Understanding Persistent Wage Disparities in Rural Colombia: Comparative Lessons from Latin America

Abstract

This research provides the first comprehensive analysis of the rural–urban wage gap in Colombia, with a focus on the coffee and cocoa sectors, over the past two decades. Using household survey microdata from 2001 to 2023 and international sources, we estimate wage differentials and apply econometric models—including Mincerian wage regressions and Blinder–Oaxaca decompositions. Results reveal a persistent and substantial wage gap: on average, rural coffee and cocoa workers earn roughly half as much as urban manufacturing workers. Even after controlling for education, experience, and other characteristics, a substantial share of the gap remains unexplained, indicating structural issues such as lower productivity, elevated levels of informality, and labor market segmentation in rural areas. Moreover, time-series evidence from the past two decades shows no significant convergence between rural and urban wages. Comparative analysis with Brazil, Mexico, and other Latin American countries highlights how policy interventions, such as rural social protection programs, labor formalization, and support for agricultural cooperatives, have helped narrow rural–urban wage disparities elsewhere. Drawing on these lessons, we discuss policy implications for Colombia and recommend measures to boost rural human capital, strengthen labor institutions, expand social safety nets, and promote rural economic development. These recommendations aim to gradually close the rural–urban wage divide, reduce rural poverty, and foster inclusive growth.

1. Introduction

Large income disparities between rural and urban areas pose a critical development challenge in Colombia. Wages earned in traditional rural sectors remain far below those in urban industries, contributing to higher poverty in the countryside and fueling rural-to-urban migration. Recent statistics underscore the severity of this divide. As of 2019, the rural poverty rate (about 34.5%) was almost three times the urban poverty rate (~12.3%), reflecting the low incomes and precarious livelihoods of rural households. Rural workers not only have fewer employment opportunities, but when employed, they tend to receive much lower wages than their urban counterparts.

This gap is especially pronounced in Colombia’s coffee and cocoa sectors—two cornerstone agricultural industries that support hundreds of thousands of small farmers and laborers. Coffee and cocoa are labor-intensive cash crops and vital exports for Colombia, yet the workers who cultivate and harvest these crops often earn incomes at or near subsistence levels. Field evidence illustrates this stark reality: a typical coffee picker’s daily wage ranges from only about COP 15,000 to 30,000 (USD 4–8) for 8–10 h of work, well below the statutory daily minimum wage for formal urban workers (~COP 42,000, including mandatory benefits). Cocoa farm labor shows similar patterns of low pay and informality, contributing to persistent poverty in cocoa-growing communities.

The wage gap between Colombia’s rural agricultural sectors and the urban industrial sector is not a mere statistical artifact; it has deep structural roots and significant implications for development. In theory, competitive labor markets would equalize wages across sectors once productivity differences and the cost of living are accounted for. In practice, however, many farmers in Colombia rely on verbal agreements, which sustain elevated levels of rural informality and limit access to formal private finance for farmers and their families. By contrast, urban workers are more likely to have formal written contracts governed by labor law. The absence of effective legal protections for rural workers contributes to persistently low pay, reinforcing poverty among workers in the cocoa and coffee subsectors. As a result, remuneration for agricultural activities remains extremely low. Colombia’s experience resembles classic dual-economy models in which a traditional rural sector characterized by surplus labor and low productivity coexists with a modern urban sector offering higher wages. According to the seminal Lewis (1954) model, a subsistence agricultural sector can sustain a large labor surplus, keeping rural wages low, while the urban sector must pay a premium to attract workers, leading to a persistent wage gap until surplus labor is absorbed. Colombia appears not to have reached the so-called Lewis turning point, as evidenced by continued rural out-migration and low farm wages (Nguyen The Kang 2025; Chen et al. 2021; ILO 2019; Moffitt 1983).

Decades of internal conflict have further complicated this picture: violence and insecurity in rural areas drove millions of Colombians to cities, effectively increasing the urban labor supply while depleting the countryside of human capital. This exodus creates a self-reinforcing cycle: meager rural wages push workers to migrate, but the loss of younger and more skilled labor can depress agricultural productivity, keeping rural incomes low. Meanwhile, urban labor regulations (such as relatively high minimum wages and better enforcement of labor standards in cities) create a wage floor in formal urban jobs that is out of reach for most informal rural jobs, sustaining a large earnings gap (Nguyen The Kang 2025; Ananian and Dellaferrera 2024; Chen et al. 2021; Fajgelbaum and Redding 2021; Franco and Ramos 2010).

Migration models help explain why the rural–urban wage disparity persists. The Harris–Todaro framework posits that migration decisions are based on expected income differentials: rural workers will migrate if the expected urban income (urban wage multiplied by the probability of finding a job) exceeds the rural income. Equilibrium is reached only when rural wages adjust such that expected urban earnings are no higher than rural earnings. A key implication is that urban wages can remain higher than rural wages even in equilibrium if urban unemployment or informal employment absorbs the excess labor inflow (Bryan and Morten 2019; Cuecuecha Mendoza et al. 2021; Zhao 2020; Wang and Yu 2018; Michaelsen and Haisken De-New 2011; Taylor and Martin 2001; Moffitt 1983; Harris and Todaro 1970; Todaro 1969).

In Colombia, the relatively high minimum wage in the formal sector and better social protections in urban areas have sustained a wage gap, even as high urban unemployment and informality persist; many migrants are willing to risk joblessness or take informal work in the city in the hope of eventually securing a higher-paying formal job. Indeed, empirical observations show that rural migrants often endure spells of urban unemployment or informal work before finding formal employment. Thus, the coexistence of a low-productivity rural sector and a constrained urban job market produces an equilibrium with a persistent wage gap. This mirrors the findings of Harris and Todaro (1970) as well as observations by Shaw (1974); Todaro (1969) in other developing contexts (Ananian and Dellaferrera 2024; Chen et al. 2021; IADB 2018; Harris and Todaro 1970; Todaro 1969; Lewis 1954).

In summary, a persistent wage gap exists between rural and urban workers in Colombia because of a combination of surplus labor in rural areas and institutional factors that favor urban workers. This means that even as Colombia develops, rural wages remain chronically lower than urban wages (Lewis dual-sector model and Harris–Todaro migration model) (Bryan and Morten 2019; Cuecuecha Mendoza et al. 2021; Zhao 2020; Wang and Yu 2018; Michaelsen and Haisken De-New 2011; Taylor and Martin 2001; Moffitt 1983; Harris and Todaro 1970; Todaro 1969).

This study examines that divide through the lens of Colombia’s coffee and cocoa sectors and a comparison to urban manufacturing. Focusing on coffee and cocoa is particularly insightful because these sectors are emblematic of rural Colombia: coffee has long been the country’s largest agricultural employer and export crop, and cocoa is an emerging sector with poverty-alleviation potential (often promoted as an alternative to illicit crops). Both sectors consist predominantly of smallholder farmers and seasonal laborers working outside the formal wage labor market. By comparing wages in these agricultural sectors to wages in urban manufacturing (including related industries like coffee processing and chocolate manufacturing), we can gauge the extent of labor market duality and the barriers preventing rural workers from contributing to overall economic progress (Nguyen The Kang 2025; Gáfaro and Pellegrina 2022; Adamopoulos et al. 2022; Adamopoulos and Restuccia 2020; Bryan and Morten 2019; Adamopoulos and Restuccia 2014; Ávila and Evenson 2010).

We also place Colombia’s wage gap in a broader international context. Latin America as a region has historically exhibited wide rural–urban income gaps, but some countries have made progress in narrowing these differences. Brazil is a particularly instructive comparison: as a major coffee (and cocoa) producer and Latin America’s largest economy, Brazil likewise had a large rural–urban wage chasm, yet over the past few decades, it implemented robust rural social protections (such as a rural pension scheme and Bolsa Família cash transfers) and saw significant growth in agribusiness productivity. These developments helped reduce poverty and narrow income gaps in rural Brazil (though significant regional disparities remain) (Ananian and Dellaferrera 2024; Gáfaro and Pellegrina 2022; Otero-Cortés and Acosta-Ariza 2022; IMF 2022; Chen et al. 2021; ILO 2019; Wang and Yu 2018; IADB 2018; Franco and Ramos 2010; Harris and Todaro 1970; Todaro 1969).

By comparing Colombia with Brazil—and drawing examples from Mexico and other Latin American countries—we aim to identify which differences in economic structure and policy account for variations in rural wage outcomes. Lessons from these countries can inform policy options for Colombia.

Organization of the paper: The remainder of this paper is organized as follows. Section 2 (Literature Review) presents relevant literature, including theoretical perspectives on rural–urban wage gaps and empirical findings from Colombia and comparable economies. Section 3 (Materials and Methods) describes our data sources, econometric models, and analytical approach. Section 4 (Results) presents the magnitude and decomposition of the wage gap and evidence on whether the gap has changed over time. Section 5 (Discussion) provides a comparative perspective, summarizing lessons from Brazil, Mexico, and other Latin American contexts. Section 6 (Policy Implications) discusses recommendations for closing the rural–urban wage gap in Colombia. Section 7 (Conclusions) summarizes the findings and offers suggestions for future research.

Purpose of the study. The purpose of this research is to examine the extent and causes of the rural–urban wage gap in Colombia, with a focused comparison between the rural coffee and cocoa sectors and the urban manufacturing sector. We aim to determine how much of the wage disparity can be explained by measurable factors and what remains unexplained.

Research questions. Specifically, this study addresses three key questions:

(1)

How large is the wage gap between rural coffee/cocoa workers and urban manufacturing workers, and has this gap changed over the past two decades?

(2)

To what extent can differences in worker characteristics (education, experience, etc.) explain the wage gap, and how much of the gap is due to other factors?

(3)

What policy lessons from other countries (e.g., Brazil, Mexico) can inform strategies to close the rural–urban wage gap in Colombia?

