. 165-88.

Ghai, Dharam. 2003. “Decent work: Concept and indicators,” International Labour Review 142

(2), 113-145.

Gobierno de El Salvador, Plan quinquenal de desarrollo 2010-2014, Segunda edición, El Salvador,

2011

Hundley, G. 2001. Why women earn less than men in self-employment. Journal of Labor Research

22, 817-829.

Instituto Salvadoreño para el Desarrollo de la Mujer (ISDEMU), 2011. “Política nacional de

las mujeres: actualizada: medidas al 2014 Instituto Salvadoreño para el Desarrollo de

la Mujer (ISDEMU), 1ª edición, San Salvador, El Salvador, 2011.

Klapper, Leora and Simon Parker, 2010, “Gender and the Business Environment for New Firm

Creation,” The World Bank Research Observer (advance access February 25, 2010 at

http://wbro.oxfordjournals.org/content/early/2010/02/25/wbro.lkp032.short).

Ley No. 648, Ley de Igualdad de Derechos y Oportunidades. Gaceta Diario Oficial No. 51, Nicaragua,

2008.

Lohmann, H., and S. Luber, 2004. “Trends in Self-Employment in Germany: Different Types,

Different Outcomes?” In W. Muller and R. Arum ed. The Reemergence of Self-

Employment. Princeton University Press: Princeton, New Jersey, 36-74.

OECD (2011) Women’s economic empowerment. Issues paper Paris: OECD-DAC GENDERNET

Pagés, C., Piras, C. “El Dividendo de Género. Cómo Capitalizar el Trabajo de las Mujeres”. Banco

Interamericano de Desarrollo, 2010.

Tokman, Victor, 2007, “Modernizing the Informal Sector,” UN/DESA Working Paper No. 42.

World Bank (2012) “The effect of Women´s Economic Power in Latin America and The Caribbean”,

Document of the World Bank, August 2012.

APPENDIX A. Tables

Table A1: Number of households and individuals matched according to the criteria

Table A2: Evolution of labor force in El Salvador

Table A3: Women and men occupied, by category

Table A4: Why not seeking a job?

Table A5: Women and men occupied by economic sector

Table A6: Women and men occupied, by education

Table A7: Women and men occupied, by economic sector and education level.

Table A8: Women and men active contributors to Salvadoran Social Security Institute

Table A9: Women and men by categories.

Table A1: Number of households and individuals matched according to the criteria

No. Household Panel 2008- 2009 2009- 2010 2010- 2011 2011 -2012

1 housing, sex, month, age (+1 / -1) and year of birth 2,324 2,715 3,178 3,124

2 housing, sex, month, and age (+1 / -1) 16 20 48 42

3 housing, sex, month and year of birth 19 46 69 50

4 housing, sex, age and year of birth 306 353 402 390

5 housing, sex, month 1,510 1,732 1,897 1,936

6 housing, sex, month, age (+1 / -1) and year of birth, last year

between the spouse and the head of the next year. 250 272 298 332

7 housing, sex, month, age (+1 / -1) and year of birth, including the

former head and spouse the following year. 3 1 1 2

TOTAL 4,428 5,139 5,893 5,876

No. Panel per individual 2008- 2009 2009- 2010 2010- 2011 2011 -2012

1 headed home identifier ( same family ), sex, month, age (+1 / -1)

and year of birth 10,042 11,880 13,444 13484

2 headed home identifier, sex, month, and age (+1 / -1) 54 74 142 138

3 headed home identifier, sex, month and year of birth 86 159 251 209

4 headed home identifier, sex, age and year of birth 990 1,116 1,335 1240

5 headed home identifier, sex, month 4,146 4611 5,085 5137

6 Heads that were matched in the above process 110 103 121 111

TOTAL 15,428 17,943 20,378 20,319

Sample of each year 68,457 83,194 85,159 85,291

Percentage Panel in the sample 22.54 21.57 23.93 23.82

Sign women each year 35,813 43,556 44,575 44,482

Panel of Women each year 8,006 9,400 10,554 10,720

Percentage women in the sample Panel 22.36 21.58 23.68 24.10

Table A2: Evolution of labor force in El Salvador

Definition 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Total population 6 ,046,257 6 ,154,079 6 ,272,353 6 ,428,672 6 ,510,348 6 ,639,010 6 ,756,786 6 ,864,080 6 ,980,279 5 ,744,575 6 ,122,413 6 ,150,953 6 ,181,405 6 ,213,730 6 ,249,262

Women 3 ,154,382 3 ,221,570 3 ,270,285 3 ,375,686 3 ,425,723 3 ,474,450 3 ,526,383 3 ,616,302 3 ,691,850 3 ,036,350 3 ,221,152 3 ,240,662 3 ,259,023 3 ,249,627 3 ,296,491

Men 2 ,891,875 2 ,932,509 3 ,002,068 3 ,052,986 3 ,084,625 3 ,164,560 3 ,230,403 3 ,247,778 3 ,288,429 2 ,708,225 2 ,901,261 2 ,910,291 2 ,922,382 2 ,964,103 2 ,952,771

Working age population>=16 3 ,664,767 3 ,811,586 3 ,903,781 4 ,056,521 4 ,123,180 4 ,171,480 4 ,299,726 4 ,363,743 4 ,449,431 3 ,738,671 3 ,980,187 4 ,065,439 4 ,127,560 4 ,212,174 4 ,308,637

PAW >=16, Woman 1 ,981,853 2 ,064,787 2 ,112,551 2 ,210,447 2 ,257,728 2 ,267,549 2 ,321,647 2 ,398,584 2 ,459,771 2 ,065,562 2 ,177,757 2 ,219,421 2 ,255,794 2 ,280,267 2 ,332,324

PAW>=16, Men 1 ,682,914 1 ,746,799 1 ,791,230 1 ,846,074 1 ,865,452 1 ,903,931 1 ,978,079 1 ,965,159 1 ,989,660 1 ,673,109 1 ,802,430 1 ,846,018 1 ,871,766 1 ,931,907 1 ,976,313

Inactive 1 ,381,839 1 ,465,309 1 ,501,971 1 ,563,171 1 ,649,246 1 ,572,980 1 ,677,050 1 ,700,923 1 ,703,088 1 ,417,725 1 ,484,279 1 ,513,772 1 ,547,276 1 ,571,041 1 ,583,883

Inactive>=16, Woman 1 ,076,722 1 ,125,501 1 ,152,857 1 ,206,136 1 ,240,998 1 ,207,116 1 ,272,562 1 ,300,406 1 ,304,241 1 ,099,926 1 ,148,123 1 ,163,254 1 ,189,633 1 ,207,809 1 ,215,389

