Labor market policies

From the perspective of Keynesian thought, the idea of full employment is an objective that is not satisfactorily achieved by the market (^{Davidson, 2001}). Meanwhile, according to ^{Toporowski (2018)}, one of the most important contributions to political economy of Kalecki’s wartime work was related to the political aspects of full employment and the restrictions that are established in a capitalist system where unemployment is an integral part of a normal capitalist system, so if at some point capitalism coincides with full employment that will be fundamentally different capitalism. Lerner, who coined the idea of functional finance (¹⁹⁴³, ¹⁹⁴⁴), assumes that the counterpart of inflation has to do with unemployment and assigns an important role to the action of the State as a regulator of spending that permits equilibrium in the economy. Likewise, under the theoretical umbrella of functional finance and some post-Keynesian principles, Modern Monetary Theory (MMT) also supports using monetary and fiscal policy to achieve employment objectives. As stated by ^{Wray (2020)}, MMT favors spending aimed at achieving full employment, even though, according to ^{Pérez and Vernengo (2020)}, attention must be paid to the peculiarities of the labor market in developing countries, such as structural or hidden unemployment.

In public policies related to the labor market (LMP), an important division has to do with how much the government intervenes in the labor market. Accordingly, a difference is made between active and passive policies (ALMP and PLMP, respectively) (see Table 1), where the difference in the objective pursued is that passive policies aim to protect the needs produced by the lack of employment. In contrast, active policies balance labor supply and demand to reduce or eliminate unemployment (^{Álvarez, 2016}).

^{Martínez (2008)}, based on a review of other authors, points out that the difference between active and passive policies is that the latter focus on an income substitution component of transfers whose objective is to maintain the income of the unemployed, while the former aim to establish a process of integration into the labor market so that while the logic of passive policies is to protect, compensate or indemnify, the objective of active policies is to achieve the participation or activation of people in the labor market. A trend seems to be emerging in favor of active policies over passive ones, given that PLMPs can generate perverse incentives by discouraging job search (^{Bucheli, 2005}). ^{Vooren et al. (2019)} additionally mention that passive policies such as unemployment benefits can be more costly, in addition to the fact that they cannot shorten the gap between needs and skills that is achieved by ALMPs through programs such as training; additionally, they also favor employment and avoid the loss of income and the depreciation of human capital that is amplified as the worker becomes inactive.

In practice, the average spending on ALMP in OECD countries has been around 0.5% of their GDP, which has remained relatively stable and, at the same time, counter-cyclical since in the two moments of crisis occurring between 2004 and 2020, spending showed an increase, while in the case of the PLMPs, there was a greater increase in the setbacks and a more marked reduction than in the active policy, which shows its function as an automatic stabilizer (Figure 1). Also noteworthy is the increase in spending in 2020, which would be expected to return to levels similar to those of 2008 or 2018 as the economy recovers.

Source: created by the authors with OECD data

Figure 1 Spending on Assets and Liabilities in the labor market (%GDP) 

On the other hand, Figure 2 provides information regarding expenditures made in 2018¹ by the OECD countries. In principle, only 9 countries of the total spending on LMP-more than 50%-is allocated to ALMP. Denmark together with France are the countries that allocated the largest number of resources to LMP, while the United States and Japan were the ones that allocated the least resources to that purpose. In the case of Mexico, although constitutionally, it establishes a broad reference to the issue of labor and social welfare,² in practice from 2004 to 2016 it allocated less than 0. 01%. Subsequently, with the adjustment made in 2017 due to the increase in indebtedness, which framed the return to fiscal austerity, the budget allocation for LMP was less than 0.005%. Even in 2020 due to the pandemic crisis, spending on average increased by more than one percentage point relative to GDP in OECD countries, as seen in Figure 1. By way of comparison, Chile, since 2008, has recurrently increased the percentage of resources allocated to LMP: 0.26% in 2008, 0.5% in 2019, and 2020 1.24%.