2. Literature Revision

2.1. Theoretical Perspectives: Rural–Urban Wage Gaps

The existence of a persistent wage gap between rural and urban sectors has long been explained by development economics theories. A foundational framework is Lewis’s dual-sector model (Lewis 1954), which portrays an economy divided into a traditional agrarian sector and a modern industrial sector. In Lewis’s model, the traditional rural sector is characterized by labor surplus and subsistence-level earnings, while the urban sector features higher productivity and wages that attract rural workers. Lewis famously posited that capitalist firms must pay a premium above the subsistence wage—about 30% higher—to entice workers to leave the farm (Economic Development with Unlimited Supplies of Labour). This creates an initial rural–urban wage gap by design. As industrialization proceeds and surplus rural labor is absorbed, the gap should eventually close at the “Lewis turning point,” when rural labor becomes scarce. However, until that point is reached, a significant wage differential can persist, reflecting the higher marginal productivity of labor in the urban sector. Colombia’s rural–urban dynamics closely echo this theory: the country has not yet reached a Lewis turning point, as evidenced by continued rural out-migration and low farm wages. Marginal productivity in smallholder coffee and cocoa farming remains low, which keeps rural wages at subsistence levels, whereas urban industries can pay more due to higher capital intensity and greater output per worker (Ananian and Dellaferrera 2024; Chen et al. 2021; Franco and Ramos 2010; Moffitt 1983; Lewis 1954; Harris and Todaro 1970; Todaro 1969).

Another seminal framework is the Harris–Todaro migration model (Harris and Todaro 1970), which explains rural–urban wage gaps through expected income differentials and migration equilibrium. The Harris–Todaro model assumes a fixed high wage in the urban formal sector (often due to minimum wage laws or unions) and a flexible wage in the rural sector. Rural workers will migrate to the city if the expected urban income (the formal wage multiplied by the probability of finding a formal job) exceeds the rural income. Equilibrium is reached only when rural wages adjust such that expected urban earnings are equalized with rural earnings. A key implication is that urban wages can remain higher than rural wages even in equilibrium, as long as urban unemployment or informality absorbs the excess labor. In other words, a rural wage gap may persist because it is offset by the risk of not securing a formal job in the city. This model is highly pertinent to the Colombian context. Colombia has a relatively high statutory minimum wage and stronger labor protection in formal urban employment, which has maintained a wage floor in cities. Meanwhile, there is a large informal sector in urban areas for those who cannot find formal jobs (Ananian and Dellaferrera 2024; Otero-Cortés and Acosta-Ariza 2022; Franco and Ramos 2010; Lewis 1954; Harris and Todaro 1970; Todaro 1969).

Rural workers tend to migrate when the chance of a higher urban wage raises their expected income, even if they must start out unemployed or in informal jobs. In Colombia, rural migrants face periods of urban unemployment or informality before moving into formal work. The Harris–Todaro perspective helps explain why simply raising rural wages via policy (or, conversely, raising urban wage floors) can have complex effects. If urban wages are pushed up without expanding employment, more migration may ensue, and if urban wages are somehow forced (e.g., through a high rural minimum wage), it could reduce rural employment or encourage mechanization, potentially exacerbating rural unemployment. Thus, the model underscores the importance of balancing job creation with wage policies. Beyond these classic models, other theories shed light on rural–urban wage gaps. Labor market segmentation theory argues that rural and urban labor markets may be institutionally distinct and imperfectly connected. Due to factors like mobility costs, information gaps, and social networks, workers might not freely arbitrage wage differences, resulting in segmented markets (Ananian and Dellaferrera 2024; Otero-Cortés and Acosta-Ariza 2022; Franco and Ramos 2010; Moffitt 1983; Rosenbaum and Rubin 1983; Harris and Todaro 1970; Todaro 1969; Lewis 1954).

In many developing countries, rural labor markets operate largely under informal arrangements (family labor, casual day jobs) with weak enforcement of national labor standards, whereas urban labor markets have greater formalization. This segmentation means that wages need not equalize across rural and urban areas, even for workers with similar skills, because the institutional context differs. In Colombia, for instance, rural workers often lack coverage by minimum wage laws in practice, have little access to collective bargaining, and rely on seasonal incomes, unlike urban workers. Institutional economists point out that legal frameworks (like land tenure and labor rights) and social norms can entrench a dual structure. Colombia’s case illustrates this: historically, labor laws like Estatuto del Trabajador Rural (Rural Worker Statute) were introduced much later or enforced weakly, keeping rural workers outside the formal protections that urban workers enjoy. Consequently, rural wages are determined in more informal, unregulated settings. Human capital theory also contributes to explaining wage differentials. On average, rural workers have lower educational attainment and skill levels than urban workers, which would lead to lower productivity and lower wages. If rural areas have inferior schools and fewer training opportunities, a skill gap emerges. This is indeed a factor in Colombia: rural adult populations have significantly fewer years of schooling on average, and illiteracy rates remain higher in rural zones. According to human capital theory, these differences should account for part of the earnings gap, as more educated workers command higher pay. However, as we will see in empirical studies, the human capital differences explain only a portion of Colombia’s rural–urban wage gap, suggesting that other structural factors are also at play (Ananian and Dellaferrera 2024; Haggblade et al. 2010; Otero-Cortés and Acosta-Ariza 2022; Gáfaro and Pellegrina 2022; Chen et al. 2021; Franco and Ramos 2010; Lewis 1954; Harris and Todaro 1970; Todaro 1969).

New economic geography and agglomeration theories introduce another angle: urban areas may pay a premium due to agglomeration benefits. Cities are concentrated in infrastructure, advanced services, and knowledge spillovers, boosting productivity for firms and workers. This urban wage premium is observed worldwide, including in Latin America, even after controlling for individual characteristics. In Colombia, large cities like Bogotá and Medellín are centers of finance, manufacturing, and services, where higher productivity and living costs both contribute to higher nominal wages. In contrast, rural coffee or cocoa towns may lack scale economies and infrastructure, limiting local productivity and wages. Migration of ambitious workers to cities can also lead to a selection effect: those who remain in rural areas have disproportionately fewer skills or opportunities, which could widen wage disparities (Ananian and Dellaferrera 2024; Haggblade et al. 2010; Otero-Cortés and Acosta-Ariza 2022; Gáfaro and Pellegrina 2022; Chen et al. 2021; ILO 2019; Franco and Ramos 2010; Lewis 1954; Harris and Todaro 1970; Todaro 1969).

In summary, multiple theoretical frameworks consistently predict a rural–urban wage gap in developing economies and help interpret why it persists. The dual-sector and migration models provide a macrostructural view: until substantial structural change occurs (industrial absorption of labor or major productivity gains in agriculture), a wage gap is a natural outcome. Labor market segmentation and institutional factors explain why market forces alone may not eliminate the gap: mobility barriers to mobility and unequal enforcement of standards allow divergent wage equilibria. Human capital differences and urban agglomeration further entrench the local wage gap (Ananian and Dellaferrera 2024; Haggblade et al. 2010; Otero-Cortés and Acosta-Ariza 2022; Gáfaro and Pellegrina 2022; Chen et al. 2021; Moffitt 1983; Rosenbaum and Rubin 1983; Harris and Todaro 1970; Todaro 1969; Lewis 1954).

2.2. Empirical Studies on Rural–Urban Wage Gaps

A considerable body of empirical literature has documented rural–urban wage disparities in Colombia and comparable developing economies. These studies typically confirm large wage differentials and attempt to decompose the gap into explained and unexplained components. In Colombia, research dating back decades notes the persistent income gap between rural and urban workers. Franco and Ramos (2010) analyzed rural–urban earnings differentials using household survey data from 2002 to 2009. Using matching and decomposition methods, they compared workers with similar observable characteristics. Even after controlling for education, age, and occupation, a large rural–urban earnings gap persisted: about fourteen percentage points of the 50% differential were “unexplained” by standard human-capital variables (Ananian and Dellaferrera 2024; Li et al. 2024; Gáfaro and Pellegrina 2022; ILO 2019; Franco and Ramos 2010).

This implies that only a portion of the rural penalty could be attributed to measurable differences such as lower schooling or experience, while a sizeable residual gap hinted at structural or discriminatory factors. Notably, their time-series analysis found no significant divergence or convergence trend over the period. The rural–urban wage ratio remained stable, indicating that the wage gap was entrenched and did not automatically shrink with economic growth (Otero-Cortés and Acosta-Ariza 2022; Anderson et al. 2013; Delgado et al. 1998). These results underscore the need to look beyond individual worker attributes to understand why rural Colombians earn so much less. Recent studies have reinforced and expanded this knowledge. Otero-Cortés and Acosta-Ariza (2022) examine urban and rural labor income inequality in Colombia for the period 2010–2019, using the Blinder–Oaxaca decomposition techniques (Otero-Cortés and Acosta-Ariza 2022).

They report that around 60% of the wage gap can be explained by observable characteristics of workers—chiefly educational level, age, and job sector—while 40% remains unexplained. The explained portion reflects the fact that rural workers in Colombia generally have lower human capital: for instance, rural workers on average have fewer years of schooling (often only primary education) and are older, since many young people migrate out. These factors naturally lead to wage differences. However, the unexplained 40% suggests other forces at work, consistent with segmentation or institutional exclusion. Otero and Acosta highlight three salient features of Colombia’s rural labor market that perpetuate the wage gap: (i) low female labor force participation and higher unemployment in rural areas relative to urban areas (indicating that many rural women receive no wage at all, and those seeking work face scarcity of jobs); (ii) significantly higher informality in rural employment (most rural workers lack formal contracts, social benefits, or minimum wage coverage); and (iii) a greater incidence of child labor in rural areas (which both reflects and reinforces poverty) (Otero-Cortés and Acosta-Ariza 2022).

These conditions contribute to lowering average incomes and bargaining power for rural workers. The labor market in rural Colombia functions very differently from the urban formal labor market, leading to a persistent wage gap that is not fully justified by productivity or skills alone. Studies focusing specifically on the coffee sector have provided additional context. For example, Leibovich et al. (2007) (as referenced in the Colombian economic literature) examined wages in coffee-growing regions during the 1990s and 2000s. They observed that Colombia’s uniform national minimum wage policy had very uneven effects: in urban areas, the minimum wage often sets a floor for the lowest-skilled formal jobs, but in rural coffee regions, many farm workers earn below the legal minimum because enforcement is weak (Ananian and Dellaferrera 2024; Murialdo 2023; Chen et al. 2021; Franco and Ramos 2010; Rosenbaum and Rubin 1983).

