




































American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

75 | P a g e  

THE IMPACT OF TRADE UNION MEMBERSHIP ON WAGE 
LEVELS IN BENIN 

 
 

1Aimee Mahougbe and 2Achille Assouto 
1Faculty of Economics and Management (FASEG), University of Abomey-Calavi (UAC), 

2Orléans Economics Laboratory (LEO), University of Orléans, CNRS, FRE 2014 
 

Abstract: The labor market, particularly the role of trade unions, has garnered significant attention in 
developed countries. This research explores the impact of trade unions on wage levels and wage 
inequality, drawing from established literature. Trade unions play a crucial role in advocating for higher 
wages, better working conditions, and increased job security for their members, which in turn has 
implications for wage disparities. This study aims to shed light on the strategies employed by trade 
unions to improve the well-being of their members, including negotiation processes with employers 
and public authorities, as well as the use of strike actions. By delving into the functioning of labor 
markets and the influence of trade unions, this research contributes to our understanding of wage 
dynamics and inequality in developed economies. 
Keywords: labor market, trade unions, wages, wage inequality, negotiation, strike actions 
  
 
1. Introduction  
The functioning of the labor market, in particular the role of institutions such as trade unions, has been 
the subject of much work in developed countries in recent decades (Schmidt, 2021; Card et al., 2020). 
Since the seminal work of Dunlop (1944), Ross (1948) and Lewis (1963), it has been accepted that trade 
unions contribute to the increase in the wages of their members and are therefore likely to reduce any 
increase in the wages of their members.  wage inequalities. The main objective of trade unions is to 
improve the well-being of union members by negotiating higher wages, better working conditions and 
increased tenure (Manda et al., 2005). To achieve this, trade unions use many legal means at their 
disposal, ranging from platforms for discussions with employers and/or public authorities to strike 
action.  
However, restrictions on union activities limit their ability to use strikes to gain benefits from 
employees. However, unions can influence the level of pay and working conditions of their members 
through collective bargaining. It is noted that these unions sometimes lack the resources to hire 
negotiators whose skills and capabilities are comparable to those of full-time employer negotiators. 
Despite the difficulties, union leaders were required to insist on workers' demands for substantial 
results in order to continue to enjoy the support of their rank and file (Sandbrook, 1975).   
Despite the ever-increasing body of research on the impact of unions on wages and wage inequality in 
developed countries, empirical evidence of the impact of unions on incomes in developing countries, 
particularly in Africa, is still scarce (Kerr and Wittenberg, 2021; Manda et al, 2005). Worldwide, most 
empirical studies on the influence of trade unions on wages are based on the experiences of 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

76 | P a g e  

industrialized countries, and few attempts have been made to estimate the relationship for less 
developed countries, particularly in sub-Saharan Africa. Freeman (2010) argues that there has been 
relatively little research on the impact of trade unions in developing countries. Better still, he believes 
that the few works available are mainly based on small, unrepresentative samples. Benin is an 
interesting context for studying the impact of trade unions on wages.  
Indeed, several changes have been made in recent years to the regulatory framework, including the 
Labor Code and the law on the exercise of the right to strike.  Périsse (2014) notes that these changes 
often stem from endemic labor disputes that stem from the existence of a concentration of industries 
and low- or unskilled workers that cannot be regulated by the labor code focused on the individual 
employment contract. Thus, recent policy reforms in Benin's labor market are expected to both promote 
private sector expansion and ensure the continuity of public services. However, it is believed that the 
new environment thus defined seems to have caused a weakening of trade union power because of the 
right to strike abolished for some trades and reframed for others.  The relevance of the role of trade 
unions in improving working conditions must therefore be questioned.  Torm (2018) questions the 
effectiveness of trade unions due in part to their marginal independence, employer dominance, and the 
generally low incidence of collective agreements.  
Amendments to labor regulations in Benin in recent years reinforce the Beninese context, which is 
becoming a relevant case study for analyzing the functioning of labor markets.  In addition, the 
evolution of the guaranteed interprofessional minimum wage (SMIG) since 1994 reflects a certain 
upward rigidity in wages. Indeed, it has increased annually by an average of CFAF 1,150 from 1994 to 
2022. In a context where Benin’s economy is dependent on exogenous shocks, the most recent of which 
are the Covid-19 pandemic and the Russian-Ukrainian conflict, the erosion of purchasing power is likely 
to exacerbate tensions between workers and employers.  
The objective of this article is to assess the wage impact of unionization of workers in Benin. The 
contribution of the article is twofold.  First, it is one of the few studies exploring the effects of workers' 
union membership in the West African sub-region.  Less is known, therefore, about the gains from 
union membership for the worker in Africa.  The countries dealing with this theme are mostly developed 
countries (Keune, 2018; Borland, 1996) or developing regions other than Africa (Liu et al, 2020; Torm, 
2018).  The work of Manda et al.  (2005) in Kenya and Ngom (2021) in Senegal are exceptions. But this 
article differs from the study by Manda et al (2005) in that it deals with all sectors of activity from 
primary to tertiary. Similarly, unlike Ngom (2021), this study uses a methodology to correct for 
potential problems of endogeneity and unobserved heterogeneity that may occur in these regressions.  
The decision to join a trade union is voluntary or prompted by the existence of trade union 
organizations in the branch of activity. In addition, awareness of the benefits inherent in union struggle 
could be associated with the worker's level of education and experience.  This results in self-selection 
that generates endogeneity problems, the failure to take into account of which can lead to biased results.  
Second, we consider duality, which is characteristic of labor markets in the majority of African countries 
where the formal and informal sectors coexist. Taking this aspect into account is of particular 
importance because the informal sector occupies a prominent place in developing economies, 
particularly in Africa (Freeman 2010). For example, in Benin, about 80% of workers work in the 
informal sector (Yedomon, 2016). The methodological approach used to achieve the objective relies on 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