Inactive>=16, Men 305,117 339,808 349,114 357,035 408,248 365,864 404,488 400,517 398,847 317,799 336,156 350,518 357,643 363,232 368,494

Economically active population 2 ,282,928 2 ,346,277 2 ,401,810 2 ,493,350 2 ,473,934 2 ,598,500 2 ,622,676 2 ,662,820 2 ,746,343 2 ,320,946 2 ,495,908 2 ,551,667 2 ,580,284 2 ,641,133 2 ,724,754

EAP>=16, Women ,905,131 ,939,286 ,959,694 1 ,004,311 1 ,016,730 1 ,060,433 1 ,049,085 1 ,098,178 1 ,155,530 ,965,636 1 ,029,634 1 ,056,167 1 ,066,161 1 ,072,458 1 ,116,935

EAP>=16, Men 1 ,377,797 1 ,406,991 1 ,442,116 1 ,489,039 1 ,457,204 1 ,538,067 1 ,573,591 1 ,564,642 1 ,590,813 1 ,355,310 1 ,466,274 1 ,495,500 1 ,514,123 1 ,568,675 1 ,607,819

Occupied 2 ,116,763 2 ,188,861 2 ,239,199 2 ,319,065 2 ,320,957 2 ,418,050 2 ,445,577 2 ,472,958 2 ,568,742 2 ,173,963 2 ,349,050 2 ,364,579 2 ,398,478 2 ,466,375 2 ,559,315

Occupied>=16, Women ,850,935 ,896,147 ,924,168 ,951,380 ,981,125 1 ,022,802 1 ,008,393 1 ,046,504 1 ,110,878 ,929,438 ,992,232 1 ,004,061 1 ,011,696 1 ,025,678 1 ,068,524

Occupied>=16, Men 1 ,265,828 1 ,292,714 1 ,315,031 1 ,367,685 1 ,339,832 1 ,395,248 1 ,437,184 1 ,426,454 1 ,457,864 1 ,244,525 1 ,356,818 1 ,360,518 1 ,386,782 1 ,440,697 1 ,490,791

Unemployment 166,165 157,416 162,611 174,285 152,977 180,450 177,099 189,862 177,601 146,983 146,858 187,088 181,806 174,758 165,439

Unemployment>=16, Woman 54,196 43,139 35,526 52,931 35,605 37,631 40,692 51,674 44,652 36,198 37,402 52,106 54,465 46,780 48,411

Unemployment>=16, Men 111,969 114,277 127,085 121,354 117,372 142,819 136,407 138,188 132,949 110,785 109,456 134,982 127,341 127,978 117,028

Real rate of economic growth 3.7 3.4 2.2 1.7 2.3 2.3 1.9 3.6 3.9 3.8 1.3 -3.1 1.4 2.2 1.9

EAP/WAP for women 45.7 45.5 45.4 45.4 45.0 46.8 45.2 45.8 47.0 46.7 47.3 47.6 47.3 47.0 47.9

EAP/WAP for men 81.9 80.5 80.5 80.7 78.1 80.8 79.6 79.6 80.0 81.0 81.3 81.0 80.9 81.2 81.4

Table A3 Women and men occupied, by category

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years

Women 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Patron 26,188 28,333 29,473 27,217 33,399 144610 2.6% 2.8% 2.9% 2.7% 3.1% 2.8%

Self- employ with own local 57,103 55,727 57,450 52,318 52,871 275469 5.8% 5.6% 5.7% 5.1% 4.9% 5.4%

Self- employ with out own local 316,864 333,330 330,929 332,947 344,146 1658216 31.9% 33.2% 32.7% 32.5% 32.2% 32.5%

Cooperative 113 0 0 0 0 113 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Relative unpaid 74,124 72,021 76,853 83,023 85,373 391394 7.5% 7.2% 7.6% 8.1% 8.0% 7.7%

Permanent wage earners 360,922 347,917 344,717 352,011 370,587 1776154 36.4% 34.7% 34.1% 34.3% 34.7% 34.8%

Temporal wage earners 57,986 64,064 80,072 83,580 78,886 364588 5.8% 6.4% 7.9% 8.1% 7.4% 7.1%

Apprentice 248 192 0 65 69 574 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Domestic work 97966.0 102069.0 92136.0 94202.0 102197.0 488570 9.9% 10.2% 9.1% 9.2% 9.6% 9.6%

Others 718 408 66 315 996 2503 0.1% 0.0% 0.0% 0.0% 0.1% 0.0%

TOTAL 992,232 1004,061 1011,696 1025,678 1068,524 5102191 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Men 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Patron 76,476 73,816 69,112 64,912 74,807 359123 5.6% 5.4% 5.0% 4.5% 5.0% 5.1%

Self- employ with own local 24,364 26,865 26,994 28,775 26,549 133547 1.8% 2.0% 1.9% 2.0% 1.8% 1.9%

Self- employ with out own local 293,000 320,151 320,610 337,162 338,751 1609674 21.6% 23.5% 23.1% 23.4% 22.7% 22.9%

Cooperative 264 257 53 0 0 574 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Relative unpaid 83,957 97,150 90,672 101,300 112,740 485,819 6.2% 7.1% 6.5% 7.0% 7.6% 6.9%

Permanent wage earners 564,861 529,472 540,952 565,463 600,124 2800872 41.6% 38.9% 39.0% 39.2% 40.3% 39.8%

Temporal wage earners 292,954 298,674 328,093 331,170 323,346 1574237 21.6% 22.0% 23.7% 23.0% 21.7% 22.4%

Apprentice 4,194 3,813 3,812 3,621 4,689 20,129 0.3% 0.3% 0.3% 0.3% 0.3% 0.3%

Domestic work 10,385 9,876 6,346 8,106 9,583 44,296 0.8% 0.7% 0.5% 0.6% 0.6% 0.6%

Others 6,363 444 138 188 202 7,335 0.5% 0.0% 0.0% 0.0% 0.0% 0.1%

TOTAL 1356818 1360518 1386782 1440697 1490791 7035606 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table A4 Why not seeking a job?