Their work suggested that even when coffee prices rose and farm productivity improved, rural wages did not rise proportionally, partly because of surplus labor and the absence of bargaining mechanisms, which kept agricultural wages low. Another study on coffee by Hataya (1992) (an urban–rural linkages analysis) found that the development of the coffee sector increased demand for labor, but much of it was met by temporary or migrant workers, and rural wages remained tied to the volatile coffee price cycle rather than steadily converging to urban levels (Nguyen The Kang 2025; Otero-Cortés and Acosta-Ariza 2022; Franco and Ramos 2010). This implies that while booms in commodity prices (like coffee) can boost rural incomes in the short run, they often do not permanently alter the structural wage gap unless accompanied by improvements in productivity and institutions (Otero-Cortés and Acosta-Ariza 2022; Franco and Ramos 2010). Internationally, cross-country research by organizations like the ILO finds that Colombia’s experience is common across the developing world. A recent ILO working paper (Ananian and Dellaferrera 2024) compiled data from fifty-eight countries on rural–urban wage differences. They found that in developing countries, rural workers earn significantly less, often with a wider gap than the global average 24% hourly pay gap between urban and rural workers (Ananian and Dellaferrera 2024; Murialdo 2023; Otero-Cortés and Acosta-Ariza 2022; Benguria et al. 2021; Chen et al. 2021; Franco and Ramos 2010).

Importantly, in their global sample, only about half of the wage gap could be explained by differences in education, experience, or occupation, while the remaining half was unexplained. This aligns closely with the Colombian studies, reinforcing that unobservable or structural factors (such as location-based disadvantages, access to infrastructure, network effects, or outright discrimination) are responsible for a large part of rural wage penalties. The ILO paper also noted that rural women often face a double disadvantage, earning less than rural men, thereby compounding both gender and location gaps. For our purpose, the key takeaway is that rural–urban wage gaps are pervasive and tend to persist even after accounting for worker characteristics, suggesting that policies targeting those structural factors are necessary to bridge the divide (Ananian and Dellaferrera 2024; Murialdo 2023; Otero-Cortés and Acosta-Ariza 2022; Rosenbaum and Rubin 1983).

Case studies from other Latin American countries provide comparative insights. During the commodity boom in the 2000s, many Latin American economies experienced rising rural incomes due to high agricultural and mineral prices, which, in some cases, narrowed rural–urban wage gaps (Nguyen The Kang 2025; World Bank 2017; Gáfaro and Pellegrina 2022; Otero-Cortés and Acosta-Ariza 2022). A World Bank analysis of wage inequality in Latin America noted that the urban–rural wage gap in the region shrank significantly between 2002 and 2013, from an average 46% urban wage premium to about 25%. This compression was especially pronounced in South American countries that benefited from the boom (e.g., Argentina, Brazil, Peru), where increased demand for rural labor and higher export revenues lifted rural wages (Takanohashi et al. 2025; Li et al. 2024; Xi 2023; Benguria et al. 2021; World Bank 2008, 2017).

However, even after this improvement, substantial gaps remained, and in some countries (Central America, Mexico), the narrowing was modest. Rodríguez Castelán et al. (2016) concluded that only a few countries (Colombia, El Salvador, Paraguay) showed a notable reduction in the rural–urban gap in the 1990s, and many more experienced improvements after 2003 with the commodity boom. This suggests that external economic conditions can influence the gap, but sustained progress often requires domestic internal policy changes and structural transformation (Adamopoulos and Restuccia 2014, 2020; World Bank 2017). In Brazil, extensive research documents the historical evolution of the rural–urban wage gap. During Brazil’s period of rapid economic growth and structural change (1960s–1980s), rural wages rose markedly. One World Bank report highlighted that the gap between urban and rural unskilled wages in Brazil fell from 56% in 1968 to just 16% in 1977, a dramatic convergence (Benguria et al. 2021; Delgado et al. 1998; Barro and Sala-i-Martin 1992; Barro 1991).

This was attributed to a combination of factors: robust industrial job creation in cities drawing workers out of rural areas (thus tightening rural labor supply), government policies that improved rural incomes (including partial enforcement of wage regulations and rural extension services), and migration that reduced labor surplus. However, Brazil’s experience also shows that convergence can stall or reverse if conditions change (World Bank 2017, 2020, 2021, 2022, 2023). In the 1980s and 1990s, economic crises and adjustment policies in Brazil saw less focus on rural support, and gaps in the poorest regions (e.g., the Northeast) remained high. By the early 2000s, Brazil still had significant spatial inequalities—for instance, one source noted that in the Northeast, 44% of rural dwellers were poor compared to 7% of urban residents in the richer southern states. The literature on Brazil underscores how interventions like rural pensions and minimum wage policies (discussed further in the comparative section) helped reduce poverty and indirectly support rural wages, even if not always closing the direct wage gap (Ananian and Dellaferrera 2024; World Bank 2017, 2023; Anderson et al. 2013).

Finally, it is worth mentioning the empirical studies on migration and labor mobility in Colombia. Research has found that migration flows respond to wage differentials in expected ways: rural Colombian households are more likely to send a member to migrate if the expected urban–rural wage ratio is high, controlling for other factors. However, studies also point out that migration is selective—younger, more educated individuals are disproportionately likely to leave rural areas (often worsening the average human capital in villages). Colombian scholars (e.g., at DANE or in academia) have applied the Harris–Todaro model to domestic data and found it to be consistent with the observed patterns. Additionally, Colombia’s armed conflict created a unique form of forced migration (internal displacement, people moving to cities), which, in the 1990s and 2000s, added to rural depopulation; yet many of these migrants lacked the skills for formal urban employment, often ending up in urban informal sectors or unemployed. This dynamic arguably kept a lid on urban wage growth at the bottom end (due to a larger pool of low-skilled labor) while rural areas lost labor without commensurate wage gains, because remaining rural jobs were still low in productivity. Empirical works by analysts at the Banco de la República and think tanks like Fedesarrollo have suggested that conflict-related migration temporarily tightened rural labor supply in some areas, but not enough to equalize wages, given the low productivity and weak bargaining position of those who stayed. These nuanced findings reinforce that purely economic drivers of the wage gap intertwine with historical and social factors in Colombia (Chen et al. 2021; Delgado et al. 1998; Moffitt 1983).

Literature gaps: While many studies confirm the rural–urban wage disparity, there are gaps that this study aims to fill. First, there is little focus on specific agricultural sub-sectors (like coffee and cacao) in the context of wage gaps; most analyses treat rural areas monolithically. By zooming into coffee and cocoa, we consider whether cash–crop sectors that are connected to global markets behave differently in terms of wage convergence or divergence compared to general rural employment. Second, the existing Colombian literature has analyzed wages within Colombia; a direct comparison with Brazil’s experience (or other countries) has been missing. This study contributes a comparative perspective, asking what Colombia might learn from Brazil’s relative success in raising rural incomes through policy. Third, our approach integrates both econometric analysis and policy analysis. Much of the literature is either purely empirical (identifying the gap and its proximate causes) or purely narrative/policy-oriented. We bridge these by using econometric results to inform the policy discussion. In the process, we also incorporate factors like informality, social protection, and migration more explicitly into the analysis of wage disparities. These factors are often acknowledged qualitatively but not directly measured in wage gap decompositions. Our study brings them to the forefront by discussing how Colombia’s lack of rural social security coverage, high informality, and labor market segmentation contribute to wage differences, and by drawing lessons from cases where these institutional factors were addressed (Nguyen The Kang 2025; Ananian and Dellaferrera 2024; Gáfaro and Pellegrina 2022; ILO 2019; IADB 2018; Otero-Cortés and Acosta-Ariza 2022; Franco and Ramos 2010).

3. Materials and Methods

3.1. Data and Methodology

3.1.1. Data and Sample

Our primary data source for Colombia is the Gran Encuesta Integrada de Hogares (GEIH), the nationwide household and labor force survey conducted by the National Administrative Department of Statistics (DANE). We utilize microdata from the GEIH spanning the years 2001 through 2023, which provides individual-level information on employment, wages, education, and other demographics. From these surveys, we extract a sample of wage-earning workers in three groups of interest: (1) rural coffee sector workers, (2) rural cocoa sector workers, and (3) urban manufacturing workers.

We define the rural coffee and cocoa sector workers as individuals living in rural areas who are employed in either coffee cultivation or cocoa cultivation/harvesting activities (including both hired laborers and sharecroppers on coffee/cocoa farms). The urban manufacturing group serves as a benchmark for higher-productivity urban employment; it includes workers in manufacturing industries, and we pay particular attention to those in sectors related to coffee and cocoa processing (such as coffee processing factories, food manufacturing, and chocolate/confectionery manufacturing) to allow a closer comparison of industries along the value chain. By comparing rural coffee/cocoa farm wages to urban manufacturing wages (especially in related agro-industries), we attempt an “apples-to-apples” comparison insofar as all workers are linked to the coffee/cocoa commodity chain but differ by location and sector (farm vs. factory) (Benguria et al. 2021; Chen et al. 2021; FAO 2020; Kadjo et al. 2019; Chowdhury et al. 2005; Hobbs and Young 2000).

We restrict the sample to prime-age workers (15–65 years) to focus on the core workforce. We exclude self-employed farmers to concentrate on wage earners. In rural Colombia, many agricultural workers are informal and do not report wages; contracts in the coffee and cocoa subsectors are predominantly verbal, and there are no robust legal protections for agricultural workers’ social rights. As a result, pay for these activities is often very low. Because our study examines wage gaps between rural and urban wage workers, we limit the analysis to individuals who report wages and exclude those with no reported rural remuneration. Our analysis, therefore, mainly captures hired labor on farms (including day laborers, known as jornaleros). Wage data are converted to real terms (inflation-adjusted to constant 2020 COP) to analyze trends over time. For cross-sectional comparisons, we often use the latest available year (e.g., 2023) to illustrate the current gap, as well as pooled data for the 2002–2023 period to estimate long-run average differences. (DANE—GEIH—Sipsa system).

For comparative analysis, we compile secondary data from several sources. We use the International Labour Organization (ILO) online database (ILOSTAT) and the World Bank’s World Development Indicators (WDI) for country-level metrics, including average wages by sector, rural versus urban income levels, poverty rates, and informality rates in Brazil, Mexico, and select other Latin American countries. For Brazil, we utilize data from its National Household Survey (PNAD) to compute rural and urban wage differentials similar to our Colombian analysis. We also draw on published papers, reports, and statistical yearbooks for information on historical trends (for instance, if detailed time-series data on Brazil’s rural–urban wage gap in earlier decades are needed, we reference World Bank reports and academic studies). The comparative data serve to place Colombia’s situation in context and are not used for new econometric estimations; rather, they inform a qualitative assessment of how distinct factors (like social programs or labor policies) have impacted rural wages in those countries (World Bank 2021, 2022, 2023).