77 | P a g e  

regime-change regression models to address potential issues of selectivity and endogeneity bias. The 
results show that membership of a trade union organization positively affects the level of pay. There 
has been a 33.06% increase in average hourly wages associated with unionization. In addition, this 
remuneration is influenced by the skill levels of the workers as well as the length of time spent in work. 
On the other hand, employees in the formal sector are the most likely to unionize.  
The rest of the article is organized as follows. Section 2 presents Benin's institutional framework, while 
Section 3 summarizes the available literature on the effects of trade union membership. The fourth 
section is devoted to the description of the methodological approach as well as the data. The results and 
discussions are discussed in Section 5 while the last section concludes.  
2. Institutional framework of the trade union movement in Benin  
The trade union movement in Benin is historically linked to the main characteristic phases of the 
country's political history, from the pre-independence period to the post-conference of the nation's vital 
forces of February 1990, marked by multiple structural reforms (Friedrich Ebert Stiftung, 2015). The 
Beninese trade union landscape underwent a major upheaval in the 1990s with the emergence of trade 
union pluralism, thanks to the full multiparty system devoted to the national conference of February 
1990. Trade union pluralism is in harmony with Benin’s labor regulations, which recognize the right of 
workers to form or join trade unions.       
In the era of democratic renewal, there was an increase in the number of trade unions organized into 
federations or confederations. The Danish Trade Union Development Agency counts in 2019, eight 
trade union confederations for more than 500 affiliated unions.  Trade unions operate in both formal 
and informal sectors. The formal sector is characterized by strong unionization. Environ 75% of 
employees in this sector belonged to a union in the early 2010s. However, formal employment 
accounted for only 10% of total employment. As for the informal sector, trade union action began by 
covering workers in this sector with the existence of a confederation dedicated to informal workers, 
namely the “Centrale des Syndicats du Privé et de l’Informel du Bénin” (CSPIB) even if the latter has 
seen a decline in the number of its members.  In addition, 23% of unions affiliated to the National Union 
of Workers' Unions of Benin (UNSTB) operate with organized workers from the informal economy 
(DTUDA, 2022).  
The public sector has a strong unionization characterized by a preponderance of organizations in the 
education subsector followed by the health subsector. The education subsector alone accounts for 
nearly 40 per cent of rank-and-file unions in the public sector. The importance of unionization in 
education is explained both by the large proportion of teachers out of all public sector workers and by 
the visibility of their union actions. Indeed, the staff of the ministries of the three levels of education, 
as shown in Figure 1, represented in 2017 more than 50% of the civil staff of the State (Ministry of Labor 
and the Public Federation (MTFP), 2018). In addition, the strike movements usually called in the 
education subsector have the particularity of affecting most Beninese households that have their 
children enrolled in public education. The health subsector comes in second place with a total of forty-
three (43) rank-and-file unions divided into corporatist unions, national unions or representative 
unions of departmental and university hospitals. In addition to these two main subsectors, there are a 
significant number of trade unions in the public administration, particularly in the various ministerial 
departments.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

78 | P a g e  

Figure 1: Number of civilian government employees between 2011 and 2017 in the three levels of 
education  

  
In contrast to the public sector, there is low unionization in the private sector. In this private sector, a 
distinction must be made between the informal and the formal. In the formal private sector, trade 
unions are mainly concentrated in the cement industry, the banking sector, the press and the service 
sector. On the other hand, in the informal private sector, they are made up of actors from the 
agricultural and craft sectors (Friedrich Ebert Stiftung, 2019). The low visibility of trade union action 
in the formal sector can be explained either by a relative success of social dialogue or a lack of knowledge 
of the texts governing workers' rights. Although the Beninese State has adopted a political and 
regulatory framework favorable to their development, trade unionism in the informal sector is 
unsuccessful. This is due to illiteracy, the strong presence of women and the lack of awareness of 
workers' rights. However, the informal sector remains the main sector providing employment, with 
more than 80% of the workforce employed. It should be noted, however, that the informal economy is 
experiencing slow but increasing unionization with the emergence of large unions, even if they have 
limited influence (Friedrich Ebert Stiftung, 2019).  
All in all, Benin's economy has seen an increase in the membership of rank-and-file unions in recent 
decades. There are eight (08) confederations that took part in the 2021 professional elections for more 
than 674 rank-and-file unions. Three of the eight confederations were elected to be interlocutors in 
negotiations with the State.  However, the unionization rate of employment fell by four percentage 
points during the 2010s, falling to 11% in 2019, a relatively high level compared to other West African 
countries.  There is also a proliferation of trade unions sometimes operating in the same sector with the 
same objectives. The result is a war of leadership, political clientelism and internal governance 
problems that relegate to the background the mission of defending workers' rights.  
Despite problems of organization and governance, trade union struggles have for decades produced 
significant social and economic results in some sectors. For example, in the public sector, education 
workers have been granted an increase in the index point through negotiations with the government. 
For example, the salary index for kindergarten, primary and secondary school teachers increased 1.25-
fold from January 2011.  
Since October 2010, higher education teachers have benefited from an improvement in their index 
point, which has been multiplied according to grade by coefficients ranging from 1.5 to 3 (General 
Directorate of the Economy, 2014). Similar benefits were also granted to workers in the Ministry of 
Finance as well as to the corporation of magistrates. The positive results have been achieved following 

  
Source: Authors, based on MTFP data (2018)   

44903   44811   44811   44688   43611   44542   47912   
66144   70475   70494   69041   73090   72285   77546   

0   

50000   

100000   

2011   2012   2013   2014   2015   2016   2017   

Trois ordres d'enseignement   Total   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