Source: created by the authors with OECD data

Figure 2 Total resources allocated to LMP (%GDP, 2018) and their composition 

Finally, regarding Figure 2, in which of the categories mentioned in Table 1 is the spending concentrated in the countries that spend the most, namely France and Denmark: regarding ALMP, about a third is focused on training and supported employment and rehabilitation, respectively; in the case of PMLP, in Denmark 90% is allocated for maintenance and support of income without work, while in France it is 100%. In the countries at the extremes, Chile and Hungary, in terms of ALMP, both spend more on direct job creation with 33 and 63 %, respectively, while in PLMP, spending is similar to France.

Empirical evidence supports government participation in LMPs and especially in ALMPs; nevertheless, ^{Bucheli (2005)} states the need to pay attention to program design, considering the causes of unemployment, its duration, the profile of the unemployed, and their probability of finding employment. An appropriate design will lead to acceptable results. In evaluating ALMPs, the Public Employment Service and Management is considered one of the best performers (^{Bucheli, 2005}). Regarding design, ^{Calmfors (1995)} mentions that a scheme should be sought that provides the unemployed with better opportunities for employment while avoiding the generation of perverse incentives that lead them not to do so, and the best way to achieve a good result is to have a portfolio of programs with all available policies since each of them offers different alternatives that can usefully complement each other.

Although there is some consensus on the benefits of ALMPs, there are also some questions related to the heterogeneity of the results according to the country in question, in addition to whether the programs that seek greater employability are also valid at times when labor demand is low or only for good times (^{Martín, 2014}). According to ^{Kluve (2010)}, a large part of the result that can be obtained will depend on the program chosen, where programs such as employment services and management can improve employability if they are accompanied by training and employment subsidies, while what should be avoided are direct employment policies in the public sector since they are not effective and do not offer greater employment opportunities for participants.

As ^{Shanoun and Abdennadher (2018}) and ^{Vooren et al. (2019}) also argue, ALMPs are a better option than PMLPs. Shanoun and Abdennadher (2018) performed an analysis of 18 OECD countries distributed in 4 groups with data corresponding to the period 2000-2014. The authors found an overall negative relation of ALMPs to the unemployment rate, with differentiated effects individually: training improves productivity and growth, generating more employment; the public employment service and management establishes a link between supply and demand, impacting unemployment downwards; employment policies and incentives are not effective, and if they are also very focused they generate inefficiency derived from the idea of dead weight and the increase in the cost of the subsidy; direct job creation policies, besides being irrelevant are also inefficient due to the creation of new supply with a low marginal product; incentives to start-up according to the authors is the best policy, besides being the one that entails the lowest costs.

^{Vooren et al. (2019}), through the analysis of a sample of 57 experimental and quasi-experimental studies that offered 654 estimates published between 1990 and 2017, provide evidence of short- and long-run effects of ALMPs, although it is relatively inconclusive. Within their findings, services and public employment management stand out especially when there are sanction systems. Positive coefficients are found in training, but they are not significant. As for subsidized labor, direct job creation presents insignificant short-run results, and only subsidized labor is significantly positive in the long run.

Thus, different results should be considered to improve the dynamics of the labor market and, to this end, promote LMPs in those activities that have a high potential for generating employment, expanding the effect of the policy, as well as the growth and welfare of the population, thus satisfactorily complying with social policy.

Product input

In the economic literature, several studies with an input-output approach analyze multipliers and their impact on production, income, and employment. These analyses are carried out at the national and regional levels using different methodologies. Some of them are highlighted below.

^{Boundi (2016)} identifies the key industries of the Mexican economy with the application of different methodologies based on input-output analysis, proposed by ^{Chenery-Watanabe (1958)}, ^{Rasmussen (1963)}, ^{Leontief (1985)} and ^{Ghosh’s (1958}, ¹⁹⁶⁸⁾ supply approach. The results of these applications show that the key sectors in Mexico are the intermediate goods and capital goods industries. In calculating production linkages, the Rasmussen method is more robust for measuring backward linkages (BL) and forward linkages (FL) effects. These results show that the industrial sectors are the most important in the Mexican economy.