Artificial intelligence assistance (ChatGPT-5, OpenAI) was used to improve the translation and language editing of the manuscript from preliminary versions drafted in Spanish. All content was subsequently reviewed and validated by the authors.

3.1.2. Econometric Approach

We implement a combination of descriptive analysis, Mincerian wage regressions, Blinder–Oaxaca decompositions, and convergence tests to quantify the rural–urban wage gap and investigate its underlying determinants systematically:

1. Descriptive Analysis of Wage Gaps: We first calculate the raw rural–urban wage gap without any controls. This involves computing mean (and median) wages for the three groups: rural coffee workers, rural cocoa workers, and urban manufacturing workers. We express the gaps in relative terms for interpretability—for example, we calculate the rural coffee wage as a percentage of the urban manufacturing wage. This provides a baseline measure of the wage disparity. We also examine the distribution of wages (to see if the gap persists across the wage distribution or is concentrated among certain percentiles). In addition, we break down the wage gap by year to assess trends: for each year 2001–2023, we compute the average rural and urban wages and the rural/urban ratio. This helps determine whether the gap has narrowed, widened, or remained stable over time. We further disaggregate by region, when possible (comparing, for instance, coffee-growing regions like Antioquia or Caldas to major urban centers), to check for within-country spatial variation. This descriptive step sets the stage by confirming the magnitude of the wage gap and providing a first look at its evolution (Allen and Atkin 2015; Allen 2014; Franco and Ramos 2010).

2. Mincerian Wage Regressions: To adjust for differences in worker characteristics between rural and urban samples, we estimate earning functions of the Mincer (1974) form. The basic regression model is as follows:

ln(wage_i) = β₀ + β₁*Agriculture + β₂*Age (experience rural/urban) + β₃*Education + β₄*Gender +
β₅*Children + β₆*Contract + β₇*Owner place + ε

(1)

where [ln(wage_i)] is the natural logarithm of the hourly (or monthly) wage of individual ‘i.’ Agriculture is a dummy variable equal to ‘1’ if the individual is a rural coffee/cocoa worker and ‘0’ if the individual is an urban manufacturing worker. The coefficient ‘beta’ on this dummy captures the ceteris paribus wage difference between rural and urban workers. The other variables are age, education, gender, number of children, contract type, and ownership status. These form the vector of control variables, including years of schooling (or highest educational attainment), potential labor-market experience (proxied by age and age squared), gender, and other relevant characteristics—for example, whether the job is informal and whether the worker has a permanent or temporary contract.

The error term ε captures unobserved factors. We estimate these regressions on the pooled sample as well as for specific years or sub-samples, using ordinary least squares (OLS). Standard errors are clustered at an appropriate level (such as the municipality or region level) to account for spatial autocorrelation in wages. The key parameter of interest is β: a significantly negative β after controlling for Education and Age would indicate a structural “rural penalty”—i.e., rural coffee/cocoa workers earn less than observationally similar urban workers, suggesting the presence of factors beyond human capital that depress rural wages. Conversely, if β is greatly reduced in magnitude relative to the raw gap, this implies that much of the gap is due to differences in education, experience, and other observable characteristics (Franco and Ramos 2010).

As another way to adjust for differences in worker characteristics between rural and urban samples, we estimate earnings functions of the Mincer (1974) type. The regression model with double logarithmic methodology is as follows:

log(wage_i) = β₀ + log(β₁*Agriculture) + log[(β₂*Age (experience rural/urban)] + log(β₃*Education) +
log(β₄*Gender) + log(β₅*Children) + log(β₆*Contract) + ln(β₇*Owner place) + ε

(2)

We run separate specifications to test robustness. In one variant, we restrict the urban sample to low-skill manufacturing workers (those with similar education levels as the rural workers) to ensure a more like-for-like comparison. In another, we include additional controls such as occupation dummies or region fixed effects. We also test interactions (for example, interacting the rural dummy with a time trend to see if the controlled gap has changed over time). These regressions help identify how much of the wage gap can be “explained” by observable traits. If the R² of the model is low and β remains large and significant, it indicates that unobserved or structural factors are important. Indeed, in our results, we find that even with controls, a sizeable rural–urban wage differential persists (see Section 4) (Franco and Ramos 2010).

3. Blinder–Oaxaca Decomposition: To quantify the portion of the wage gap due to differences in characteristics versus differences in returns to those characteristics, we perform a Blinder–Oaxaca decomposition (Blinder 1973; Oaxaca 1973). This technique decomposes the mean wage difference between two groups (urban vs. rural) into an “explained” component and an “unexplained” component. The explained component reflects the part of the gap attributable to group differences in average endowments (e.g., rural workers having lower average education, less experience, different job types). The unexplained component reflects differences in the returns to those endowments (e.g., the same level of education yielding lower pay in rural areas) as well as any other structural or unmeasured factors (often interpreted as a combination of discrimination, segmentation, or omitted variables like productivity differences). In practice, we estimate separate wage equations for the two groups and then use the decomposition formula to split the mean log wage gap. We include in the decomposition the key covariates such as education, age/experience, gender, job informality, and region. This allows us to see, for instance, how much of the wage gap is statistically accounted for by rural workers’ lower education levels or the higher incidence of informal (lower-paid) jobs in rural areas. We perform the decomposition for coffee vs. urban and cocoa vs. urban comparisons separately, given that the sectors may differ. Additionally, we apply pooled specifications to the combined rural and urban samples. We note that the decomposition has an output that is sensitive to the choice of reference coefficients (we use the urban sector as the reference). We report the percentage of the gap that is explained and unexplained, and we discuss which variables contribute most to the explained portion (Chen et al. 2021; Franco and Ramos 2010; Blinder 1973; Oaxaca 1973).

4. Results

4.1. Descriptive Findings: Magnitude and Trends of the Wage Gap

Our analysis confirms a large raw wage gap between rural coffee/cocoa workers and urban manufacturing workers in Colombia. Over the full 2001–2023 period, rural agricultural wages have averaged roughly 50% or less of urban manufacturing wages. In the most recent data (early 2020s), the gap remains striking: for example, in 2023, the average monthly wage of a rural coffee farm worker was approximately COP 600,000, compared to about COP 1,200,000 for the average urban manufacturing worker, indicating that the rural worker earned only about 50% of the urban worker’s wage on average. A similar ratio is observed for cocoa farm workers. Figure 1 illustrates the trend in average wages for rural coffee/cocoa workers versus urban manufacturing workers from 2001 to 2023. The rural–urban wage ratio has fluctuated between 0.4 and 0.6 (40% to 60%) over this period, without a clear upward trajectory.

In other words, we do not observe meaningful convergence of rural wages toward urban wages over time: the gap appears entrenched. In some years, the gap even widened: for instance, during periods of low global coffee prices (which depressed farm revenues) or when urban wages were boosted by inflation-indexed minimum wage hikes, the rural wage fell to under 40% of the urban wage. On the other hand, there were brief episodes (e.g., mid-2000s commodity boom) when rural wages rose and the gap narrowed slightly, but these gains were not sustained. Statistical tests reinforce this visual impression: we found that the rural–urban wage ratio series is not trend-stationary. A unit root test on the log wage ratio failed to reject the null hypothesis of a unit root, implying no mean reversion or deterministic trend in the gap. Furthermore, a simple convergence regression yielded an estimated φ (convergence coefficient) that was statistically indistinguishable from zero (and even slightly positive in some specifications, though not significant), indicating no tendency for the gap to diminish systematically. Thus, the wage disparity has persisted over two decades of economic growth and structural change in Colombia (Benguria et al. 2021; Franco and Ramos 2010; Taylor and Martin 2001; Harris and Todaro 1970; Delgado et al. 1998; Todaro 1969).

To put the magnitude in perspective, consider that a typical rural coffee picker earns around COP 20,000 per day, whereas a worker in a coffee processing plant in town might earn on the order of COP 40,000–50,000 per day (including benefits). These differences compound to large annual income gaps. The implications are reflected in poverty statistics: despite overall poverty reduction in Colombia in recent decades, rural poverty remains roughly double or triple urban poverty, largely because of depressed rural earnings. Our findings are consistent with national statistics and prior studies that found a sizable wage premium for urban jobs. For example, an analysis by the ILO found that globally rural workers earn 24% less on average, with only half the gap explained by observable factors—our Colombia-specific figure of roughly a 50% shortfall is even larger, highlighting the severity in this country (Ananian and Dellaferrera 2024; ILO 2012, 2019, 2020; Haggblade et al. 2010).

In terms of within-country variation, we found that the wage gap is narrower in certain regions (for instance, rural wages in coffee zones closer to major cities) and wider in more remote regions (like parts of Cauca or Nariño). However, the gap exists in all regions. We also found that the gap is present at different points of the wage distribution: even comparing median wages or the 25th percentile, urban workers earn significantly more than rural workers. This suggests the result is not driven solely by a few high-earning urban workers, but a broad-based disparity. One notable difference is that informality rates are much higher in rural areas. Over 80% of rural workers in Colombia are informal (lacking formal contracts and social security), compared to around 50% in urban areas. Informal jobs generally pay less and lack protection, a crucial factor to which we return in the decomposition analysis (Ananian and Dellaferrera 2024; Murialdo 2023; Otero-Cortés and Acosta-Ariza 2022; Gáfaro and Pellegrina 2022; Franco and Ramos 2010).

Table 1. Average wages by sector (Colombia, 2023). Note: Illustrative values; 1 USD = 4.000 COP. Rural coffee/cocoa wages are roughly half of urban manufacturing wages.

Sector & Location	Average Hourly Wage (COP)	Rural Wage as % of Urban
Urban Manufacturing Workers	5.000	-
Rural Coffee Farm Workers	2.500	50%
Rural Cocoa Farm Workers	2.300	46%

summarizes key wage statistics for the rural and urban groups (placeholder for an embedded table). Table 1 (below) shows the average hourly wage for each group in COP, along with the rural/urban percentage:

As shown in Table 1, rural coffee workers earn, on average, about half of what urban manufacturing laborers earn. Cocoa sector wages are remarkably similar (if anything, slightly lower on average), reflecting the fact that cocoa is a smaller sector with less institutional support than coffee. These gaps are stark, and they persist even when considering daily or monthly earnings. Next, we move beyond these raw figures to examine how much of this gap can be accounted for by differences in worker characteristics.