79 | P a g e  

numerous strikes that reduce the performance of the economy. In 2014, the General Directorate of the 
Economy estimates the losses induced by the 2010 strike at around 94,189,680,000 FCFA and 
33,747,624,076 FCFA respectively for GDP and tax revenues.  
Following an inventory of the trade union struggle, the public authorities undertook a reform with a 
view to reorganizing trade union action in Benin. The reform consisted of the adoption of new laws to 
regulate the right to strike as well as the scope of action of the various trade union organizations. For 
example, the National Council for Social Dialogue (CNDS) was created by decree in 2017 to ensure the 
organization and management of social dialogue on socio-economic issues of common interest to the 
State, employers and trade unions. The CNDS introduced changes in hiring, placement of labor and 
termination of the employment contract. It was dissolved in 2022 and replaced by the National 
Commission for Consultation, Consultation and Collective Bargaining. Another reform concerns the 
law on the exercise of the right to strike has been amended, limiting the maximum duration of a strike 
to 10 days per year for all employees. Several groups are banned from strikes, such as the military, 
paramilitaries (police, customs, water, forests, hunting, etc.) and health personnel. Another provision 
provides that strikes motivated by the violation of universally recognized trade union rights may not be 
deducted from wages (Danish Trade Union Development Agency, 202-2).  
The reforms thus implemented can be interpreted as the consequence of potential adverse effects on 
economic performance.  But analyzed fromthis point of view, it would obscure the ability of trade unions 
to improve the living conditions of workers, who can in turn be encouraged to increase their 
productivity. Such reforms then raise questions about the usefulness of trade union organizations in 
the various sectors of activity in Benin. The idea is to compare the effects of these trade union 
organizations on the lives of Beninese workers with the results provided by the literature.  
3. Literature Review  
Recent research has failed to identify a general law on the effects of institutions on labor market 
outcomes, but it has led to new, sometimes surprising, conclusions about how institutions affect 
outcomes (Kerr and Wittenberg, 2021; Card et al., 2020; Ge, 2014; Liu et al., 2020; Yao and Zhong, 
2013).  Economic circumstances and institutions probably vary too much from country to country to 
allow for a single generalization, may be the reason. The absence of such a general law leads to a more 
measured view of what institutions do.  The synthesis of previous work proposed here first examines 
the general theoretical framework before highlighting the main results obtained in the empirical 
literature. 
Economists use three types of theories to analyze labor institutions. The first, which Freeman (2010) 
calls distortions, suggests that institutions disrupt the equilibrium of a competitive market. Union-led 
wage negotiations would result in higher labor costs, which in turn leads unionized companies to reduce 
employment. As a result, some workers turn to lower-paying and less productive non-unionized jobs, 
reducing economic efficiency. The higher the elasticity of labor demand, the greater the distortion in 
resource allocation. In contrast to the first theory, the second type of theory views institutions as 
effective negotiation mechanisms that promote optimal resource allocation.  In an effective negotiation 
model, the consensus reached benefits both managers and employees, thus avoiding the waste of 
resources. This is Coase's theorem at work in the world of labor institutions (Freeman, 1993).  As a 
result, rules determined by the institutions, such as employment protection legislation, affect the 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

80 | P a g e  

distribution of resources but not production. The third theory focuses on the idea that institutions 
eliminate information asymmetries and facilitate cooperative behavior, which increases productivity. 
Freeman and Lazear (1995) view work councils as institutions that increase communication within 
firms and enable management and workers to make more informed decisions. On the other hand, where 
there is pronounced wage inequality due to lack of competition, as is the case in monopsony (Manning, 
2013) or due to information failures or other factors, collective bargaining or government regulations 
may bring wages closer to the market equilibrium level. 
There are theoretical arguments that unions increase, reduce or do not affect the well-being of workers. 
Freeman (1980) and Freeman and Medoff (1984) argued that in addition to increasing their members' 
wages, unions reduced inequality and had other non-wage benefits, which they argued stemmed from 
the collective voice/institutional response aspect of union activity. These include solidarity and impacts 
on wage inequality, improving firms' productivity by providing a voice for workers that facilitates 
communication between workers and management, and a voice in broader societal debates. The 
reduction in wage inequality was achieved through the "standard rates" that unions negotiated both 
within and between companies. Freeman (1980) called the reduction of inequality in the union sector 
the intra-sectoral effect. He showed that this intra-industry effect was large and negative in the United 
States in the 1970s, meaning that the overall impact of unions was to reduce inequality, particularly in 
the manufacturing sector.  
Empirical work is also part of the same logic but emphasizes more the benefits of union membership 
for workers (Card et al., 2020; Ge, 2014; Bryson, 2014; Laroche, 2004; Coutrot, 1996; Najem and 
Paquet, 2007). Research to estimate the extent to which unions raise their members' wages has indeed 
stimulated a growing literature on causal inference (Card et al., 2020; Ge, 2014; Yao and Zhong, 2013; 
Read et al., 2010). But the work in general is more the responsibility of developed countries or 
developing countries (DCs) other than those of Africa. For example, Lu et et al.  (2010) using firm-level 
data, find that unions lead to better benefits and more contract signings. However, workers in unionized 
companies do not enjoy higher wages and bonuses than their counterparts in non-unionized 
companies. Zhang et al. (2011) also conclude that trade unions in China do not play a significant role in 
the incidence of industrial labor income. For its part, Ge (2014) notes that unions in the workplace 
significantly improve wage and non-wage compensation, as well as employee training. Based on 
company data, the author shows that the presence of unions within the same region and industry 
generates positive spin-offs for employee compensation. Yao and Zhong (2013) found that unionized 
firms were associated with significantly higher average wages and a larger fraction of employees 
covered by a pension. In contrast, Budd et al.  (2014) found that union density does not affect average 
wage levels, even though it is positively associated with employee productivity and overall output.  
When it comes to work in developing countries, particularly in Africa, very few are interested in the 
effects of unionization (Kerr and Wittenberg, 2021; Ngom, 2021; Manda et al., 2005). Manda et al.  
(2005) report a positive wage effect of trade union membership in Kenya. In other words, the 
unionization of the worker generates a wage premium. The study further shows that elite workers tend 
to refrain from unionizing. More recently, Kerr and Wittenberg (2021) note extremely high union wage 
premiums in South Africa leading to a slight increase in pay inequality in recent years. But unlike 
Manda et al.  (2005), Kerr and Wittenberg (2021) find that union membership has become increasingly 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