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

Women inactive 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Lack of knowledge 4,983 5,011 6,383 6,242 8,118 30,737 0.4% 0.4% 0.5% 0.5% 0.7% 0.5%

Wait harvest 4,035 3,736 2,583 2,811 2,087 15,252 0.3% 0.3% 0.2% 0.2% 0.2% 0.3%

Study 165,431 176,883 182,157 188,503 182,033 895,007 14.2% 15.0% 15.2% 15.5% 14.8% 14.9%

Retired 27,475 35,618 37,557 34,226 35,513 170,389 2.4% 3.0% 3.1% 2.8% 2.9% 2.8%

Domestic work or family obligations 811,480 813,573 827,649 846,885 847,749 4147,336 69.6% 69.0% 68.9% 69.5% 69.0% 69.2%

Disability or illness 139,003 130,608 135,314 131,052 143,432 679,409 11.9% 11.1% 11.3% 10.7% 11.7% 11.3%

Others 13,530 13,626 9,396 9,573 9,339 55,464 1.2% 1.2% 0.8% 0.8% 0.8% 0.9%

Total 1165,937 1179,055 1201,039 1219,292 1228,271 5993,594 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Men inactive 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Lack of knowledge 20,412 28,808 21,706 24,847 27,084 122,857 5.1% 6.8% 5.3% 5.9% 6.4% 5.9%

Wait harvest 25,387 29,064 23,119 25,156 19,995 122,721 6.3% 6.9% 5.7% 6.0% 4.8% 5.9%

Study 145,363 151,916 163,621 174,224 174,435 809,559 36.2% 35.9% 40.1% 41.5% 41.4% 39.1%

Retired 45,558 49,051 45,119 44,354 44,176 228,258 11.3% 11.6% 11.1% 10.6% 10.5% 11.0%

Domestic work or family obligations 15,291 17,816 20,599 17,510 13,122 84,338 3.8% 4.2% 5.1% 4.2% 3.1% 4.1%

Disability or illness 109,574 105,922 109,323 104,415 114,392 543,626 27.3% 25.0% 26.8% 24.9% 27.2% 26.2%

Others 40,166 40,606 24,283 29,001 27,697 161,753 10.0% 9.6% 6.0% 6.9% 6.6% 7.8%

Total 401,751 423,183 407,770 419,507 420,901 2073,112 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table A5 Women and men occupied by economic sector

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

Women 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012

Agriculture and mining 47,736 48,049 54,503 67,590 53,915 4.8% 4.8% 5.4% 6.6% 5.0%

Manufacturing/Construction 116,324 106,693 110,828 114,153 118,988 11.7% 10.6% 11.0% 11.1% 11.1%

Textile manufacturing 86,859 76,529 74,083 74,418 75,280 8.8% 7.6% 7.3% 7.3% 7.0%

Commerce 406,000 425,711 424,542 416,680 442,608 40.9% 42.4% 42.0% 40.6% 41.4%

High complexity services 167,583 155,835 165,397 164,504 174,918 16.9% 15.5% 16.3% 16.0% 16.4%

Low complexity services 167,730 191,244 182,343 188,333 202,815 16.9% 19.0% 18.0% 18.4% 19.0%

TOTAL 992,232 1 004,061 1 011,696 1 025,678 1 068,524 100.0% 100.0% 100.0% 100.0% 100.0%

Men

Agriculture and mining 394,497 447,313 445,177 465,913 484,501 29.1% 32.9% 32.1% 32.3% 32.5%

Manufacturing/Construction 291,129 260,304 277,634 279,505 292,888 21.5% 19.1% 20.0% 19.4% 19.6%

Textile manufacturing 41,827 34,658 37,204 38,100 37,128 3.1% 2.5% 2.7% 2.6% 2.5%

Commerce 271,590 253,958 266,671 271,778 279,121 20.0% 18.7% 19.2% 18.9% 18.7%

High complexity services 306,394 258,911 256,873 274,427 279,461 22.6% 19.0% 18.5% 19.0% 18.7%

Low complexity services 51,381 105,374 103,223 110,974 117,692 3.8% 7.7% 7.4% 7.7% 7.9%

TOTAL 1 356,818 1 360,518 1 386,782 1 440,697 1 490,791 100.0% 100.0% 100.0% 100.0% 100.0%

Table A6 Women and men occupied by education

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

Woman 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

No Education 131,581 123,465 125,437 124,114 126,785 631,382 13.3% 12.3% 12.4% 12.1% 11.9% 12.4%

Incomplete primary 235,805 233,198 227,230 234,377 234,937 1165,547 23.8% 23.2% 22.5% 22.9% 22.0% 22.8%

Complete primary 156,792 165,369 166,964 174,442 180,657 844,224 15.8% 16.5% 16.5% 17.0% 16.9% 16.5%

Incomplete secondary 138,434 139,083 135,258 140,404 143,458 696,637 14.0% 13.9% 13.4% 13.7% 13.4% 13.7%

Complete secondary 172,766 180,642 190,856 200,825 214,784 959,873 17.4% 18.0% 18.9% 19.6% 20.1% 18.8%

Somekind higher level 156,854 162,304 165,951 151,516 167,903 804,528 15.8% 16.2% 16.4% 14.8% 15.7% 15.8%

TOTAL 992,232 1004,061 1011,696 1025,678 1068,524 5102,191 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Men 2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

No Education 173,491 178,638 173,941 172,392 170,496 868,958 12.8% 13.1% 12.5% 12.0% 11.4% 12.4%

Incomplete primary 324,559 332,301 330,524 334,684 328,749 1650,817 23.9% 24.4% 23.8% 23.2% 22.1% 23.5%

Complete primary 202,460 205,718 206,653 221,551 226,205 1062,587 14.9% 15.1% 14.9% 15.4% 15.2% 15.1%

Incomplete secondary 271,093 267,434 286,947 297,544 311,952 1434,970 20.0% 19.7% 20.7% 20.7% 20.9% 20.4%

Complete secondary 218,418 220,772 237,381 254,168 275,778 1206,517 16.1% 16.2% 17.1% 17.6% 18.5% 17.1%

Some kind of higher level 166,797 155,655 151,336 160,358 177,611 811,757 12.3% 11.4% 10.9% 11.1% 11.9% 11.5%

TOTAL 1356,818 1360,518 1386,782 1440,697 1490,791 7035,606 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table A7 Women and men occupied by economic sector and education level.

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

Women No

Education

Incomplete

primary

Complete

primary

Incomplete

secondary

Complete

secondary

Somekind

higher level

Total No

Education

Incomplete

primary

Complete

primary

Incomplete

secondary

Complete

secondary

Somekind

higher level

Total

Agriculture 79,799 97,268 29,081 40,461 11,576 2,925 261,110 12.6% 8.3% 3.4% 5.8% 1.2% 0.4% 5.1%

Fishing 5,634 3,227 674 479 199 0 10,213 0.9% 0.3% 0.1% 0.1% 0.0% 0.0% 0.2%

Mining 152 148 0 0 170 0 470 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Industry 91,621 129,335 99,995 78,488 111,595 49,082 560,116 14.5% 11.1% 11.8% 11.3% 11.6% 6.1% 11.0%

Clothing Industry 16,401 78,117 108,516 79,462 81,706 12,627 376,829 2.6% 6.7% 12.9% 11.4% 8.5% 1.6% 7.4%

Electricity Gas and Water 69 223 381 390 2,483 3,106 6,652 0.0% 0.0% 0.0% 0.1% 0.3% 0.4% 0.1%