^{Walle et al. (2022)} present the main linkages and clusters of the State of Tamaulipas based on the regionalized input-output matrix (IOM). They perform a principal component analysis with the methodology of ^{Feser and Bergman (2000)} and a study of the linkages with the Chenery and Watanabe indices and the Rasmussen indices of power and dispersion sensitivity. Using Feser and Bergman’s methodology, 11 clusters were identified, indicating where public resources and efforts to promote the State’s economic development should be directed. The classification obtained by Rasmussen’s power and dispersion sensitivity indices allowed them to identify eight key sectors. With these results, the authors conclude that the government should prioritize the food, energy, electricity, electronics, and auto parts sectors in its public investment agenda.

^{Mattas and Shrestha (1991)} use input-output, output, employment, and income elasticities to identify key sectors of the Greek economy. They estimate linkages to compare elasticities and better understand the structure and dynamics of the economy. Their results find that agriculture, food, construction, transportation, services, trade, banking, and textiles are the most important sectors for the Greek economy regarding potential output, employment, and income and recognize that sectoral classifications depend on the index used: elasticities or linkages. Elasticities reflect the relative capacities of sectors to generate impacts according to their sizes. They conclude that the elasticity approach produces more robust estimates of sectoral output, employment, and income potentials than the backward linkage approach.

^{Pérez-Santillán (2022)} derives the key sectors from the linkage measures developed by ^{Rasmussen (1963)}. The author uses IOMs published by the OECD in its 2018 edition, disaggregated into 36 sectors for the period 2005-2015. The key sectors for the Mexican economy are mining support services; wood and wood products and cork; paper and printing products; coke and refined petroleum products; chemical and pharmaceutical products; rubber and plastic products; other non-metallic mineral products; basic metals; fabricated metal products; and electrical equipment.

Through a comparative analysis of the matrices of regional and national technological coefficients for Chile, ^{Arriagada (2004)} distinguishes in the production structure of the region those branches with the potential to become dynamic axes of growth. The results show great similarities, such as the linkages of the production sectors of the national and regional structure, so that the classification of key, driving, strategic, and independent activities is similar, except for commerce, hotels, and restaurants.

^{Mendoza (2020)} presents the relations of economic interdependence in the production structure of Sonora using the 2013 input-output matrix. The author’s analysis shows that primary and industrial activities are key to the development of the production structure. In the classic approach of key sectors, the author found that the activities that generate the most economic interdependencies are agriculture, stockbreeding, hunting, and fishing; non-oil mining, input manufacturing, textile finishing, wood industry, paper industries, printing, and related industries; manufacture of petroleum and coal products, manufacture of products based on non-metallic minerals, and basic metal industries. Finally, the author recommends that the promotion of state economic development has to consider the existing economic interdependencies in primary and industrial activities, as a way to increase the possibilities of growth and state sectorial development.

Alarcón (2018) performs an analysis for Mexico, using graph theory to hierarchize the technology-based sectors (TBS) and their production linkages based on the 2012 national Input-Output matrix. The analysis describes the hierarchies through their sales and purchases and the sectoral interrelationships generated by their relative position with the rest of the industrial sectors, showing that the chemical and pharmaceutical subsector is key in the production linkages. In contrast, the other technology-based subsectors are positioned mostly as drivers or independent.

^{Fuentes and García-Andrés (2009)} emphasize that the search for key sectors of the economy based on production linkages does not consider that key sectors are not defined exclusively in terms of economic weight or the intensity of direct economic relations but rather in terms of the role played by productive interrelationships. The strength and stability of a structure depend not only on the weight of the relations between sectors but also on their relative position. This method allows them to carry out a sectoral hierarchization considering the dependency relations.

These studies focus mainly on the input-output analysis of intersectoral linkages, highlighting the importance of this methodology for identifying the activities that generate the most employment. Others include the approach of elasticities and intersectoral linkages to perceive key, driving, and strategic sectors. Therefore, this research is consistent with the methodology that seeks to discover the main economic activities that are generating employment for the Mexican economy, which is very useful for the design of active policies in the labor market focused on education and training with programs that improve employability in the most dynamic and interconnected activities.