81 | P a g e  

concentrated at the top of the wage distribution and in the public sector. Kerr and Wittenberg (2021) 
nevertheless argue that unions do more for those at the bottom than for those at the top. Card et al.  
(2004) note that this result could be generated by positive selection in unions at the bottom and 
negative selection at the top, so that unions appear to reduce inequality more than they actually do.  
Ngom (2021), for his part, highlights the lack of a correlation between the unionization rate and labor 
productivity in Senegal.  However, unionization has positive effects in terms of wage levels.  
It emerges from this literature that the effects of affiliation lead to rather contrasting results in terms 
of wages or working conditions. Overall, the empirical literature supports a positive effect of union 
membership, although in some cases the authors highlight a lack of correlation. In Benin as in the West 
African region, few studies have focused on the impact of unionization. This study attempts to examine 
the role of unionization in improving workers' wages in Benin. It is based on an adapted methodology 
for estimating the impact of unionization on wages.  
4. Methodology and data used  
In this section, we describe the estimation strategy adopted and then present the data used.   
4.1. Estimation method  
Since the purpose of this article is to examine the effects of membership of a trade union organization 
on the level of wage compensation, it is important to consider the methodological problems associated 
with such an analysis.  Estimating the effects of unionization on wages comes up against the problem 
of endogeneity which, if not taken into account, would lead to biased results. Indeed, membership of a 
trade union is either voluntary or induced by the existence of a trade union in the branch of activity and 
the awareness of the advantages inherent in the trade union struggle.  Knowledge of the interests of 
union membership can be improved through education and experience. For example, workers with a 
high level of education have a deeper knowledge of the issues associated with unionization and are likely 
to join than workers with no or low levels of education.  Thus, selfselection in union membership is an 
important source of endogeneity.    
Moreover, the literature has shown that affiliation to a union depends not only on the observable 
characteristics of workers, but also on certain unobservable characteristics (Laroche, 2017).  Failure to 
take these into account can lead to overestimating, underestimating or reporting the impact where it 
does not exist at all.  Several methods have been developed and proposed in the literature to address 
potential self-selection bias and to allow comparability of the two groups.  These are Heckman selection 
methods, instrumental variables, propensity score matching and regime-change regression.  
In addition to the limitations specific to each method, they are, for the most part, unable to correct 
selection biases related to the endogeneity of the treatment variable (Heckman et al., 1997; Khonje et 
al., 2015), resulting in biased results. A suitable method for such situations is   regression with regime 
change (Lokshin and Sajaia, 2004; Maddala, 1983; Lee, 1978). Nahm et al. (2017) also draw attention 
to these methodological issues in the analysis of the effects of unionisation.  The approach used in this 
article to estimate the gain from union membership is therefore based on a regression model with 
regime change.  
To do this,we start from two earnings functions, one for unionized workers (equation 1a) and the other 
for non-unionized workers (equation 1b).  

  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

82 | P a g e  

 where is the log hourly earnings of unionized workers 

and the log hourly earnings of non-unionized  
workers. The vector includes the standard explanatory variables, while and are the error terms.

 Equation 2 represents the choice of union membership  
       

si 
( ) si 
Where is a latent variable determining whether union membership, , is observed, and is our estimate 
of union premium  is a vector of instrument variables which do not directly influence the worker's wage 
level but which intervene in the decision on union membership. In this study we use as an instrument 
the availability or absence of trade union organization in the company.  Theavailability 

of trade union organization determines the decision to join, but has no effect on the level of wages unless 
the workers join and are active in it.  

 , and are parameters to be estimated. The error terms in both regimes ( , , )  are assumed to 

follow a trivariate normal distribution with a zero mean and a covariance matrix equal to .   
The earnings equations and the choice equation are a regime change regression model and can be used 
to estimate the determinants of union membership and its effect in terms of wage premiums. It is 
highly likely that the error terms in equations 1 and 2 are correlated, resulting in selectivity bias. Indeed, 
in equations 1 and 2, represents a vector of variables likely to influence both membership of a trade 
union organization and wage level. These are the characteristics of the worker such as qualification, 
age, sex, level of education, length of employment, etc.    

Unobserved factors such as skill and motivation that affect the selection system could also affect the 
employee's level of compensation.  For example, more skilled workers might be those who are more 
likely to join the union in the belief that they are asserting their rights.  The same workers are those who 
are more likely to have high productivity and therefore high pay.  Lee (1978) and Maddala (1983) note 
that the terms of errors and can be correlated and render inconsistent the estimators resulting from the 
application of ordinary least squares (OLM).  
To address this problem posed by the regime change regression model, the selection and wage 
equations  are estimated simultaneously using the full-information maximum likelihood method. 

This method has the advantage of providing robust standard error estimates in contrast to step-by-step 
methods by estimating equations separately (Guirkinger and Boucher, 2008; Lee, 1978). Under the 
assumptions made about the distributions of the error terms of equations 1 and 2, and according to 
Lokshin and Sajara (2004), the likelihood log function of the regression model with change of regime 
is given by:  

  

Where is a cumulative normal distribution function, is a density distribution normal function and:
  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

83 | P a g e  

, avec   

With the correlation coefficient between and ;  the correlation coefficient between and , with and 
respectively the covariances of and , and . , and represent the respective standard deviations of , and .

 
   

The estimation results of equation 4 using the full-information maximum likelihood method will be 
used to determine the potential gains or losses from union membership or the level of compensation 
that could be obtained by non-unionized employees if they had been unionized. The procedure will 
therefore consist of estimating for non-unionized employees. According to Maddala (1983), conditional 
expected gains are calculated using equation 6.   

  

Where and are parameters to be estimated from the regime change regression model.  The higher the 
value of the differential forecast, the greater the loss due to non-affiliation. If necessary, it would then 
be urgent to set up a mechanism to better organize the social partners that are the trade unions.  
4.2. Data used and descriptive statistics  
The data used in this article come from the Integrated Regional Survey on Employment and the 
Informal Sector (IRS-EIS) in the Member States of the West African Economic and Monetary Union 
carried out in 2017. The integrated regional survey on employment and the informal sector was 
conducted using a two-stage area probability survey with first-stage stratification. The objective of the 
survey was to produce statistically reliable estimates of indicators, at the national level, for urban and 
rural areas, and for each of the twelve departments of the country.    
A total of 23 survey strata were defined. Within each stratum, the samples of Zones of Enumeration 
(ZD) and households were divided according to the largest entities of the administrative division of the 
country and the place of residence. At first level, 680 ZDs were selected out of the 13,000 defined during 
the census cartographic work as part of the fourth RGPH (General Population and Housing Census) 
carried out in 2013. The ZDs were drawn with a probability proportional to the number of households. 
At the second level, a fixed number of fifteen (15) households was selected in each of the ZDs selected 
at the first degree with three (3) replacement households. The sample size of the ERI-ESI is 10,200 
households.  
The ERI-ESI survey covers two parts, the first of which concerns the collection of data on the 
sociodemographic characteristics and employment of the population and the second relates to the 
collection of data from informal non-agricultural production units identified in the first component. 
Four types of questionnaires were used, including an employment questionnaire administered in each 
household to all individuals aged 10 and over. The individual employment questionnaire comprises 
several modules including employment status, main activity, general job satisfaction, social protection, 
vocational training and further training, unemployment, trajectory and prospects, income outside 
employment, etc. The employment database obtained from the information collected using this 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