4.2. Regression Results: Controlled Wage Differentials

We estimated Mincerian wage regressions to control for education, experience, gender, and other factors. The regression, including basic controls, yields a large and statistically significant coefficient on the rural dummy variable. Specifically, in a pooled regression for 2023, controlling for an individual’s years of education, age (as a proxy for experience), gender, and whether their job is informal, we find that rural coffee/cocoa workers earn about 30–35% lower wages than comparable urban workers, on average. In logarithmic wage terms, the coefficient β on the rural indicator is approximately −0.35 (significant at the 1% level), implying rural workers have wages about 30% lower (ε{−0.35}\approx. USD 0.70) than observationally similar urban manufacturing workers. This controlled gap is slightly smaller than the raw gap (which was about 50% lower wages, corresponding to a log difference of −0.70), but it remains quite substantial. The fact that a large gap persists after accounting for education and experience suggests that differences in human capital explain only part of the disparity (Aminur 2025; Li et al. 2024; Mincer 1974).

The regression results indicate that education is a significant predictor of wages and that rural workers indeed have lower educational attainment on average. This contributes to the wage gap. However, the rural dummy’s coefficient remains large even with education in the model, meaning that even equally educated workers earn less in rural areas. We also included a dummy for informal employment, which was strongly negative (informal workers earn significantly less, other things equal). Since most rural workers are informal, this variable captures a lot of the rural effect. In fact, when we add a control for job informality, the absolute size of the rural coefficient drops (in one specification it fell from about −0.5 to −0.3), indicating that part of the rural penalty operates through the informality channel. This aligns with the idea that the rural labor market’s informal nature is a key factor in the wage gap. Other controls like age (experience) showed a concave relationship (earnings rise with experience at a decreasing rate) and did not differ drastically between rural and urban samples; interestingly, the rural workforce has a slightly older average age (many young people have migrated out), which, if anything, reduced the explained gap because older workers usually earn a bit more.

It is worth noting that the regression’s overall explanatory power is modest—the R² for the pooled 2023 sample regression is around 0.10 (10%). This is not unusual for wage regressions, as there is substantial unobserved heterogeneity in wages, but it underscores that much of the wage variation (and by extension, much of the rural–urban gap) is not captured by just education and basic demographics. Our findings here mirror those of Ñopo et al. (2012) in a study of wage gaps in 64 countries. Controlling for observable characteristics narrows wage gaps but leaves a large unexplained portion. In our case, the unexplained rural wage penalty is strongly suggestive of structural factors. This motivates the use of the Blinder–Oaxaca decomposition to quantify the explained vs. unexplained shares more formally (Aminur 2025; Li et al. 2024; Ananian and Dellaferrera 2024; Murialdo 2023; Otero-Cortés and Acosta-Ariza 2022; Gáfaro and Pellegrina 2022; Franco and Ramos 2010; Mincer 1974).

Four OLS linear models of hourly wages were estimated (

Table 2. OLS regression of hourly wage (linear model).

Variable	Model 1	Model 2	Model 3	Model 4
Intercept (COP)	8781.75 ***	4597.04 ***	4333.96 ***	4243.73 ***
Agricultural sector (1 if rural coffee/cocoa worker)	−3684.61 ***	−1260.31 ***	−1162.41 ***	−1146.77 ***
Age (years)		24.33 ***	24.66 ***	20.59 ***
Education (years)		108.16 ***	107.71 ***	104.79 ***
Female (1 = yes)		−464.05 **	−447.81 **	−423.69 **
Children < 6 yrs		54.46	34.72	–
Has formal contract (1 = yes)		254.76 *	528.68 ***	–
Owns land/farm (1 = yes)		867.86 ***	–	–
R2	0.0383	0.0197	0.0204	0.0236
Observations	5236	5057	5057	5057

Note: Dependent variable is hourly wage (Colombian pesos-COP). Models (1)–(4) add controls sequentially as described in the text (Model 1 is baseline; later models include age, education, gender, children, contract status, and land ownership dummies). Significance: *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Author’s calculations from GEIH/DANE (2023) data.

). The agricultural-sector indicator (rural coffee/cocoa worker) is negative and highly significant across all specifications; the baseline gap is −3684 COP and narrows to about −1147 COP after controls are added. Age and education are positively associated with wages; the female indicator is associated with lower wages; formal contracts and land/farm ownership are positively associated with wages; and having children under six is not statistically significant in the fuller models. The R² values are low (≈0.02–0.04), as is common in micro wage regressions, yet signs and magnitudes remain stable across specifications, supporting the robustness of the main results.

In addition to the four linear regression models, four double-logarithmic (log-linear) models were estimated, yielding low R² values (

Table 3. OLS regression of log hourly wage (log model).

Variable	Model 1 (log)	Model 2 (log)	Model 3 (log)	Model 4 (log)
Intercept	8.8701 ***	8.5458 ***	8.3591 ***	8.3567 ***
Agricultural sector (1 if rural coffee/cocoa worker)	−0.4911 ***	−0.3308 ***	−0.2509 ***	−0.2505 ***
Age (years)		0.0021 ***	0.0021 ***	0.0020 ***
Education (years)		0.0139 ***	0.0135 ***	0.0134 ***
Female (1 = yes)		−0.1798 ***	−0.1649 ***	−0.1643 ***
Children < 6 yrs		0.0047	0.0041	–
Has formal contract (1 = yes)		0.1975 ***	0.2050 ***	–
Owns land/farm (1 = yes)		0.0236	–	–
R2	0.0790	0.0635	0.0996	0.0998
Observations	5236	5057	5057	5057

Note: Dependent variable is ln (hourly wage). Other notes and significance are as in Table 1. Source: Author’s calculations from GEIH/DANE (2023). The dependent variable is the natural logarithm of hourly wage. Standard errors are robust. *** ρ < 0.01.

) No evidence of heteroskedasticity, multicollinearity, or autocorrelation was observed; the disturbance term appears uncorrelated with the regressors and the dependent variable. After estimating the models and analyzing the results—using DANE time-series—the estimates are consistent with Gauss–Markov assumptions; for these reasons, the four log-linear models are considered robust and relevant for this study. Notably, the agricultural-sector coefficient declines from −0.491 in the baseline to about −0.251 with controls—roughly a 39% raw penalty that narrows to about 22% after controls.

4.3. Decomposition Results: Explained vs. Unexplained Gap

Figure 2 presents the Blinder–Oaxaca decomposition of the mean wage gap (placeholder for an embedded table). We performed the decomposition for the coffee vs. urban gap and the cocoa vs. urban gap. In both cases, we included education, age (experience), gender, and job informality as the key covariates. The decomposition yields two main components: “explained” (due to endowment differences) and “unexplained” (due to different returns or residual factors).

As shown in Figure 2, in the case of coffee vs. urban, 55–60% of the wage gap can be explained by differences in observable characteristics, while the remaining 40–45% is unexplained. For cocoa vs. urban, the split is roughly 50/50, with a slightly larger unexplained portion (up to about 50%). These results indicate that about half of the wage gap is due to rural workers having, on average, lower levels of education, dissimilar experience profiles, and a higher incidence of informal employment, whereas the other half of the gap is due to factors beyond those observables. The unexplained portion captures the rural disadvantage in return to characteristics—effectively, even if a rural worker has the same education and experience as an urban worker, they earn less, which could reflect lower productivity in rural firms, discrimination, lack of bargaining power, or other unmeasured influences like infrastructure and access to markets (Chen et al. 2021; Costinot et al. 2016).

A regression-based Blinder–Oaxaca decomposition (

Table 4. Blinder–Oaxaca decomposition.

Variables	Average Group 1 (Agricultural)	Average Group 0 (Non-Agricultural)	Group 0 Beta	Group 1 Beta	Explained Portion	Unexplained Portion
Age	42.86691312	34.81775701	0.005788171	0.006808568	0.046589891	0.043741294
Education	5.768396846	13.86192469	0.004897276	0.015269325	−0.039636239	0.059830094
Gender Female	0.080134827	0.400934579	−0.147759374	−0.012073062	0.047401171	0.010873199
Children< 6 yrs	0.286071545	0.288785047	−0.145641722	−0.044821412	0.000395199	0.028841822
Contract	0.486571708	0.887850467	0.423313843	0.153878612	−0.169866853	−0.131099561
TOTAL	8.04364681	8.531353927			−0.115116831	0.012186848

Blinder–Oaxaca decomposition, data extracted from GEIH (DANE), own calculations, lapse 2002–2023.

) was estimated to separate the explained and unexplained portions of the rural–urban wage gap for workers in permanent coffee/cocoa crops and related processing across Colombia’s 32 departments. A simple linear wage model was implemented in Python 3.14.0 using GEIH (DANE) data for 2002–2023, with controls for Age, Education, Female, Children < 6, and Contract. The explained component reflects differences in observed characteristics, most notably contract status, schooling, and age, whereas the unexplained component captures residual disparities after accounting for those traits, consistent with structural differences in returns. As shown in Table 4, the explained portion is negative and accounts for most of the observed gap, with contract status and schooling as the leading contributors (Song 2023; OECD 2020; Adamopoulos and Restuccia 2020; Shahe and Forhad 2017; World Bank 2008; Blinder 1973; Oaxaca 1973).

In our decomposition, we found that urban workers had, on average, about eight more years of schooling than rural workers, and this difference alone accounted for about 4% of the wage gap. In other words, if rural workers had the same education levels as urban workers, the gap would shrink, but only by about one-twentieth. Another crucial factor was informality (and related job benefits): roughly half of rural workers in our sample have a formal contract, compared with nearly nine in ten among urban workers. We included an indicator for whether the worker had a formal contract and found that this variable explained a considerable portion of the gap—about 16%. When a worker is informal, their wage tends to be lower; since rural workers are predominantly informal, this drags down the average rural wage.