Construction 444 1,094 2,263 1,962 2,818 8,629 17,210 0.1% 0.1% 0.3% 0.3% 0.3% 1.1% 0.3%

Commerce 278,998 534,803 395,543 320,631 432,516 153,707 2 116,198 44.2% 45.9% 46.9% 46.0% 45.1% 19.1% 41.5%

Services 158,264 321,332 207,771 174,764 316,810 574,452 1 753,393 25.1% 27.6% 24.6% 25.1% 33.0% 71.4% 34.4%

TOTAL 631,382 1165,547 844,224 696,637 959,873 804,528 5102,191 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Men

Agriculture 585,429 794,244 297,009 333,721 112,618 22,821 2 145,842 67.4% 48.1% 20.7% 31.4% 9.3% 2.8% 30.5%

Fishing 21,668 33,317 13,024 12,916 2,196 293 83,414 2.5% 2.0% 0.9% 1.2% 0.2% 0.0% 1.2%

Mining 2,291 2,638 1,385 1,053 514 264 8,145 0.3% 0.2% 0.1% 0.1% 0.0% 0.0% 0.1%

Industry 35,522 133,882 219,326 123,753 187,660 83,628 783,771 4.1% 8.1% 15.3% 11.6% 15.6% 10.3% 11.1%

Clothing Industry 3,097 20,239 58,172 25,941 57,031 17,571 182,051 0.4% 1.2% 4.1% 2.4% 4.7% 2.2% 2.6%

Electricity Gas and Water 2,341 8,613 6,911 3,785 13,338 13,066 48,054 0.3% 0.5% 0.5% 0.4% 1.1% 1.6% 0.7%

Construction 59,826 184,712 142,076 134,846 76,523 26,572 624,555 6.9% 11.2% 9.9% 12.7% 6.3% 3.3% 8.9%

Commerce 82,739 230,599 335,442 198,889 348,675 156,212 1 352,556 9.5% 14.0% 23.4% 18.7% 28.9% 19.2% 19.2%

Services 76,045 242,573 361,625 227,683 407,962 491,330 1 807,218 8.8% 14.7% 25.2% 21.4% 33.8% 60.5% 25.7%

TOTAL 868,958 1650,817 1434,970 1062,587 1206,517 811,757 7035,606 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table A8 Women and men active contributors to Salvadorean Social Security Institute

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Women 295,569 285,399 291,325 298,349 308,109 1478,751 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Agriculture and fishing 3,743 3,164 3,143 3,312 2,856 16,218 1.3% 1.1% 1.1% 1.1% 0.9% 1.1%

Mining 85 111 93 104 79 472 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Industry 83,482 72,159 73,571 76,151 74,984 380,347 28.2% 25.3% 25.3% 25.5% 24.3% 25.7%

Electricity Gas and Water 506 566 622 647 684 3,025 0.2% 0.2% 0.2% 0.2% 0.2% 0.2%

Construction 3,027 2,483 2,611 2,653 2,699 13,473 1.0% 0.9% 0.9% 0.9% 0.9% 0.9%

Comerce, restaurants 61,112 56,339 55,448 57,988 59,599 290,486 20.7% 19.7% 19.0% 19.4% 19.3% 19.6%

Transportation, warehouse and communication 5,982 7,125 8,186 8,850 9,920 40,063 2.0% 2.5% 2.8% 3.0% 3.2% 2.7%

Financial services and other high complexity services 38,352 41,650 43,730 44,524 44,825 213,081 13.0% 14.6% 15.0% 14.9% 14.5% 14.4%

Other services 38,858 44,238 42,337 39,282 41,187 205,902 13.1% 15.5% 14.5% 13.2% 13.4% 13.9%

Municipal and state 60,422 57,564 61,584 64,838 71,276 315,684 20.4% 20.2% 21.1% 21.7% 23.1% 21.3%

Men 409,146 396,051 404,424 420,370 425,829 2055,820 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Agriculture and fishing 9,510 9,151 9,476 9,985 10,332 48,454 2.3% 2.3% 2.3% 2.4% 2.4% 2.4%

Mining 614 501 506 563 583 2,767 0.2% 0.1% 0.1% 0.1% 0.1% 0.1%

Industry 81,645 78,927 82,701 85,601 86,206 415,080 20.0% 19.9% 20.4% 20.4% 20.2% 20.2%

Electricity Gas and Water 3,361 3,430 4,042 4,199 4,198 19,230 0.8% 0.9% 1.0% 1.0% 1.0% 0.9%

Construction 28,221 20,663 21,052 21,387 21,790 113,113 6.9% 5.2% 5.2% 5.1% 5.1% 5.5%

Comerce, restaurants 80,158 79,483 77,149 80,683 83,284 400,757 19.6% 20.1% 19.1% 19.2% 19.6% 19.5%

Transportation, warehouse and communication 22,120 24,019 26,258 28,391 29,951 130,739 5.4% 6.1% 6.5% 6.8% 7.0% 6.4%

Financial services and other high complexity services 80,638 73,880 74,395 75,747 79,435 384,095 19.7% 18.7% 18.4% 18.0% 18.7% 18.7%

Other services 36,829 33,524 33,512 31,094 34,931 169,890 9.0% 8.5% 8.3% 7.4% 8.2% 8.3%

Municipal and state 66,050 72,473 75,333 82,720 75,119 371,695 16.1% 18.3% 18.6% 19.7% 17.6% 18.1%

Table A9 Women and men by categories

Source: Data base of Multipurpose Household Survey (EHPM), Publication of Multipurpose Household Survey (EHPM), DIGESTYC, several years.

2008 2009 2010 2011 2012 Total 2008 2009 2010 2011 2012 Total

Women

1. Formal salaried 290,278 269,425 271,250 272,174 282,233 1385,360 13.3% 12.1% 12.0% 11.9% 12.1% 12.3%

2. Informal salaried 227,562 245,225 245,741 257,999 270,502 1247,029 10.4% 11.0% 10.9% 11.3% 11.6% 11.1%

3. Advantageous self-employed not agriculture 105,127 107,543 113,251 106,759 114,840 547,520 4.8% 4.8% 5.0% 4.7% 4.9% 4.9%

4. Unfavorable self-employed not agriculture 281,961 295,314 287,748 286,216 301,142 1452,381 12.9% 13.3% 12.8% 12.6% 12.9% 12.9%