84 | P a g e  

questionnaire includes 22,027 observations. 1,775 individuals declared themselves employees while 
10,318 declared themselves self-employed. 1,184 receive family help.  
We extracted from the employment database, the observations relating to the 1,775 employees spread 
throughout the national territory. After clearance, a total of 1 135 employees were retained on whom 
the analyses were carried out.  The remaining observations were suppressed for reasons related to the 
lack of declared salaries.  The workers surveyed are concentrated in the Littoral and Atlantic 
department with 21.76% and 12.69% respectively. The importance of workers on the coast is justified 
by the fact that this department is home to the municipality of Cotonou, the main metropolis and 
generally referred to as the economic capital of the country.  
The variables included in the analysis are composed of the dependent and independent variables. The 
dependent variables are union status for the probit model, and log hourly wages for earnings functions. 
Union status is a binary variable that takes the value 1 if the worker declares that he belongs to a trade 
union or similar association of employees who can defend his labor rights or participate in collective 
bargaining and 0 otherwise.  Hourly wages are calculated by dividing monthly wages by the total 
number of hours worked in a month (that is. hours worked in a week multiplied by 4.3 weeks in a 
month). Explanatory variables include socio-demographic characteristics such as age, sex, marital 
status and education level.   
Other variables used to determine wage levels and unionization were mobilized. These are sectors of 
activity (public/private and formal/informal), vocational qualification, length of service and number of 
working hours per month.   
Table 1 shows the differences in the socio-economic characteristics of workers according to whether 
they belong to a trade union organization or not. The average comparison tests show that the two groups 
of employees differ fundamentally in terms of socioeconomic characteristics such as age, salary, sex, 
marital status and length of service.    
Table 1: Socioeconomic characteristics of employees by affiliation 

Variables  Together  Unaffiliated  Affiliated  T-test  

Average hourly wage  535,554  515,892  616,419  -1,269***  

Age  36.165  35.398  39.315  -
4.795***  

Formal sector  36.39 %  26.51 %  77.03 %  -
15.421***  

Female sex  23.00 %  24.10 %  18.47 %  1.788*  

Educational 
attainment  

        

None  27.05 %  29.24 %  18.02 %  3.391***  

Primary  14.71 %  16.32 %  8.11 %  3.108***  

Secondary  35.68 %  34.50 %  40.54 %  -1.685*  

Upper  22.56 %  19.93 %  33.33 %  -4.316***  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

85 | P a g e  

Marital status          

Bachelor  13.04 %  14.56 %  6.76 %  3.110***  

Polygamous married  10.40 %  9.64 %  13.51 %  -1.697***  

Monogamous married  65.81 %  64.73 %  70.27 %  -1.561  

Widowed/Divorced  10.75 %  11.06 %  9.46 %  0.691  

Length of service          

15 to 24 years  3.35 %  3.40 %  3.15 %  0.180  

10 to 14 years  12.69 %  11.17 %  18.92 %  -3.121***  

5 to 9 years  17.71 %  15.22 %  27.93 %  -
4.482***  

2 to 5 years  25.90 %  27.93 %  17.57 %  3.172***  

Less than 2 years  40.35 %  42.28 %  32.43 %  2.688***  

Actual  1 135  913  222    

Note: ***, ** and * indicate significance levels at thresholds 1%, 5% and 10%, respectively. Source: 
Authors using ERI-ESI data, 2017  
The average hourly wage of unionized workers is higher on average than that of their non-unionized 
counterparts. In addition, unionized employees appear to be older and more in the formal sector. On 
the other hand, women are less prone to unionization, as are single people and workers with no level of 
education or those with a primary level. It should be noted, however, that employees who have reached 
secondary and higher levels are more likely to join trade unions. In addition, we realize that the 
proportion of unionized workers decreases with seniority measured by duration in work. As the worker 
spends more time in the company, he or she is less likely to join a trade union organization. No doubt 
that the employee over time, manages to develop a relationship with the employer and has a mass of 
information about the company that reduces information asymmetries.  
5. Results and discussions 
The results presented are of two kinds. First, we analyze the determinants of union membership and 
wage compensation depending on whether the individual is unionized or not.  It should be noted that 
this result is intermediate which leads us, in a second step, to estimate the gain in remuneration 
associated with union membership that is the impact of membership of a trade union organization.  
5.1. Analysis of the determinants of union membership and compensation   
Table 2 presents the main econometric results for the effects of union membership on workers' 
compensation. The estimates lead to two main results. On the one hand, they make it possible to 
identify the main determinants of workers' trade union membership. The main factors explaining wage 
remuneration, depending on whether or not a trade union is affiliated to a trade union, are also 
highlighted.   
The results essentially show that industry, level of education, marital status, duration of employment 
or seniority and the existence of trade union organization in the enterprise affect the likelihood of union 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

86 | P a g e  

membership. The coefficient associated with the variable "existence of union organization" is positive 
and significant, supporting the idea that this variable is a valid instrument in this model. A first 
interesting result concerns the higher tendency of workers in the formal sector to join trade unions 
compared to their peers in the informal sector. Such a result is not surprising in view of the Beninese 
context where most unions recruit more in the formal sector, particularly the public sector, as we have 
shown in section 2 above.   
With regard to employment status, our results show that workers with more than 25 years' experience 
are less likely to join a trade union organization than their counterparts with less than 10 years' 
seniority.   One explanation for the higher probability of older workers joining a union may be explained 
by the desire of entrylevel workers to engage in organizations in order to improve their employment 
experience and fully enjoy their rights. The significant effect of job tenure is at odds with the results of 
Manda et al.  (2005). Manda et al. (2005) realizes that the probability of being unionized increases with 
seniority.  
Table 2: Regime change regression model estimation results  