Our results suggest that differences in the rate of informality and job type explain 30–35% of the gap. This aligns with the intuition that Colombia’s labor regulations (like minimum wage laws) lift wages in the formal urban sector, and their absence in rural areas leaves rural wages to be determined by market forces, often at very low levels. Occupational differences also mattered: rural workers are mostly unskilled farm laborers, whereas urban manufacturing workers might include skilled occupations or machine operators. The occupational effect overlaps with education and informality, but it adds some explanation as well (rural jobs being concentrated in lower-paying occupations) (Foster and Rosenzweig 2022; ILO 2020; Caliendo et al. 2018; Arkolakis 2010; Barrientos and Hulme 2009; Baum-Snow et al. 2017; Allen 2014; Chaney 2008; Baum-Snow 2007; Bernard et al. 2007; Barrientos 2003).

The unexplained component (40–50%) is substantial. This unexplained portion can be interpreted in several ways. It may reflect productivity differentials not captured by education—for example, rural workers might have lower access to technology or capital, making their labor less productive (hence paid less) even if they have similar schooling. It could also reflect compensating differentials or preferences—rural jobs offer non-monetary benefits (like living at home, subsistence farming side income) that partially compensate for lower wages, though this is debatable given the poverty outcomes. A likely interpretation is labor market segmentation and institutional exclusion: rural workers may not have access to the same high-wage sectors or strong labor unions that urban workers do, so even equally skilled rural workers are stuck in lower-paying segments.

Discrimination or social exclusion could also play a role if rural origin workers are undervalued. Moreover, the unexplained gap might capture geographic isolation—people working in remote rural areas are effectively in a different labor market with lower demand and fewer employers, which suppresses wages relative to urban centers (Gáfaro and Pellegrina 2022; World Bank 2021; Chen et al. 2021; Blinder 1973; Oaxaca 1973). Our decomposition findings are in line with other studies. For example, an analysis by the World Bank (2017) and by Otero-Cortés and Acosta-Ariza (2022) similarly found that about half of Colombia’s rural–urban wage gap was unexplained by observable factors. International comparisons show a range of unexplained shares—in some Asian countries with massive rural–urban divides, the unexplained portion can be even larger (suggesting strong structural barriers), while in some more integrated economies it can be smaller. Our result of ~50% unexplained indicates significant structural issues in Colombia’s case.

Additionally, the decomposition showed that when we controlled for an informality indicator, the gap that remained unexplained was closely related to what one might call a “sector premium”, a worker in agriculture is paid less than a worker in industry, even after accounting for personal traits. This is consistent with a dual economy interpretation and suggests that raising rural wages would require changing the sectoral and institutional context, not just the characteristics of workers. Finally, to corroborate the decomposition, we also computed a counterfactual wage for rural workers that had been paid on the urban wage schedule. This exercise (sometimes called reweighting) showed that if rural workers had the same characteristics and returns as urban ones, their wages would be dramatically higher, again indicating that a large share of the gap is due to differences in wage structure (returns) rather than just differences in workers.

In summary, our empirical results paint a consistent picture: a large wage gap exists between Colombia’s rural coffee/cocoa sectors and the urban manufacturing sector; this gap has not significantly narrowed over time; and roughly half of it cannot be explained only by education, experience, or other worker characteristics, reflecting structural disadvantages faced by rural labor (Ananian and Dellaferrera 2024; ILO 2019).

4.4. Interpreting the Results in Light of Theory

Our findings confirm elements of both the Lewis and Harris–Todaro models discussed earlier. The persistent wage gap alongside ongoing rural-to-urban migration aligns with Harris–Todaro’s notion of an equilibrium with expected wage equalization: rural workers keep migrating because of the wage incentive, but many end up in urban informal employment, preventing the urban wage advantage from dissipating. The fact that we do not see convergence over time suggests Colombia has not yet absorbed its surplus rural labor—there is still effectively an excess supply in rural areas willing to work at low wages, consistent with the Lewis model of an economy before the turning point.

The surplus labor condition is evidenced by the low marginal productivity on small farms and the ability of the rural sector to continue functioning with low wages without triggering massive labor shortages (indeed, if anything, rural depopulation is a concern in some zones, but wages remain low, indicating low labor demand or productivity) (Ananian and Dellaferrera 2024; Gáfaro and Pellegrina 2022; Chen et al. 2021; ILO 2019; Costinot et al. 2016; Moffitt 1983).

The structural factors we identified (like informality and segmentation) tie into broader institutional analyses. For instance, Colombia’s labor regulations (minimum wage) are national in scope, but enforcement in rural areas is weak. The minimum wage sets a standard for formal urban employment but is often not honored in rural day labor arrangements. One interpretation of our unexplained gap is that it reflects this regulatory gap—rural workers effectively operate outside the protection of labor laws. This resonates with the Harris–Todaro idea that raising rural wages via policy could simply result in fewer jobs if not accompanied by productivity gains. It also resonates with global evidence that countries with more inclusive labor institutions (covering rural areas) tend to have smaller rural–urban wage gaps (Ananian and Dellaferrera 2024; Gáfaro and Pellegrina 2022; Chen et al. 2021; ILO 2019).

In summary, our results highlight that market forces alone have been insufficient to equalize rural and urban wages in Colombia’s context of surplus labor and institutional weaknesses. Without intervention, the gap shows no signs of closing. We now turn to a comparative perspective to see how other countries have addressed similar challenges, which will inform our policy discussion for Colombia.

5. Discussion

5.1. Comparative Perspectives: Lessons from Latin America

While Colombia’s rural–urban wage gap is pronounced, it is by no means unique. Many Latin American countries have historically faced similar divides between an impoverished rural sector with the highest informality and a more prosperous urban sector. In this section, we summarize key lessons from other countries, focusing on Brazil and Mexico, with additional insights from other Latin American experiences, to shed light on policies and structural factors that can influence rural wages. These comparative lessons will help contextualize our policy recommendations for Colombia.

5.2. Brazil: Rural Social Protection and Agrarian Development

Brazil provides a compelling case study for addressing rural–urban disparities. Like Colombia, Brazil for much of the 20th century had a large gap between its rural agricultural workforce and urban industrial workers. Rural poverty and low wages were especially acute in regions like the Northeast of Brazil. However, from the 1990s onward, Brazil implemented a series of policies that significantly improved rural incomes and reduced the rural–urban wage gap (though not eliminating it entirely). One of the cornerstone policies was the introduction of a non-contributory rural pension program in the early 1990s (Prêvidencia Rural). This program provided a basic pension to rural elderly (including farmers and rural laborers) even if they had not paid formally into a social security system. The impact was substantial: studies found that the rural pension lifted millions of rural Brazilians out of extreme poverty and effectively set a floor on rural household income (IADB 2018).

By guaranteeing older rural workers a minimum income, the pension reduced the need for the elderly to continue working for low wages (in Colombia, by contrast, many older rural people must keep working due to a lack of pensions). The Brazilian rural pension thus raised the reservation wage for rural labor—younger workers could be less pressured to accept extremely low pay if their household had some pension income support. Barrientos (2003) estimates that Brazil’s rural pension was responsible for significant poverty reduction; one study noted it cut the rural poverty gap by up to 40% (IADB 2018; Barrientos 2003).

Brazil also expanded conditional cash transfer programs (like Bolsa Família), which, while not wage policies per se, injected cash into poor rural households, improved nutrition and schooling, and indirectly bolstered the human capital of rural youth. Over time, these measures contributed to better opportunities for rural children and a gradual improvement in their earning potential. Empirical analyses credit Bolsa Família with reducing extreme poverty and inequality in Brazil’s rural areas in the 2000s (IADB 2018; Barrientos 2003).

Another area where Brazil made strides is labor market formalization in rural industries. Through stricter enforcement of labor laws and targeted programs, Brazil increased the share of rural workers with formal contracts. For example, large agricultural enterprises (such as sugar cane mills, orange plantations, coffee estates) were pushed to formalize their workforce, often through inspections and legal requirements. Consequently, many rural laborers in those industries started receiving at least the legal minimum wage and benefits. While small family farms remained informal, the formalization of larger agribusiness labor helped raise average rural wages. There is evidence that in states like São Paulo, rural wages climbed closer to urban levels once formal sector coverage expanded (albeit urban wages also rose). It is noteworthy that Brazil’s minimum wage policy had broad effects: Brazil raised its real minimum wage significantly in the 2000s, which not only affected urban workers but also increased the reference for rural formal wages and even informal wages via a lighthouse effect. Research suggests that these policies contributed to a narrowing of the wage gap during the 2000s Barrientos 2003).

Additionally, Brazil invested heavily in agricultural productivity and rural development projects. There were programs for improving yields in family agriculture, technical assistance and rural extension services provided by agencies (often with support of the National Federation of Agriculture), and credit programs (PRONAF) aimed at smallholders. These interventions helped increase the output and income of small farms, enabling them to pay higher wages to hired labor or at least improve self-employment earnings. Brazil also has a robust system of cooperatives in some regions (e.g., dairy cooperatives, coffee cooperatives in Minas Gerais and São Paulo) that improve market access for small producers. Cooperatives can help farmers receive better prices for their products, indirectly supporting higher labor incomes. For example, some Brazilian coffee cooperatives provide profit-sharing to farm workers or invest in community services Chen et al. 2021; ILO 2012).

The net result of these efforts in Brazil was a sharp reduction in rural poverty (Barrientos 2003). By the mid-2010s, the urban wage premium in Brazil had fallen compared to the 1990s. However, Brazil’s experience also shows that convergence can stall or even reverse under adverse conditions. During periods of economic crisis or when there was less emphasis on rural support (like structural adjustment in the 1980s), rural progress slowed, and regional inequalities persisted. As of the late 2010s, Brazil still had notable gaps, especially between the prosperous South/Southeast and the poorer Northeast. But the lesson for Colombia is that proactive social policy and inclusive labor institutions can be effective. Extending social security to rural workers (pensions, health insurance), enforcing basic labor rights in rural employment, and investing in rural education and productivity were key in Brazil’s case. Colombia’s context differs (for instance, farm sizes are smaller on average, and conflict has been a major factor), but Brazil’s success in reducing rural hardship provides a policy template worth examining IMF 2022; Chen et al. 2021; Haggblade et al. 2010; Barrientos 2003).