Estimated direct and indirect employment generated by the Mexican economy

The importance of economic activity in the generation of employment can be identified by its volume, establishing:

This differentiation is relevant because there are subsectors, on the one hand, with a high level of direct employment generation and that therefore stand out in job creation; nevertheless, they may have low inter-industrial linkages, which will be reflected in a lower ratio of indirect employment. On the other hand, others register greater inter-industrial linkages, which will present higher levels of indirect employment in relation to the direct employment generated by the activity but which, in terms of the number of jobs, is not as significant (^{Sánchez, 2015; 11}).

Estimates of total direct and indirect employment indicate that the largest share is in the Retail Trade (17%), Agriculture (10%), Business Support Services (8%), Building (6%), Legislative and Government Activities (4%), Educational Services (4%), Households with Domestic Employees (4%), and Food and Beverage Preparation Services (4%) subsectors. (Table 2).

Direct employment represents 70% of total employment, with the following major economic subsectors: Retail Trade (21%), Building (9%), Agriculture (7%), Legislative and Government Activities (6%), Educational Services (6%), Households with Domestic Employees (6%), and Food and Beverage Preparation Services (5%), which together account for 61% of total direct employment (Table 2, Figure 3).

Source: created by the authors based on information from the input-output matrix 2013

Figure 3 Direct and Indirect Employment 

Indirect jobs represent 30% of total jobs generated by the productive interrelation of intermediate demand. The subsectors that in volume generate the most indirect jobs are: Business Support Services (25%), Agriculture (18%), Repair and Maintenance Services (7%), Retail Trade (7%), Wholesale Trade (5%), Animal Husbandry and Farming (5%), Specialized Works for Construction (4%), Professional Services (4%), Food and Beverage Preparation Services (2%), and Food Industry (2%), which together account for 73% of the total indirect jobs in the economy. (Table 2)

The subsectors that stand out most in indirect employment as a ratio of total employment in their economic activity are Business support services (94%), Specialized work for construction (86%), Professional, scientific, and technical services (84%), Manufacture of products based on non-metallic minerals (77%), Animal husbandry and farming (71%), Repair and maintenance services (66%), Wholesale trade of groceries, food, beverages and tobacco (60%), and Agriculture (62%), while in the manufacturing sector, with approximately 50% of indirect linkage in the generation of jobs, the subsectors of the plastics and rubber industry, the chemical industry, and the manufacture of metal products are the most prominent (Figure 4).

Source: created by the authors based on information from the input-output matrix 2013

Figure 4 Share of indirect jobs in relation to total jobs in the subsector 

Estimation of employment coefficients, multipliers, and employment-product elasticities

There is a set of employment indicators for economic activities that can contribute to the construction of a public employment policy. The appropriate estimators for this purpose are the coefficients of direct and indirect employment, which are obtained from the matrix of direct and indirect employment requirements and the employment multipliers; additionally, the calculation of employment-product elasticities enables the identification of the economic activities with the greatest impact on total employment in the face of changes in the sector’s demand. This is very important because it indicates the potential employment in the event of increases in the final demand of the subsector.

These indicators are presented below, following the methodological approaches of ^{Arriaga et al. (2021)}.

Employment coefficients

Direct employment coefficients measure the employment requirement of each sector, i.e., employment per monetary unit of output.

The matrix product of direct employment coefficients and the Leontief inverse obtains the total employment matrix. The direct and indirect employment coefficients (CL) of each sector of economic activity are obtained from the column sum of the CL^T matrix of total employment. For each sector j, the corresponding column of the total employment matrix indicates the employment requirement toward sector j and the indirect requirements toward the other sectors per unit of j’s GVP. The indirect coefficients are calculated as the difference between the total and direct coefficients.

This results in

Where CLjT is the total employment of each j-th economic subsector (sums per column of the matrix resulting from the above multiplication). Thus, for each j-th subsector, the column of the CLjT matrix shows the total (direct and indirect) employment requirements toward the industry itself and the indirect requirements toward the other industries per unit of j’s GVP.