Independent variables  

Dependent  variable: 
 Average hourly 
wage  

Selection 
equation   
  

Unaffiliated  Affiliated  
Formal (Ref.: Informal)  0.097  -0.364**  0.461***  
  (0.059)  (0.159)  (0.159)  
Public sector (Ref: Private sector)   0.186***  -0.012  0.062  
  (0.058)  (0.177)  (0.193)  
Female (Ref.: Male)  -0.001  0.047  0.001  
  (0.051)  (0.101)  (0.122)  
Log Age  0.121  -0.069  0.144  
  (0.088)  (0.203)  (0.208)  
Marital status (Ref.: Widowed/Divorced)        
Bachelor  0.206**  0.115  -0.095  
  (0.102)  (0.144)  (0.189)  
Polygamous married  0.240**  -0.374**  0.555***  
  (0.117)  (0.179)  (0.197)  
Monogamous married  0.206**  0.004  0.180  
  (0.100)  (0.108)  (0.146)  
Level of education (Ref, : None)        
Primary  0.126*  0.068  -0.151  
  (0.071)  (0.157)  (0.191)  
Secondary  0.147**  0.139  0.083  
  (0.062)  (0.109)  (0.137)  
Upper  0.143*  0.268**  0.017  
  (0.081)  (0.131)  (0.165)  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

87 | P a g e  

Length of service (Ref.: Less than 10 years)        
10 to 25 years  0.028  0.010  0.057  
  (0.057)  (0.097)  (0.116)  
Over 25 years  -0.315*  0.328  -0.505*  
  (0.162)  (0.232)  (0.282)  
Qualification (Ref.: Unskilled or semi-skilled 
employee)  

      

Qualified employee  0.440***  0.289**  -0.090  
  (0.052)  (0.127)  (0.121)  
Middle or senior management  0.911***  0.865***  0.083  

Independent variables  

Dependent  variable: 
 Average hourly 
wage  

Selection 
equation   
  

Unaffiliated  Affiliated  
  (0.072)  (0.127)  (0.153)  
Number of hours worked in the month  -0.006***  -0.005***    
  (0.000)  (0.000)    
Existence of trade union organization (Ref, : 
Yes)  

    1.396***  

      (0.281)  
Constant  6.156***  7.738***  -2.674***  
  (0.305)  (0.810)  (0.725)  

  0.711***  0.760**    
  (0.058)  (0.139)    
  0.860***  -0.966***    

  (0.196)  (0.281)    
Wald test of indep, Eqns, (chi2(2))  88.430    Prob  0.000  
Log pseudolikelihood  -1371.007    Pseudo R2    
N  1.135    

* p<0.1; ** p<0.05; p<0.01  

  
Source: Authors using ERI-ESI data, 2017  
With respect to the determinants of the average hourly wage, the estimation results show that the wage 
level of affiliates is significantly determined by the level of education. However, only the coefficients of 
the modalities related to secondary and higher education are significant at the 10 percent and 5 percent 
thresholds, respectively. For non-union members, we note that the coefficient increases with the level 
of education, corroborating the idea of an increase in the average private return to education 
highlighted by Manda et al (2005). Similarly, among affiliates, workers in the formal sector earn less 
than those in the informal sector in hourly terms. The difference can be explained by the absence of 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

88 | P a g e  

taxes in the informal sector affecting wage compensation. The same result is obtained for polygamous 
marriages, which appear to have a lower average hourly wage than widowed or divorced marriages.   
Not surprisingly, the level of qualification significantly affects the wage remuneration of workers for 
both affiliated and non-affiliated workers. It appears that the average hourly wage increases with the 
worker's qualification. Thus, laborers and other semi-skilled workers are paid less than middle or senior 
managers.   
5.2. Gain from unionization  
Table 3 shows the average effect of treatment on the Treaties, that is the difference in the average hourly 
wage of affiliated workers and their wage level if they had not been affiliated. The results show that the 
wage premium associated with unionization is positive and significant. In other words, workers' 
membership of trade unions allows them to increase their wage remuneration level by an average of 
33.06%.   
The positive impact of trade union membership highlighted confirms the predictions of previous work 
(Schmidt, 202-1; Ngom, 2021; Nahm et al., 2017). For example, Schmidt (202-1) shows that 
membership of a local union for expatriate workers results in a reduction in the wage gap between 
expatriate workers and national workers. Similarly, Nahm et al.  (2017) also estimate a positive gain 
associated with unionization. Using a gender analysis, the authors show that unionized men and women 
earn 12% and 18% more respectively than their nonunionized counterparts.    Maleka et al.  (2021) argue 
that union membership has a dampening effect on satisfaction levels. The explanation for this is that 
unions negotiate higher wages, which ultimately leads to greater job satisfaction.  
Table 3: Wage differential (en FCFA) by union affiliation or not  

  Affiliated   Unaffiliated   Effect of 
treatment   

Unionized workers   (a) 539,962   (b) 405,808   134,154***   
  

Non-unionized workers   (c) 1241,077   (d) 1119,158   121,919***   

Note: (a) and (d) represent the average hourly wage. (b) and (c), the expected counterfactual average 
hourly wage.   
Source: Authors, based on estimation results 
5.3. Analysis of robustness of results  
To test the sensitivity of results from regime-change models, we use propensity score matching (PSM) 
methods typically used in impact assessment studies. Impact analysis using MSP is preceded by the 
estimation of propensity scores for treatment variables from the probit model, the results of which are 
presented in Tableau A1 in the Appendix.  
Propensity scores, as Lee (2013) points out, are useful for balancing the distribution of observed 
covariates between treated and untreated groups. In general, most variables in the model have the 
expected signs. The figure 2 presents the distribution of estimated propensity scores and the region of 
common support. Visual analysis of density distributions for both groups reveals that all treated and 
untreated individuals are in the common support region. In other words, each individual had a positive 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

89 | P a g e  

probability of being either a union affiliate or a non-affiliate.  Thus, the common support hypothesis 
that requires each treated individual to have a corresponding untreated individual as a match, is 
satisfied. 