5.3. Mexico and Other Latin American Countries: Social Programs and Migration Dynamics

Mexico offers another instructive comparison, though its rural development path has been different. Historically, Mexico underwent a major land reform in the 20th century that created the ejido system (community-owned lands) aimed at empowering peasant farmers. While this gave rural households access to land, it did not guarantee higher incomes—many ejidos remained low-productivity. In modern times, Mexico’s rural economy has been heavily influenced by migration, both internal and to the United States. Massive out-migration of rural Mexicans to the U.S. labor market has functioned as a “pressure valve,” reducing rural labor supply and bringing in remittances that bolster rural incomes. This is a unique factor: remittances in some Mexican rural communities significantly raise household income, albeit not by raising local wages but by providing external income. Despite this, Mexico still exhibits large rural–urban income gaps, partly because areas without migration outlets remain poor and because more productive workers often leave (a brain drain effect) (Chen et al. 2021; Taylor and Martin 2001; Harris and Todaro 1970).

One of Mexico’s most influential policies was the introduction of Progresa/Oportunidades (now called Prospera), a conditional cash transfer (CCT) program launched in 1997 targeting poor rural families. This program provides cash stipends to families conditioned on keeping children in school and attending health checkups. Evaluations have shown it had multiple benefits: it reduced short-term poverty and improved nutrition, and it increased educational attainment among rural youth (especially girls). Over time, better-educated rural children either found better jobs or migrated to cities, which can help narrow wage gaps in the long run. While cash transfer does not directly raise market wages, it improves the human capital of the next generation and gives households a fallback income source, potentially increasing their reservation wage.

Indeed, Prospera and similar programs have been emulated in many countries. Mexico’s experience underscores that investing in human capital in rural areas is crucial—better education and health eventually enable rural workers to access higher-paying opportunities (often by leaving low-paying farm work). Additionally, Mexico’s social programs have mitigated extreme poverty: rural poverty in Mexico (pre-pandemic) was around 20–25%, which, while higher than urban poverty, is lower than Colombia’s rural poverty (above 40%). This suggests that Colombia could improve its situation by more aggressively implementing social protection programs in rural areas (Colombia’s own CCT program, ‘Familias en Acción’, is smaller in coverage and benefit levels compared to Mexico’s, and Colombia lacks a rural pension system like Brazil’s) (IADB 2018).

Other Latin American countries provide further lessons. In Central America, many countries share Colombia’s challenges of agrarian poverty and have seen large migration flows. One lesson from Central America is the importance of rural non-farm job creation. Because land productivity is limited for small farms, developing rural industries (agro-processing, handicrafts, rural tourism, etc.) and services in small towns can absorb surplus labor and offer higher wages than subsistence farming. For example, Costa Rica invested in rural tourism and specialty agricultural processing, providing jobs in rural regions outside of farming. Costa Rica also supported strong cooperative institutions, especially in the coffee and dairy sectors, which improved marketing and price stability for farmers.

As a result, while gaps persist, Costa Rica’s rural communities have better living standards. Another example is Chile, which, though highly urbanized, has used targeted subsidies and infrastructure development to integrate rural areas into national markets (e.g., Chile’s extensive rural road building and electrification improved access to opportunities). Infrastructure investment (roads, transport, internet connectivity) is frequently cited as a needed intervention—Pérez (2020) highlights that improving rural roads can reduce transaction costs and raise farmgate prices, thereby potentially allowing higher wages for rural workers (Chen et al. 2021).

In the Andean region, countries like Peru and Ecuador have pursued strategies such as promoting high-value agricultural exports (e.g., asparagus, broccoli in Peru; roses in Ecuador). These industries, often through foreign investment, created rural wage jobs that pay above traditional farming (though labor conditions vary). However, a known issue is that if labor supply is abundant, even expanding export agriculture can result in a mostly low-wage labor force unless accompanied by unionization or labor regulations enforcement. For instance, Peru’s agricultural export boom initially had a special labor regime with lower protections, which kept wages low. Only recently have there been reforms to improve those rural workers’ conditions. The lesson here is that simply having a booming agricultural sector does not guarantee high wages for workers; the institutional setting matters (Ananian and Dellaferrera 2024; ILO 2019; World Bank 2017).

Argentina and Uruguay, with more mechanized agriculture, have fewer farm workers but higher rural wages for those employed, partly because of stronger rural unions and collective bargaining agreements in some industries (e.g., Argentina’s rural workers’ union UATRE negotiates sectoral wage floors). This suggests that labor organization and collective bargaining could play a role in lifting rural wages, though in Colombia’s context of mostly small farms, the applicability is limited (cooperatives might be more relevant than unions for small farmer groups) (Ananian and Dellaferrera 2024; ILO 2012, 2019; World Bank 2017).

In summary, across Latin America, common themes emerge:

▪

Social protection programs (pensions, cash transfers) can significantly alleviate rural poverty and indirectly support wages by raising reservation incomes.

▪

Human capital investments (education, healthcare) in rural areas pay off in the long run, enabling rural residents to obtain better-paying jobs and be more productive.

▪

Enforcement of labor standards (minimum wage, working conditions) in rural areas, though challenging, can help formalize rural employment and push wages upward (Brazil’s case).

▪

Encouraging rural industries and value addition (e.g., processing agricultural products locally, developing rural service hubs) can provide alternative employment to absorb surplus labor at better wages.

▪

Strengthening rural institutions such as cooperatives and farmer organizations can empower small producers and workers to find better terms from the market.

▪

Facilitating migration can relieve labor pressure, but it is not a panacea—unchecked migration can lead to urban unemployment and does not raise rural wages for those left behind. The ideal is to create conditions where people are not forced to migrate out of desperation but migrate by choice.

These lessons will inform the following section, where we translate them into specific policy implications for Colombia.

5.4. Comparative Policy Dimensions: Radar Chart

This section compares six policy dimensions linked to narrowing rural–urban wage gaps—Social Protection, Human Capital, Labor Standards Enforcement, Non-farm Rural Employment, Cooperatives/Institutional Strengthening, and Migration/Remittances—for Brazil, Mexico, and Costa Rica. The radar chart summarizes how each country emphasizes different levers to reduce poverty and improve rural incomes.

Brazil shows its largest gains in social protection and labor-standards enforcement, reflecting expansive income support and stronger compliance that have raised rural earnings and formalization; meanwhile, Mexico scores highest in migration/remittances, with solid human-capital investments that boost household resources and improve schooling; in turn, Costa Rica’s strengths center on non-farm rural employment and cooperatives/institutional strengthening, where rural tourism and well-organized producer cooperatives have broadened income sources and bargaining capacity (Figure 3).

Relative to these peers, Colombia underperforms on Social Protection, Human Capital, and Labor Standards, which helps explain persistent rural informality and lower wages. The policy lesson is to combine: (i) targeted social protection that reaches rural workers, (ii) sustained human-capital upgrades, (iii) effective labor-standards enforcement, and (iv) diversification toward non-farm rural employment supported by strong cooperative institutions. This integrated approach—coordinated by the agriculture ministry, national planning, and the executive—aligns with the strategies that have shown results in Brazil, Mexico, and Costa Rica while addressing Colombia’s gaps.

6. Policy Implications for Colombia

Closing the rural–urban wage gap in Colombia is a long-term challenge that requires a multi-faceted policy approach. Based on our findings and lessons from other countries, we outline several key policy recommendations to improve rural wages and livelihoods. These recommendations target the structural issues identified—low human capital, labor market dualism, weak institutions, and low productivity in rural areas. Each recommendation is accompanied by a brief rationale and evidence:

▪

Invest in Rural Education and Skills Training: Improving the quality and accessibility of education in rural areas is crucial. Higher educational attainment will enable rural workers (especially the youth) to be more productive and to access better-paying jobs, either in rural non-farm activities or in urban areas if they choose to migrate. Currently, lower education accounts for a significant portion of the wage gap. Policies should include increasing funding for rural schools, offering targeted scholarships for rural students, and expanding vocational training programs in agricultural communities. Evidence suggests that educational disparities contribute to wage disparities; thus, closing the education gap can gradually narrow the wage gap. Moreover, extension services and technical training for farmers can boost productivity and incomes on the farm. Over time, a more skilled rural workforce can demand higher wages and diversify into higher-value activities (Gáfaro and Pellegrina 2022; Chen et al. 2021).

▪

Strengthen Labor Market Institutions and Wage Enforcement in Rural Areas: Colombia should work to enforce labor standards (such as minimum wage laws and social security coverage) for rural and agricultural workers. While challenging due to informality, incremental steps can be taken, such as formalizing labor contracts in larger farming enterprises and instituting wage boards or standard rates for certain agricultural tasks. Brazil’s experience showed that extending labor rights to rural workers (e.g., mandating formal contracts and benefits on large farms) helped lift rural wages. Colombia could introduce a rural minimum wage appropriate to regional productivity levels or strengthen the labor inspectorate to curb abusive practices. Even if not all rural jobs can be formalized immediately, setting normative wage guidelines (for instance, through local committees of the Ministry of Labor in farming regions) can empower workers to negotiate better pay. Reducing extreme exploitation is critical—when rural workers have no alternative and no legal protection, wages remain at a bare subsistence level. The International Labour Organization guidelines on wage-setting and equality recommend extending coverage to rural areas. As rural labor becomes better protected, the wage gap should reduce, though care must be taken to not simply drive employment to informality; thus, enforcement needs to go hand-in-hand with support for employers to increase productivity (so they can pay the higher wages) (Ananian and Dellaferrera 2024; IMF 2022; Chen et al. 2021; Barrientos 2003).

▪

Expand Rural Social Protection (Non-contributory Pensions and Cash Transfers): Providing a basic income floor for rural households can raise the reservation wage and reduce poverty. A non-contributory rural pension (similar to Brazil’s) would allow older rural workers to retire with dignity and reduce the labor oversupply caused by elderly people continuing to work because they cannot afford to stop. If older workers exit the labor pool on a pension, younger workers may find it easier to bargain for better wages. Barrientos (2003) finds that pensions significantly reduced poverty in Brazil’s rural areas. Colombia could build on its recent pilot programs for rural minimum income for seniors. Additionally, conditional cash transfers like ‘Familias en Acción’ should be bolstered and targeted at the poorest rural regions, potentially with higher benefits to account for gaps. These transfers, as seen in Mexico’s Prospera, improve human capital and provide immediate poverty relief. While they do not directly raise market wages, they reduce the desperation that forces rural people to accept extremely low pay, effectively setting a floor under incomes. Over time, better-nourished, better-educated rural youth will be able to command higher wages. Expanding social protection is a direct way to address the welfare gap while indirectly influencing the labor market to not tolerate ultra-low wages (Murialdo 2023; World Bank 2022; IADB 2018; Haggblade et al. 2010; Barrientos 2003).