5. Agricultural advantageous self-employed 3,928 3,712 4,143 4,168 3,276 19,227 0.2% 0.2% 0.2% 0.2% 0.1% 0.2%

6. Agriculture unfavorable self-employed 9,252 10,821 12,710 15,339 11,158 59,280 0.4% 0.5% 0.6% 0.7% 0.5% 0.5%

7. Un-paid family work 74,124 72,021 76,853 83,023 85,373 391,394 3.4% 3.2% 3.4% 3.6% 3.7% 3.5%

8. Unemployed 37,402 52,106 54,465 46,780 48,411 239,164 1.7% 2.3% 2.4% 2.1% 2.1% 2.1%

9. Non-labor / studing 176,114 179,086 187,198 192,982 186,286 921,666 8.1% 8.1% 8.3% 8.5% 8.0% 8.2%

10. Non-labor force at domestic work 797,793 807,849 819,781 838,382 839,453 4103,258 36.6% 36.4% 36.3% 36.8% 36.0% 36.4%

11. Non-labor force / others 174,216 176,319 182,654 176,445 189,650 899,284 8.0% 7.9% 8.1% 7.7% 8.1% 8.0%

Total 2177,757 2219,421 2255,794 2280,267 2332,324 11265,563 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Men

1. Formal salaried 414,300 382,571 382,013 398,791 410,613 1988,288 23.0% 20.7% 20.4% 20.6% 20.8% 21.1%

2. Informal salaried 464,457 459,708 497,328 509,757 527,331 2458,581 25.8% 24.9% 26.6% 26.4% 26.7% 26.1%

3. Advantageous self-employed not agriculture 127,230 125,137 129,509 137,774 135,069 654,719 7.1% 6.8% 6.9% 7.1% 6.8% 6.9%

4. Unfavorable self-employed not agriculture 98,099 105,185 110,069 106,148 103,293 522,794 5.4% 5.7% 5.9% 5.5% 5.2% 5.5%

5. Agricultural advantageous self-employed 51,838 58,512 54,095 63,005 66,576 294,026 2.9% 3.2% 2.9% 3.3% 3.4% 3.1%

6. Agriculture unfavorable self-employed 116,937 132,255 123,096 123,922 135,169 631,379 6.5% 7.2% 6.6% 6.4% 6.8% 6.7%

7. Un-paid family work 83,957 97,150 90,672 101,300 112,740 485,819 4.7% 5.3% 4.8% 5.2% 5.7% 5.2%

8. Unemployed 109,456 134,982 127,341 127,978 117,028 616,785 6.1% 7.3% 6.8% 6.6% 5.9% 6.5%

9. Non-labor / studing 141,326 149,759 160,457 170,613 172,570 794,725 7.8% 8.1% 8.6% 8.8% 8.7% 8.4%

10. Non-labor force at domestic work 15,143 17,468 20,221 17,262 12,938 83,032 0.8% 0.9% 1.1% 0.9% 0.7% 0.9%

11. Non-labor force / others 179,687 183,291 176,965 175,357 182,986 898,286 10.0% 9.9% 9.5% 9.1% 9.3% 9.5%

Total 1802,430 1846,018 1871,766 1931,907 1976,313 9428,434 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table A3: Women and men occupied, by category

Table A4: Why not seeking a job?

Table A5: Women and men occupied by economic sector

Table A6: Women and men occupied, by education

Table A7: Women and men occupied, by economic sector and education level.

Table A8: Women and men active contributors to Salvadoran Social Security Institute

Table A9: Women and men by categories.

APPENDIX B- : ALTERNATIVE DEFINITION OF ADVANTAGEOUS SELF-EMPLOYMENT

In the body of the paper, the category "Advantageous Self-employed/Employer" consists of

self-employed workers whose labor earnings are greater than the legal minimum wage. In this

appendix we consider a different definition of "Advantageous Self-employed/Employer,"

consisting of self-employed workers who live in families where the per capita household income

is greater than the poverty line. We use the alternative to make comparisons with the results

presented in a companion paper using Nicaraguan data that uses this alternative definition.

Tables B1 and B2 present transition matrices between the 11 states using the alternative

definition of “advantageous self-employed/employer.” Tables B3 and B4 present the probit

estimates of the characteristics associated with transitions into and out of each advantageous

labor market state using the alternative definition of advantageous.

In general, the coefficients of the probit equations for the characteristics associated with

transitions into and out of advantageous labor market states using the alternative definition of

“advantageous” have the same signs and significance levels as in the body of the paper. This

suggests that our results are robust to these two different definitions of “advantageous.” This

also suggests that the results presented in the body of this paper can be compared to the

results in the companion paper using Nicaraguan data.

Table B1. Probabilities of transition matrix for women (Sum of transitions for the period 2009-

2012)

Source: For each stata, the first row show the number of observation in each transition, the second row

should the probabilities that in status j at time t+k conditional on being in status Z at time t, the third row

shows the probabilities that an individual in status j at time t+k was in status z a time t.

Status 3 and 5, above poverty line.

1 2 3 4 5 6 7 8 9 10 11 Total in tTotal in

t - Diagonal

%

Mobility

1-Formal salaried employment 1976 112 49 8 0 0 12 34 11 93 18 2313 337 3.8

85.4 4.8 2.1 0.3 0.0 0.0 0.5 1.5 0.5 4.0 0.8 100.0

83.1 4.6 4.1 0.2 0.0 0.0 1.3 6.3 0.7 0.9 0.9 9.0

2-Informal salaried employment 166 1192 49 243 2 10 40 75 40 434 41 2292 1100 12.3