 
The results of the model of the impact of union membership on the level of wage compensation 
estimated with the Kernel Matching (KM) method, the Nearest Neighbor Matching (NNM) method and 
the radius matching (RM) method are presented in Table 4. All three matching methods indicate that 
union membership has a positive and significant impact on the hourly level of wage compensation. 
These results confirm those obtained from the regression method with regime change in the previous 
section.  They are also consistent with those obtained by Manda et al.  (2005), Bryson (2014) and Ngom 
(2021).  
Manda et al.  (2005) showed the existence of a positive effect of trade union membership on wage 
compensation. Bryson (2014) concludes that workers organized in unions benefit from higher wages 
called union wage premiums. For his part, Ngom (2021) also finds that employee compensation 
increases with the existence of a union in a company. In other words, unionized workers have higher 
wages than non-unionized workers. Our results also corroborate those found by Laroche (2004), 
Coutrot (1996) and, Najem and Paquet (2007). For example, Coutrot (1996) reported a 3% wage 
increase in organizations where there was at least one union representative. It thus confirms the 
existence of a causal link between the presence of a trade union in the undertaking and the level of 
salary. Najem and Paquet (2007) highlight the difference in hourly wages between unionized and non-
unionized employees in favour of the former. 
Table 4: Impact of Union Membership on Hourly Wages  

 
Matching methods  THAT  

Kernel Matching (KM)   0,164***   
(0,069)   

Nearest Neighbor Matching 
(NNM)   

0,092**   
(0,078)   

Radius Matching (RM)   0,326***   
(0,062)   

Figure 2:   Distribution of estimated propensity scores and common support region   

  
Source: Authors using ERI - ESI data, 2017   

0 .2 .4 .6 .8  
Propensity Score  

Untreated Treated  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

90 | P a g e  

                                                 Note: Standard errors are in parentheses. p<0.01; ** pp<0.05; * pp<0.1.                                                   
Source: Authors using ERI-ESI data, 2017 
Overall, it seems reasonable to conclude that the effects of trade unions on wages in Benin are positive. 
Such a result is very important in a Beninese context where the role of trade unions in corporate 
governance has been the subject of multiple reforms. Public authorities must consider labor 
organizations as privileged institutional partners in the service of the well-being of their union 
members. The advantages inherent in trade union action ultimately prove beneficial to the companies 
themselves because they are sources of motivation and commitment to work (Christian et al., 2011). 
However, the effect is not systematic and depends on several factors (Bryson, 2014). Ngom (2021) also 
notes for Senegal that productivity was relatively low in companies with a strong union presence. 
6. Conclusion  
The objective of this article is to empirically assess the impact of unionization of workers in Benin. To 
do this, we used a methodological approach that allows us to take into account the potential problems 
of endogeneity of the decision to join a workers' trade union organization and unobserved 
heterogeneity. The results show that the probability of union membership is significantly influenced by 
industry, level of education, marital status, duration of employment or seniority and existence of trade 
union organization in the enterprise. The level of wage compensation is determined by the sector of 
activity, the level of education and the level of qualification.  
More interestingly, unionization produces positive effects in terms of wage compensation. Concretely, 
we show that membership of a trade union organization allows the worker to increase his average 
hourly wage by about 33.06%. It is inferred that union membership promotes higher workers' wages. 
The existence of a wage premium associated with unionization suggests that trade unions play an 
important role in improving working conditions. In a Beninese context where wage levels are quasi-
rigid regardless of exogenous shocks that erode workers' purchasing power, it seems necessary for trade 
union organizations to fully invest their mission of negotiating with the State and private employers.  
The results obtained can help umbrella organizations anticipate behavioral responses by paying 
attention to salary levels and union membership. It can also influence collective bargaining processes.  
  
The union wage premium can have a positive impact on employers when they induce an increase in 
worker productivity through the worker selection mechanism, that is. if the best workers are attracted 
to wages above the market rate. Similarly, the positive effects of union membership benefit employers 
provided that workers increase the effort through efficiency wages. Firms can also benefit by becoming 
more capital-intensive in response to increases in the relative price of labor. In any case, this is an 
avenue of research to be explored in the context of developing countries. An additional dimension to be 
taken into account in this evaluation is the impact of union membership on individual employee 
satisfaction. Salary is a component of job satisfaction. Future analyses should investigate the potential 
effect on worker satisfaction. Such a line of research is interesting to explore because joining a union 
has a cost in terms of union dues and lost wages due to write-offs related to possible strikes.   
7. References  

Borland, J. (1996). Union effects on earnings dispersion in Australia. British Journal of Industrial 
Relations, vol. 34, no 2, p. 237-248.    

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

91 | P a g e  

Bryson, A. (2014), Union wage effects, IZA World of Labor.  

Budd, J. W., Chi, W., Wang, Y., & Xie, Q. (2014). What do unions in China do? Provincial-level evidence 
on wages, employment, productivity, and economic output. Journal of Labor Research, vol. 35, 
no 2, p. 185204.  

Card, D., Lemieux, T., & Riddell, W. C. (2004). Unions and wage inequality. Journal of Labor Research, 
vol. 25, no 4, p. 519-559.  

Christian, M. S., Garza, A. S., & Slaughter, J. E. (2011). Work engagement: A quantitative review and 
test of its relations with task and contextual performance. Personnel psychology, vol. 64, no 1, p. 
89-136.  

Coutrot, T. (1996). Relations sociales et performance économique : Une première analyse empirique du 
cas français : Les relations sociales en entreprise. Travail et emploi, vol. 66, p. 39-58.  

Danish Trade Union Development Agency (2021), Labor Market Profile Benin – 2021/2022, Labor 
Market Profile Benin – 2021/2022 (ulandssekretariatet.dk).  

Friedrich Ebert Stiftung (2015), Etude sur le paysage syndical au Bénin, p. 90, ISBN 978-99919-0-639-
3.  

Friedrich Ebert Stiftung (2019), Analyse de la dynamique de l’économie informelle au Bénin, p. 47, 
ISBN : 97899982-55-28-9.  