The indirect employment requirements (indirect coefficients) of employment toward the same sector (CLji) are calculated as the difference between the total requirements toward that industry (CLjT) and the direct requirements (CLjd):

With this indirect coefficient, it is possible to estimate how many jobs are generated in the rest of the sectors of the economy in the face of increases in the final demand of sub-sector j. The process shows that an increase in the final demand of subsector j generates an increase in the gross production of the subsector itself, which in turn leads to an increase in the demand for inputs necessary for the production of j. Thus, an increase in the demand for inputs leads to an increase in the gross production of the subsectors supplying those inputs chained to subsector j, which leads to an increase in employment and value added (remunerations) in those subsectors. (^{Durán and Castresana, 2016: 12})

The employment coefficient represents the number of jobs per million pesos of spending associated with final demand in the economic subsector. The following subsectors register the highest coefficients of total employment (direct and indirect). On the one hand, it is observed that the highest total employment generation is attributed to a group of subsectors where the source of job creation is the sector itself, such as Households with domestic employees (29.94), Agriculture (14.26), Repair and maintenance services (12.37), Services related to agricultural activities (11.60), Manufacture of textile products, except clothing (8.37), and Food and beverage preparation services (7.60). On the other hand, there are subsectors with outstanding total employment coefficients, where the generation of jobs is explained to a greater extent by the indirect coefficients due to their greater linkages in the subsectors. Such is the case of the Food industry (3.17), Animal husbandry and farming (3.28), Social welfare services (2.37), Manufacture of textile inputs and textile finishing (1.94), Wood industry (2.05), Electronic information processing, hosting (1.94). (Table 3).

Employment multiplier

The analysis of multipliers contributes to the study of economic impact. It shows how an increase in final demand generates an increase in production greater than the initial growth for each increase in production. This is associated with an increase in employment, which will depend on the direct employment coefficients.

Multipliers indicate how many jobs are generated in the rest of the sectors of the economy (indirect employment) when the demand for inputs of sector j increases. As the final demand of sector j increases, an increase in the sector’s gross production is generated, which leads to an expansion of the demand for inputs necessary for the production of economic activity j. The increase in the demand for inputs produces an increase in the gross production of the activities supplying those inputs, chained to sector j, which leads to an increase in employment.

Employment multipliers are obtained by dividing the total employment coefficients (direct and indirect) by the direct employment coefficients for each economic subsector:

The most prominent subsectors for their employment multipliers are Food Industry (4.10), Animal husbandry and farming (2.04), Electronic information processing, hosting (1.89), Other information services (1.82), and Textile input manufacturing and textile finishing (1.96). This means that the food industry subsector generates 1.02 direct jobs for each million pesos of investment, its employment multiplier is 4.1, and the total jobs are 4.19. It should be noted that there are economic subsectors with high indirect coefficients and employment multipliers; nonetheless, as their direct employment base is too small, they are not identified in the economic activities that generate the most jobs per million pesos of final demand spending.

It should be noted that the multipliers estimate short-run economic changes, leaving out long-run adjustments; therefore, the impacts identified are transitory. Their construction is based on input-output matrices based on inter-industry transactions occurring in five or more previous years. The above implies that if the structural relations indicated by the matrix are modified due to technological changes or the appearance of new products, the reliability of the impact analysis using the initial multipliers is diminished. (^{Ramos et al. 2017})

Employment-product elasticity

A relevant indicator for studying the economic impact on employment is the employment-product elasticity, estimated by considering the economic weight and the direct and indirect employment coefficients. This reveals the effect on total employment in the economy in the face of exogenous increases in the final demand of a particular economic activity (^{Sánchez, 2015:17}).