7.2 52.0 2.1 10.6 0.1 0.4 1.7 3.3 1.7 18.9 1.8 100.0

7.0 48.6 4.1 6.5 3.9 3.8 4.5 13.8 2.6 4.1 2.1 9.0

3. Advantageous self-employed not

agriculture 38 261 2710 213 12 14 105 25 11 662 73 4124 1414 15.8

0.9 6.3 65.7 5.2 0.3 0.3 2.5 0.6 0.3 16.1 1.8 100.0

1.6 10.6 228.3 5.7 23.5 5.4 11.8 4.6 0.7 6.3 3.8 16.1

4. Unfavorable self-employed not

agriculture 4 55 214 115 2 2 17 5 7 233 30 684 569 6.4

0.6 8.0 31.3 16.8 0.3 0.3 2.5 0.7 1.0 34.1 4.4 100.0

0.2 2.2 18.0 3.1 3.9 0.8 1.9 0.9 0.5 2.2 1.6 2.7

0 4 19 6 48 15 5 2 0 65 6 170 122 1.4

0.0 2.4 11.2 3.5 28.2 8.8 2.9 1.2 0.0 38.2 3.5 100.0

0.0 0.2 1.6 0.2 94.1 5.8 0.6 0.4 0.0 0.6 0.3 0.7

0 7 15 1 15 22 3 3 1 73 9 149 127 1.4

0.0 4.7 10.1 0.7 10.1 14.8 2.0 2.0 0.7 49.0 6.0 100.0

0.0 0.3 1.3 0.0 29.4 8.5 0.3 0.6 0.1 0.7 0.5 0.6

7-Unpaid family workers 10 58 48 85 3 3 322 14 78 253 17 891 569 6.4

1.1 6.5 5.4 9.5 0.3 0.3 36.1 1.6 8.8 28.4 1.9 100.0

0.4 2.4 4.0 2.3 5.9 1.2 36.2 2.6 5.0 2.4 0.9 3.5

8-Unemployed 54 90 9 35 1 2 20 76 29 151 10 477 401 4.5

11.3 18.9 1.9 7.3 0.2 0.4 4.2 15.9 6.1 31.7 2.1 100.0

2.3 3.7 0.8 0.9 2.0 0.8 2.2 14.0 1.9 1.4 0.5 1.9

9. Non-labor / studing 47 124 4 35 0 3 92 107 1263 236 33 1944 681 7.6

2.4 6.4 0.2 1.8 0.0 0.2 4.7 5.5 65.0 12.1 1.7 100.0

2.0 5.1 0.3 0.9 0.0 1.2 10.3 19.7 81.3 2.2 1.7 7.6

10. Non-labor force at domestic work 75 503 89 913 20 124 259 191 95 7847 607 10723 2876 32.2

0.7 4.7 0.8 8.5 0.2 1.2 2.4 1.8 0.9 73.2 5.7 100.0

3.2 20.5 7.5 24.3 39.2 47.7 29.1 35.1 6.1 74.1 31.7 41.9

9 46 8 72 2 11 15 12 19 544 1072 1810 738 8.3

0.5 2.5 0.4 4.0 0.1 0.6 0.8 0.7 1.0 30.1 59.2 100.0

0.4 1.9 0.7 1.9 3.9 4.2 1.7 2.2 1.2 5.1 55.9 7.1

Total in T+1 2379 2452 1187 3753 51 260 890 544 1554 10591 1916 25577 8934 100.0

9.3 9.6 4.6 14.7 0.2 1.0 3.5 2.1 6.1 41.4 7.5 100.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 34.9

Total in T+ 1 - Diagonal 403 1260 -1523 3638 3 238 568 468 291 2744 844 8934

% Mobility 4.5 14.1 -17.0 40.7 0.0 2.7 6.4 5.2 3.3 30.7 9.4

5. Agricultural advantageous self-

employed

6. Agriculture unfavorable self-employed

11-Not in the labor force-any other state

Table B2. Probabilities of transition matrix for men (Sum of transitions for the period 2009-2012)

Source: For each stata, the first row show the number of observation in each transition, the second row

should the probabilities that in status j at time t+k conditional on being in status Z at time t, the third row

shows the probabilities that an individual in status j at time t+k was in status z a time t.

Status 3 and 5, above poverty line.

1 2 3 4 5 6 7 8 9 10 11 Total in tTotal in

t - Diagonal

%

Mobility

1-Formal salaried employment 3349 327 96 5 24 25 20 123 16 5 41 4031 682 8.0

83.1 8.1 2.4 0.1 0.6 0.6 0.5 3.1 0.4 0.1 1.0 100.0

82.2 5.8 6.4 0.4 2.2 1.1 1.8 8.7 1.3 3.7 1.9 18.6

2-Informal salaried employment 379 3340 415 11 334 198 202 406 68 24 142 5519 2179 25.5

6.9 60.5 7.5 0.2 6.1 3.6 3.7 7.4 1.2 0.4 2.6 100.0

9.3 59.5 27.7 0.9 30.0 9.1 17.8 28.9 5.3 17.9 6.7 25.4

3. Advantageous self-employed not

agriculture 80 356 1640 28 89 54 50 81 8 18 97 2501 861 10.1

3.2 14.2 65.6 1.1 3.6 2.2 2.0 3.2 0.3 0.7 3.9 100.0

2.0 6.3 109.5 2.3 8.0 2.5 4.4 5.8 0.6 13.4 4.6 11.5

4. Unfavorable self-employed not

agriculture 4 13 28 9 1 4 5 5 2 1 15 87 78 0.9

4.6 14.9 32.2 10.3 1.1 4.6 5.7 5.7 2.3 1.1 17.2 100.0

0.1 0.2 1.9 0.8 0.1 0.2 0.4 0.4 0.2 0.7 0.7 0.4

18 325 83 2 1289 273 41 99 3 5 107 2245 956 11.2

0.8 14.5 3.7 0.1 57.4 12.2 1.8 4.4 0.1 0.2 4.8 100.0

0.4 5.8 5.5 0.2 115.6 12.5 3.6 7.0 0.2 3.7 5.1 10.3

12 228 49 2 292 276 32 44 3 2 80 1020 744 8.7

1.2 22.4 4.8 0.2 28.6 27.1 3.1 4.3 0.3 0.2 7.8 100.0

0.3 4.1 3.3 0.2 26.2 12.7 2.8 3.1 0.2 1.5 3.8 4.7

7-Unpaid family workers 24 250 51 1 57 33 531 58 111 10 43 1169 638 7.5

2.1 21.4 4.4 0.1 4.9 2.8 45.4 5.0 9.5 0.9 3.7 100.0

0.6 4.5 3.4 0.1 5.1 1.5 46.7 4.1 8.7 7.5 2.0 5.4

8-Unemployed 109 429 114 6 90 66 66 347 31 24 113 1395 1048 12.2

7.8 30.8 8.2 0.4 6.5 4.7 4.7 24.9 2.2 1.7 8.1 100.0

2.7 7.6 7.6 0.5 8.1 3.0 5.8 24.7 2.4 17.9 5.4 6.4

9. Non-labor / studing 51 165 20 7 5 12 141 116 1006 10 58 1591 585 6.8

3.2 10.4 1.3 0.4 0.3 0.8 8.9 7.3 63.2 0.6 3.6 100.0

1.3 2.9 1.3 0.6 0.4 0.6 12.4 8.3 78.9 7.5 2.8 7.3

10. Non-labor force at domestic work 6 29 16 1 5 5 15 30 4 8 38 157 149 1.7

3.8 18.5 10.2 0.6 3.2 3.2 9.6 19.1 2.5 5.1 24.2 100.0

0.1 0.5 1.1 0.1 0.4 0.2 1.3 2.1 0.3 6.0 1.8 0.7

42 148 90 16 78 82 34 97 23 27 1375 2012 637 7.4

2.1 7.4 4.5 0.8 3.9 4.1 1.7 4.8 1.1 1.3 68.3 100.0

1.0 2.6 6.0 1.3 7.0 3.8 3.0 6.9 1.8 20.1 65.2 9.3

Total in T+1 4074 5610 1498 1192 1115 2177 1137 1406 1275 134 2109 21727 8557 100.0