▪

Promote Farmer Cooperatives and Market Access Initiatives: Organizing small farmers and rural workers into cooperatives or associations can increase their bargaining power and share in value addition. Cooperatives can help members obtain better prices for crops, reduce input costs, and even set up processing facilities. In contexts like Costa Rica and some Brazilian states, cooperatives have improved rural livelihoods. Colombia has some cooperatives (including in coffee through the National Federation of Coffee Growers), but encouraging more bottom-up cooperatives in cocoa and other sectors could enable farmers to pay themselves and any laborers more fairly. Additionally, facilitating access to higher-value markets—for instance, certifications for fair trade or organic coffee/cocoa—can allow producers to earn premium prices that translate into higher wages. If small producers receive a larger share of the final consumer price, they can afford to pay hired labor more. The International Labour Organization (ILO) has emphasized cooperatives as a tool for people-centered rural development, noting their role in empowering workers and improving incomes. Government support can include providing technical assistance to nascent cooperatives, start-up grants, or preferential credit. Over time, stronger rural organizations can also serve as channels to implement minimum wage agreements or profit-sharing in communities (ILO 2012, 2019, 2020; Cosar and Demir 2016; Cosar and Fajgelbaum 2016; Bernard et al. 2007).

▪

Boost Agricultural Productivity and Value Addition: Low productivity in rural Colombia is a fundamental cause of low wages. Thus, policies that raise productivity can create room for higher wages without rendering farms non-viable. This includes investing in rural infrastructure such as roads, irrigation, storage facilities, and electrification. Better roads, for example, reduce the cost of transporting crops to market, raising the prices farmers receive. If farm revenues increase, there is more scope to increase wages for laborers. Rural road investments in other Latin countries have shown positive effects on local economies. The government should prioritize infrastructure projects in coffee and cocoa growing regions (some of which are remote and were affected by conflict). Alongside infrastructure, increasing access to finance and technology for small farmers will enable them to adopt improved seeds, fertilizers, or new crops, thus increasing output. Government programs or public–private partnerships could facilitate affordable credit and insurance for smallholders. A more productive agricultural sector can support higher labor costs. It is also important to diversify and add value: for instance, encouraging local processing of cocoa into chocolate, or coffee into roasted beans, rather than exporting raw commodities. The government could provide incentives (tax breaks, subsidies) for establishing processing facilities in secondary towns of coffee/cocoa regions. This would create better-paying jobs in rural areas (e.g., a job in a chocolate factory pays more than picking cocoa pods) and reduce the need for workers to migrate to cities. Such agro-industrialization has been part of rural development strategies in countries like Mexico (where programs tried to entice maquiladora factories to smaller towns) and can help integrate rural populations into higher value chains (Gáfaro and Pellegrina 2022; Chen et al. 2021; Benguria et al. 2021; OECD 2020; IADB 2018; Haggblade et al. 2010; ILO 2020).

▪

Facilitate Orderly Migration and Urban Integration: While the goal is to improve rural conditions so that migration is by choice rather than necessity, migration will continue to be a reality. Policymakers should ensure that those who do migrate have pathways to decent urban employment (for example, through training programs that target rural youth for urban skills, or urban job placement services). This can prevent a scenario where migrants simply swell the ranks of urban informal workers, which maintains the Harris–Todaro equilibrium. If rural migrants can more quickly secure formal jobs in cities, the expected income gap would narrow, and migration flows would balance. Additionally, rural development efforts in conflict-affected zones (now under the Peace Accord’s rural reform) need to be accelerated, as peace and security are prerequisites for any economic initiative to thrive. Reducing violence and improving land governance will make other interventions more effective (Taylor and Martin 2001; Harris and Todaro 1970).

Implementing these recommendations requires political will, significant public investment, and coordination across agencies. Crucially, these policies complement each other: for instance, better education and health amplify the benefits of cash transfers; infrastructure and cooperatives together can help farmers reach markets; pensions and labor enforcement together can set floors on income and wages. The experiences of Brazil, Mexico, and others show that policy matters—rural–urban disparities are not immutable. With targeted, consistent policies, Colombia can start to close the gap. Narrowing the wage gap will not only improve equity and social justice but also likely benefit the economy by increasing rural consumers’ purchasing power and slowing the costly rural exodus. It may also enhance social stability by reducing the grievances that stem from rural marginalization (ILO 2019).

In summary, our policy recommendations for Colombia are to invest in rural human capital, extend and enforce labor protections, provide social safety nets to rural populations, strengthen collective institutions for farmers, promote rural industries and infrastructure, and manage migration humanely. These measures, backed by both our analysis and comparative evidence, form a comprehensive approach to tackle the rural–urban wage gap.

Figure 4 summarizes the expected effects of the recommended policy package on narrowing Colombia’s rural–urban wage gap, benchmarked against Brazil, Mexico, and Costa Rica. It portrays how a coordinated mix of interventions should lift rural earnings, especially for workers in permanent coffee and cocoa crops, by reducing informality, raising productivity, and strengthening worker protections.

Limitations and Future Research

Although our study significantly advances the understanding of rural–urban wage gaps in Colombia’s coffee and cocoa sectors, there remain opportunities to enhance and extend this research. One aspect is related to data constraints; our reliance on cross-sectional household surveys limits the ability to capture detailed individual wage trajectories or firm-level productivity differences that may influence earnings. Future research would benefit from longitudinal data tracking individuals over time, enabling more precise assessments of wage dynamics and labor mobility (Chen et al. 2021; Franco and Ramos 2010).

Another opportunity for further investigation lies in examining structural and institutional dimensions that are difficult to capture quantitatively, such as labor market discrimination, social capital, and informal networks that impact rural wages (Otero-Cortés and Acosta-Ariza 2022). Qualitative and mixed-method approaches, including field interviews or case studies, could illuminate these unmeasured factors, offering deeper insights into the persistent unexplained component of the wage gap observed in this study. Lastly, while our comparative analysis with Brazil and Mexico provides useful policy insights, future studies could expand the geographic and methodological scope. Rigorous econometric evaluations of specific policy interventions within Colombia, such as rural pension schemes, conditional cash transfers, or cooperative initiatives, would offer more robust evidence regarding their effectiveness and replicability. Such research would be instrumental in informing targeted policy decisions designed to address rural wage disparities and improve rural livelihoods in Colombia and similar contexts.

7. Conclusions

This research provides the first comprehensive analysis of the rural–urban wage gap in Colombia, with a focus on the coffee and cocoa sectors in Colombia in these sectors over two decades. Our analysis confirms that a large and persistent wage disparity exists; rural workers in these traditional agricultural sectors earn, on average, about half of what urban manufacturing workers earn. Even after accounting for differences in education and experience, a substantial wage penalty remains for being a rural worker, highlighting deep structural issues. We found little evidence of market-driven convergence between rural and urban wages over the past two decades; the gap has neither systematically narrowed nor closed. This enduring disparity is rooted in Colombia’s dual economic structure, where a low-productivity, labor-abundant rural sector coexists with a more productive, protected urban sector.

Classic development theories (Lewis’s dual-sector model and Harris–Todaro migration model) are borne out in the Colombian context: surplus rural labor and institutional barriers keep rural wages depressed, while urban wage floors and limited formal jobs sustain a high urban premium. Our Blinder–Oaxaca decomposition showed that roughly half of the wage gap can be explained by observable differences (like lower education and higher informality in rural areas), but the other half is unexplained, indicating factors such as weak bargaining power, lack of opportunities, and segmentation are at play. By bringing in a comparative perspective, we illustrated that Colombia’s situation, though challenging, is not insurmountable.

Countries like Brazil and Mexico have implemented policies that demonstrably improved rural incomes and reduced gaps—from rural pensions and cash transfers to education investments and labor formalization. These examples provide a menu of policy options and underscore a key point: rural–urban wage gaps are not intractable. They are primarily man-made—the product of policy choices and historical structures—and thus can be reduced through deliberate action. Narrowing the wage gap will not only uplift millions of Colombia’s rural citizens out of poverty but also contribute to national cohesion and sustainable development. A Colombia with a smaller rural–urban divide would see benefits in terms of reduced migration pressures, a more balanced economic growth, and greater social stability (Delgado et al. 1998).

For policymakers, the implications are clear. Market forces alone will not eliminate the rural wage penalty in the near future; initiative-taking interventions are required. Strengthening rural education, healthcare, and infrastructure will tackle human capital and productivity deficits. Enhancing social protection and enforcing labor rights will provide immediate relief and empowerment to rural workers. Encouraging rural industrialization and value chain integration will create better-paying jobs where people live. These strategies, pursued in an integrated manner, can gradually chip away at the wage gap. It is essential to involve rural communities in designing and implementing these policies—local buy and adaptation to regional realities will improve effectiveness (IMF 2022).

Monitoring and evaluation should be built in as policies are rolled out; their impact on rural wages should be tracked (for example, through periodic labor force surveys and studies) to learn what works best in the Colombian context. Finally, this research points to avenues for future studies. One critical area is to further investigate the role of unobservable factors in the wage gap—for instance, measuring productivity differences directly or examining the impact of social networks and discrimination on rural workers’ opportunities. Another avenue is to conduct longitudinal analysis of individuals as they migrate from rural to urban areas: understanding how quickly (or slowly) migrants’ wages catch up in cities would inform both migration and education policies. Additionally, future research could evaluate specific policy pilots in Colombia’s rural regions (such as the impact of a new rural entrepreneurship program or the effects of land reform implementation on wages). Such micro-level studies would complement the macro comparisons and help fine-tune interventions. From a comparative standpoint, examining Colombia’s progress relative to peers in the next decade could yield insights, such as treating policy changes as experiments to learn what narrows the gap.

In conclusion, the wage gap between rural and urban Colombia, exemplified by the coffee and cocoa sectors vs. manufacturing, is a significant challenge but one that can be addressed with informed and determined policy action. By combining economic analysis with a comparative perspective, our study provides evidence-based recommendations for fostering inclusive rural development. The task ahead for Colombia’s leaders and stakeholders is to translate these insights into concrete actions—to build a country where the standard of living is less determined by being born in a rural village versus a city. Bridging the rural–urban divide will be pivotal for achieving equitable growth and realizing the full potential of Colombia’s economy and society.