Freeman, R. B. (1993). Labor markets and institutions in economic development. The American 
Economic Review, vol. 83, no 2, p. 403-408.  

Freeman, R. B. (2010). Labor regulations, unions, and social protection in developing countries: Market 
distortions or efficient institutions? Handbook of development economics, vol. 5, p. 4657-4702.  

Freeman, R. B., & Lazear, E. P. (1995). An economic analysis of works councils. In Works councils: 
Consultation, representation, and cooperation in industrial relations. University of Chicago 
Press, p. 27-52.  

Freeman, R. B., & Medoff, J. L. (1984). What do unions do. Indus. & Lab. Rel. Rev., vol. 38, p. 244.  

Ge, Y. (2014). Do Chinese unions have “real” effects on employee compensation? Contemporary 
Economic Policy, vol. 32, no 1, p. 187-202.  

Guirkinger, C. and Boucher, S. R. (2008). Credit constraints and productivity in Peruvian agriculture. 
Agricultural Economics, vol. 39, no 3, p. 295 308.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

92 | P a g e  

Kerr, A., & Wittenberg, M. (2021). Union wage premia and wage inequality in South Africa. Economic 
Modelling, vol. 97, p. 255-271.  

Keune, M. (2018). Opportunity or threat? How trade union power and preferences shape occupational 
pensions. Social Policy & Administration, vol. 52, no 2, p. 463-476.  

Laroche, P. (2004). Présence syndicale et performance financière des entreprises : une analyse 
statistique sur le cas français. Finance Contrôle Stratégie, vol. 7, no 3, p. 117-146.  

Lee, L. (1978). Unionism and wage rates: A simultaneous equations model with qualitative and limited 
dependent variables. International Economic Review, vol. 19 no 2, p. 415-433.  

Lee, W. S. (2013). Propensity score matching and variations on the balancing test. Empirical economics, 
vol. 44, no 1, p. 47-80.  

Lewis, H. G. (1963). Unionism and relative wages in the United States: an empirical inquiry. Chicago: 
University of Chicago Press.  

Liu, J., Xing, C., & Ge, Y. (2020). Does union membership reduce gender earnings differentials? 
Evidence from employer–employee matched data in China. Pacific Economic Review, vol. 25, no 
1, p. 102-117.  

Lokshin, M. & Sajaia, Z. (2004). Maximum likelihood estimation of endogenous switching regression 
models. The Stata Journal, vol. 4, no 3, p. 282-289.  

Lu, Y., Tao, Z., & Wang, Y. (2010). Union effects on performance and employment relations: Evidence 
from China. China Economic Review, vol. 21, no 1, p. 202-210.  

Maddala, G. (1983). Limited-dependant and Qualitative Variables in Econometrics. Cambridge: 
Cambridge University Press.  

Maleka, M. J., Schultz, C. M., van Hoek, L., Paul-Dachapalli, L., & Ragadu, S. C. (2021). Union 
Membership as a Moderator in the Relationship Between Living Wage, Job Satisfaction and 
Employee Engagement. The Indian Journal of Labor Economics, vol. 64, no 3, p. 621-640.  

Manda, D. K., Bigsten, A., & Mwabu, G. (2005). Trade union membership and earnings in Kenyan 
manufacturing firms. Applied Economics, vol. 37, no 15, p. 1693-1704.  

Manning, A. (2013). Monopsony in motion ». In Monopsony in Motion, Princeton University Press.  

Nahm, D., Dobbie, M., & MacMillan, C. (2017). Union wage effects in Australia: An endogenous 
switching approach ». Applied Economics, vol. 49, no 39, p. 3927-3942.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

93 | P a g e  

Najem, E. & Paquet, R. (2007). L’impact syndical sur l’entreprise canadienne et sa main-d’œuvre. Revue 
internationale sur le travail et la société, vol. 5, no 3, p. 52–73.  

Ngom, A. (2021). Trade Unionism and Enterprise Performance in Senegal. AERC Research Paper, vol. 
420, p. 32.  

Périsse, M. (2014). Le droit du travail et les migrants ruraux : instituer un nouveau salariat en Chine. 
Revue de la régulation, Capitalisme, institutions, pouvoirs, vol. 15.  

Sandbrook, R. (1975). Proletarians and African capitalism: the Kenyan case, 1960-1972. Cambridge: 
Cambridge University Press, vol. 21.  

Schmidt, T. D. (2021). Local institutions, union wage effects and native–foreign wage gaps. Regional 
Studies, p. 114.  

Torm, N. (2018). Does union membership pay off? Evidence from Vietnamese SMEs. Micro, small, and 
medium enterprises in Vietnam. p. 230-252.  

Yedomon, B. (2016). Travail informel au Bénin : Expositions professionnelles et conséquences 
sanitaires chez les forgerons-ferblantiers à Cotonou (Doctoral dissertation, Université de 
Limoges) ».  

Zhang, Y., Chen, J., & Wong, P. (2011). Effect of trade unions on industrial labor income in China. Asian 
Politics & Policy, vol. 3 no 1, p. 95-114.  

Appendix  
Table A1: Proscript regression for estimating propensity scores  

Variables  Coefficient  

Dependent variable  Trade union 
membership  

Formula  0.690***  

  (0.241)  

Female (Ref. Man)  -0.208  

  (0.264)  

Level of education (Ref. None)    

Primary  -0.789**  

  (0.401)  

Secondary  -0.204  

  (0.292)  

Upper  -0.414  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

94 | P a g e  

  (0.304)  

Age  -0.028  

  (0.062)  

Age squared  0.000  

  (0.001)  

Marital status (Widowed/divorced ref.)    

Bachelor  -0.581  

  (0.453)  

Polygamous married  0.909**  

  (0.436)  

Monogamous married  0.227  

  (0.338)  

Union availability  5.342***  

  (0.629)  

Constant  -5.002***  

  (1.286)  

Observations  1,135  
Note: Standard errors are in parentheses. p<0.01; ** pp<0.05; * pp<0.1.  
Source: Authors using ERI-ESI data, 2017 

mailto:contact@americaserial.com
mailto:contact@americaserial.com