18.8 25.8 6.9 5.5 5.1 10.0 5.2 6.5 5.9 0.6 9.7 100.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 39.4

Total in T+ 1 - Diagonal 725 2270 -142 1183 -174 1901 606 1059 269 126 734 8557

% Mobility 8.5 26.5 -1.7 13.8 -2.0 22.2 7.1 12.4 3.1 1.5 8.6

5. Agricultural advantageous self-

employed

6. Agriculture unfavorable self-employed

11-Not in the labor force-any other state

Table B3. Marginal effects on the probability of entering a successful state, comparison by gender

Notes: Table reports marginal effects evaluated at the means of all variables, from estimates of itβ in Eq. (x) using probit

regressions for samples identified by column. A positive coefficient means that an increase in the related variable,

increases the probability that a person enters a particular successful state.

n.i./Not included in this model d.c/Dropped because of collinearity Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Log Likelihood -2189.0447 -1403.2752 -2614.7208 -4667.1051 -2753.1916 -500.79721

Number of obs. 13,062 19,061 13,062 19,061 13,062 19,061

Pseudo R2 0.1319 0.2227 0.1174 0.0931 0.1338 0.1323

AGE 0.00395 *** 0.000517 *** 0.00636 *** 0.00851 *** 0.00364 *** 0.000335 ***

(0.000503) (0.000152) (0.000579) (0.000621) (0.000532) (9.08e-05)

AGE_squared -5.06e-05 *** -9.27e-06 *** -6.66e-05 *** -9.53e-05 *** -3.56e-05 *** -3.16e-06 ***

(5.76e-06) (1.83e-06) (6.16e-06) (6.66e-06) (5.34e-06) (8.99e-07)

Primary completed 0.0205 *** 0.00427 ** 0.0133 ** 0.00592 -0.0197 *** -0.00150 ***

(0.00573) (0.00212) (0.00545) (0.00533) (0.00341) (0.000482)

Enrolled in secondary 0.0337 *** 0.00664 *** 0.00724 0.00370 -0.0305 *** -0.00160 ***

(0.00586) (0.00220) (0.00520) (0.00543) (0.00366) (0.000567)

Secondary completed 0.0830 *** 0.0297 *** 0.0324 *** 0.00270 -0.0253 *** -0.00127 **

(0.0118) (0.00624) (0.00865) (0.00624) (0.00412) (0.000599)

Higher level studies 0.101 *** 0.0358 *** 0.0163 -0.00311 -0.0350 *** n.i.

(0.0163) (0.00859) (0.0100) (0.00868) (0.00409)

Head of household 0.00367 0.000707 0.0290 *** 0.0565 *** 0.0405 *** 0.00412 *

(0.00438) (0.00145) (0.00580) (0.00928) (0.00579) (0.00217)

Spouse 0.00112 -0.00202 ** 0.0419 ** 0.0350 *** 0.0186 0.000371

(0.00907) (0.00103) (0.0168) (0.00595) (0.0141) (0.000926)

Tow ards married status 0.00268 0.000640 0.0161 0.0124 0.0269 n.i.

(0.00887) (0.00297) (0.0150) (0.0174) (0.0176)

Tow ards single status -0.0164 ** 0.00377 0.0786 *** 0.0307 ** -0.0295 *** 0.0104 *

(0.00656) (0.00365) (0.0244) (0.0146) (0.00640) (0.00532)

High population density area 0.0257 *** 0.00847 *** 0.00921 * 0.00842 * -0.0361 *** n.i.

(0.00504) (0.00192) (0.00515) (0.00508) (0.00332)

Manufacturing & construction 0.00448 0.00957 ** 0.0511 *** 0.101 *** n.i. n.i.

(0.00424) (0.00450) (0.00760) (0.0178)

Commerce 0.0291 *** 0.00610 *** 0.0583 *** 0.116 *** n.i. n.i.

(0.00699) (0.00216) (0.00995) (0.0104)

Textile & clothing industry 0.0575 * 0.0347 *** 0.0589 0.143 *** n.i. n.i.

(0.0331) (0.0119) (0.0403) (0.0257)

High complexity services 0.0546 *** 0.0128 *** 0.0520 *** 0.0512 *** n.i. n.i.

(0.0135) (0.00374) (0.0143) (0.00957)

Low complexity services 0.00558 *** 0.00273 *** 0.00472 *** 0.00123 n.i. n.i.

(0.000934) (0.000519) (0.00130) (0.00331)

Age 6 or younger 0.00141 -0.000661 -0.00335 0.00766 *** 0.00328 0.000380

(0.00179) (0.000464) (0.00239) (0.00214) (0.00210) (0.000288)

Age 07_18 -0.000455 -0.000563 ** -0.00378 *** -0.000589 0.00222 ** 0.000211

(0.000941) (0.000279) (0.00125) (0.00122) (0.00110) (0.000152)

Age 19_65 0.00277 ** 0.000122 -0.00386 ** -0.00296 * 0.00239 -0.000494 **

(0.00121) (0.000320) (0.00163) (0.00166) (0.00151) (0.000237)

Age 65+ -0.00234 -0.000712 -0.00545 0.00391 -0.00168 -6.81e-05

(0.00284) (0.000743) (0.00366) (0.00377) (0.00345) (0.000503)

Tubed w ater inside dw elling 0.0217 *** 0.00205 0.0124 * 0.0154 ** -0.0361 *** -0.000606

(0.00369) (0.00175) (0.00695) (0.00698) (0.00908) (0.00102)

Electricity -0.00392 -0.000319 0.00528 0.00801 * 0.00172 -0.00117

(0.00374) (0.00112) (0.00418) (0.00439) (0.00381) (0.000770)

Water 0.000817 0.000550 0.0109 *** 0.00365 -0.00776 * -0.000593

(0.00323) (0.000895) (0.00383) (0.00420) (0.00403) (0.000604)

Remittances -0.0254 * 0.00117 0.00136 -0.0274 * 0.0796 *** 0.00183

(0.0141) (0.00200) (0.0178) (0.0152) (0.0137) (0.00127)

Non labor income -0.00636 * 0.000355 0.0119 *** -0.00367 -0.0161 ** -3.91e-05

(0.00382) (0.000733) (0.00432) (0.00509) (0.00677) (0.000843)

d09 -0.00365 0.000263 0.00133 -0.00713 -0.0141 *** 0.000536

(0.00360) (0.00102) (0.00511) (0.00486) (0.00439) (0.000855)

d10 -0.00164 7.79e-05 0.00184 -0.00356 0.00644 0.00113

(0.00360) (0.000993) (0.00493) (0.00484) (0.00500) (0.000919)

d11 -0.00660 * 0.000616 0.00410 0.00102 0.00482 0.000972

(0.00340) (0.00101) (0.00499) (0.00494) (0.00499) (0.000894)

d12 d.c. d.c. d.c. d.c. d.c. d.c.