Elasticity of sector j

Where L_i and x_i is the employment and output of product i consumed by sector j, bij is the ij-th coefficient of the direct and indirect employment requirements matrix, and y_j is the final demand of sector j. (^{Sanchez, 2015:17})

Thus, a 1% increase in the subsector means it can meet sustainable final demand increases due to its economic weight. The subsectors that contribute the most to employment in the economy, in the face of 1% increases in their final demand, are Retail grocery, food, and beverage trade (93 750), Food industry (56 404), Building (48 762), Legislative and governmental activities (32 770), Agriculture (29 767), Educational services (27 404), Households with domestic employees (23 337), Food and beverage preparation services (22 593), and Transportation equipment manufacturing (21 623) (Appendix 1).

Direct and indirect linkages: key, strategic, driving, and independent subsectors

In order to detail the structural approach to input-output analysis, multiple studies analyze economic activities and their relation to the production cluster. On the one hand, there is the work of ^{Rasmussen (1956)} to classify economic activities according to their power and dispersion sensitivity, as well as the study of ^{Chenery and Watanabe (1958)} establishing classifications for economic activities according to the value of the technical coefficients and their forward or backward linkages and the contribution of ^{Hirschman (1958)} that establishes methodological and empirical criteria to identify the key sectors of an economy through forward linkages (FL) as activities that generate inputs for other activities and backward linkages (BL), or as those that induce the development of other economic activities that provide inputs. (^{Cardenete and López, 2012:41})

The inter-industrial relations of the economic activities are located with the so-called linkages. This section shows the linkages of the economic structure, which will enable a differentiated classification into key, driving, strategic, and independent (Table 4). Applying the methodology of ^{Rasmussen (1963)}, it is possible to capture both direct and indirect effects from the Leontief matrix, which is not possible in the model proposed by ^{Chenery and Watanabe (1958)} because it only captures direct effects.

The key subsectors have high FL and BL linkages, as they are strong demanders and suppliers of inputs. Strategic subsectors are characterized by having a lower demand for inputs; nevertheless, they supply inputs to other subsectors. They are subsectors that can generate production blockages in the event of demand shocks. The driving or strong drag subsectors have low FL linkages and very high BL, as they are drivers of the economy; they have high intermediate consumption and a supply of products that mostly supply the final demand. The independent subsectors consume a small number of intermediate inputs and dedicate their production to the final demand. These subsectors do not cause significant drag effects in the economic system, nor do they react significantly to the drag effect caused by variations in the intermediate demand of other sectors (^{Arriaga and González, 2019:51-52}).

With the application of the methodology, the estimates indicate that the key subsectors are the food industry, basic metal industry, manufacture of petroleum products, generation and transmission of electric power and gas, paper industry, and repair and maintenance services with high backward and forward linkages, i.e., they are strong demanders and suppliers of inputs and a pillar of the inter-sectoral flows. Although there are few key sectors, which may be a sign of structural weakness, it should be noted that the greatest weight of the economic structure is found in the driving and strategic sectors, which increases the possibility of the emergence of new sectors that allow for greater productive diversification in the future (Table 5).

The subsectors show differentiated results when the employment estimates are combined with the above classification. In other words, the key subsectors of the economy are not the largest generators of employment since their economic weight does not give this characteristic, which is due to intersectoral flows and their high connectivity with the rest of the economic activities. An interesting result will be identifying subsectors with high intersectoral flows and a high economic weight that generate direct and indirect employment (Table 6).

According to the presentation, a complete indicator for identifying employment-generating subsectors is the employment-income elasticity, which, added to the potential identified as strategic, key, driving, and independent subsectors, provides the appropriate elements for public policy design. The generation of indirect jobs and the multipliers show the capacity of the subsectors to have a greater interrelation with the rest of the economic activities, which is reinforced in the study with the classification proposed by Rasmussen. Finally, the employment-product elasticity incorporates the economic weight of the subsector and the generation of direct and indirect employment as the most complete indicator of employment response.

For this study’s purposes, it is worth highlighting the subsectors whose classification presents a greater productive connection in the economic structure and, therefore, register high levels of employment coefficients that respond to a greater extent to the indirect coefficients and economic weight. In this regard, the food industry, building, agriculture, food and beverage preparation, transportation equipment, construction of civil engineering works, clothing manufacture, and animal husbandry and farming are the most important subsectors to encourage through a public employment policy.