Women

MARGINAL EFFECTS ON THE PROBABILITY OF ENTERING A SUCCESSFUL STATE, COMPARISON BY GENDER

Men Women

Salaried formal sector Advantageous non-agricultural SE

Men Women Men

Advantageous agricultural SEVariables

Table B4. Marginal effects on the probability of leaving a successful state, comparison by gender

Notes: Table reports marginal effects evaluated at the means of all variables, from estimates of itβ in Eq. (x) using probit

regressions for samples identified by column. A positive coefficient means that an increase in the related variable,

increases the probability that a person leaves a particular successful state.

n.i./Not included in this model d.c/Dropped because of collinearity Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Log Likelihood -1461.3613 -769.9571 -1323.0416 -2429.9628 -1391.472 -80.237314

Number of obs. 3,911 2,264 2,332 4,074 2,145 151

Pseudo R2 0.0922 0.1075 0.067 0.0644 0.0356 0.1502

AGE -0.0169 *** -0.0240 *** -0.0188 *** -0.0259 *** -0.0234 *** -0.0185

(0.00288) (0.00396) (0.00450) (0.00371) (0.00451) (0.0263)

AGE_squared 0.000200 *** 0.000262 *** 0.000196 *** 0.000268 *** 0.000235 *** 0.000205

(3.41e-05) (4.96e-05) (4.73e-05) (3.89e-05) (4.54e-05) (0.000253)

Primary completed 0.0337 0.0537 0.00641 -0.00470 0.0299 0.0476

(0.0218) (0.0423) (0.0274) (0.0215) (0.0356) (0.144)

Enrolled in secondary -0.0233 -0.0117 -0.0486 * -0.00818 -0.0734 * n.i.

(0.0166) (0.0284) (0.0258) (0.0229) (0.0416)

Secondary completed -0.0635 *** -0.0250 -0.0715 ** -0.00657 -0.0435 n.i.

(0.0154) (0.0275) (0.0291) (0.0279) (0.0674)

Higher level studies -0.0767 *** -0.0451 -0.0871 ** 0.0523 -0.235 ** n.i.

(0.0157) (0.0299) (0.0352) (0.0490) (0.119)

Head of household -0.113 *** 0.0267 -0.212 *** -0.0839 *** -0.171 *** -0.278 **

(0.0234) (0.0267) (0.0444) (0.0309) (0.0486) (0.116)

Spouse -0.0273 0.0211 -0.181 *** 0.0155 -0.101 0.155

(0.0279) (0.0205) (0.0393) (0.0284) (0.0772) (0.219)

Tow ards married status 0.00564 -0.0130 -0.0247 0.0951 -0.0276 0.136

(0.0402) (0.0380) (0.0794) (0.0722) (0.0918) (0.148)

Tow ards single status 0.104 * 0.0526 0.108 -0.0383 0.135 n.i.

(0.0603) (0.0531) (0.0797) (0.0479) (0.112)

High population density area -0.0233 ** -0.0360 *** -0.0229 -0.0113 -0.0346 n.i.

(0.0115) (0.0134) (0.0236) (0.0202) (0.0820)

Manufacturing & construction -0.00476 -0.0668 ** -0.0123 -0.238 *** n.i. n.i.

(0.0299) (0.0317) (0.0341) (0.0195)

Commerce 0.00261 -0.0458 -0.101 *** -0.292 *** n.i. n.i.

(0.0315) (0.0411) (0.0321) (0.0238)

Textile & clothing industry -0.0642 *** -0.0902 *** -0.124 ** -0.145 *** n.i. n.i.

(0.0225) (0.0290) (0.0521) (0.0291)

High complexity services -0.0208 0.0219 d.c. d.c. n.i. n.i.

(0.0294) (0.0630)

Low complexity services -0.0170 *** -0.0293 *** -0.0190 ** -0.0500 *** n.i. n.i.

(0.00601) (0.0107) (0.00785) (0.0141)

Age 6 or younger -0.00758 -0.000326 -0.000406 0.00605 -0.00248 -0.0709

(0.00763) (0.00941) (0.0139) (0.0109) (0.0133) (0.0550)

Age 07_18 0.00709 0.00257 0.00525 -0.00751 -0.00275 0.0247

(0.00439) (0.00584) (0.00782) (0.00605) (0.00710) (0.0267)

Age 19_65 -0.00999 * -0.00666 0.0276 *** 0.0152 * 0.00668 0.101 **

(0.00571) (0.00669) (0.0104) (0.00807) (0.0101) (0.0400)

Age 65+ -0.00600 0.0250 * 0.0125 0.0245 -0.0137 0.00964

(0.0138) (0.0142) (0.0246) (0.0181) (0.0234) (0.100)

Tubed w ater inside dw elling -0.150 * d.c. -0.0767 -0.145 *** 0.00689 0.0829

(0.0823) (0.0804) (0.0527) (0.0359) (0.123)

Electricity -0.0375 ** -0.0133 -0.144 *** -0.0547 ** -0.0247 0.0465

(0.0190) (0.0283) (0.0353) (0.0252) (0.0278) (0.115)

Water -0.00213 -0.0153 -0.0517 * -0.0442 ** 0.0205 -0.0369

(0.0158) (0.0242) (0.0276) (0.0209) (0.0247) (0.0960)

Remittances 0.0768 -0.0135 0.0983 0.116 -0.0490 0.390

(0.0800) (0.0707) (0.130) (0.0712) (0.0943) (0.300)

Non labor income 0.0319 *** 0.0274 ** -0.0243 -0.0498 ** -0.159 *** -0.135

(0.0116) (0.0116) (0.0254) (0.0224) (0.0412) (0.119)

d09 -0.0308 ** 0.00270 -0.0427 0.00332 0.0183 0.0517

(0.0142) (0.0188) (0.0285) (0.0233) (0.0356) (0.128)

d10 -0.0263 * -0.0243 -0.0269 -0.0297 0.0513 -0.0753

(0.0141) (0.0176) (0.0284) (0.0224) (0.0354) (0.135)

d11 -0.0442 *** -0.0159 0.0302 -0.0129 0.00898 -0.0191

(0.0136) (0.0178) (0.0288) (0.0225) (0.0352) (0.130)

d12 d.c. d.c. d.c. d.c. d.c. d.c.

Men Women Men Women Men Women

Variables

MARGINAL EFFECTS ON THE PROBABILITY OF LEAVING A SUCCESSFUL STATE, COMPARISON BY GENDER

Salaried formal sector Advantageous non-agricultural SE Advantageous agricultural SE