Livestock structural changes in Mexico: An input-output analysis approach

Introduction

Mexico’s rural development model has evolved through two broad phases, 1940 to 1983 and 1984 to 2017, marked by a transition from public investment in input supplying industries to market-oriented policies that prioritize export production and the incorporation of large producers into agri-food value chains. The primary sector, once central to the economy, has progressively weakened under unstable and uneven policy frameworks and limited coordination among social protection, productive inclusion, and territorial development instruments, thereby reinforcing persistent cycles of poverty^1-5.

Over the past two decades, Mexico's economy has expanded at an average annual rate of 1.70 %⁶, despite the impact of the COVID-19 pandemic⁷. By the end of 2023, the economy reached record levels, resulting in significant structural changes. According to official data from INEGI (2024) ⁸, the agricultural sector has contributed between 3 and 4 % to the Gross Domestic Product (GDP), with the livestock subsector accounting for less than 1.5 %. These values are consistent with those observed in emerging and advanced economies, where agriculture and livestock typically contribute between 1 and 7 % of GDP. The sector remains strategically important because of its broad production linkages across the economy. In Mexico, Sector 11 (agriculture, animal production, forestry, fishing, and hunting) is linked to 718 of the 834 registered economic classes⁹ .

In Mexico, the largest multiplier effects reported for 2012 are associated with petroleum and coal products manufacturing on the supply side (multiplier of 21) and audio and video equipment manufacturing on the demand side (multiplier of 4.32), as derived from the national input-output matrix¹⁰.

Input-output (IO) analysis enables the study of structural economic changes and provides data to evaluate industries and their interrelations within the economy. Despite inherent methodological contradictions, it remains one of the most significant contributions to economics in the 20th century¹⁰. The IO model, introduced by Wassily Leontief in the 1930s¹¹, offers internal consistency as its main advantage¹². It assumes that each industry’s output requires specific inputs, including raw materials, inter-industry services, household labor, or government-provided amenities.

The output comprises a variety of sector-specific products and services. A conventional input-output table is based on double-entry bookkeeping, where column totals equal row totals. These tables can represent global, national, or regional economic systems¹².

Globally, Input-Output Model (IOM) has been widely used to analyze economic structures and serve as a fundamental tool for public policy. In Brazil, it was used to estimate the economic value of sea activities and their significance to the national economy¹³. The structural characteristics of the global agricultural embodied land-transfer network were comprehensively analyzed, identifying key economies and sectors and revealing multidimensional pathways¹⁴. The importance of interregional economic transactions and related resource transfers at fine spatial scales, especially in large countries like China with pronounced disparities in development and resource availability, was identified¹⁵. Numerous other studies illustrate the versatility and importance of IOM, including analyses of carbon tax impacts across nearly all countries worldwide¹⁶. Evidence indicates that, in the Netherlands, the direct effects of multifunctional agriculture are modest relative to primary agriculture¹⁷. In Mexico, this approach has been used across sectors to examine local and regional economic structures and to guide public policy in agriculture, automotive, steel, and high-tech. Some of these analyses are used for the identification of strategic subsectors in others states like Tamaulipas¹⁸ or the impacts of certain subsectors over the economy like oil surpluses¹⁹.

Accordingly, this study estimated supply and demand multipliers and productivity indices for selected livestock branches in 2008, 2012, and 2018, in order to characterize structural change in Mexico’s livestock sector.

Material and methods

This study was based on a historical dataset from INEGI, comprising variables like production, paid prices to the producers, total production value and growth rates among others that reflect the performance of major livestock systems in Mexico in the period from 2008 to 2018. All calculations including operations with matrices and indexes were made using Microsoft Excel(and were obtained from INEGI⁸.

To analyze the data in the IOM, a series of matrix operations were conducted to calculate product multipliers, which are classified as supply or demand multipliers. Supply multipliers are derived by summing the rows for each branch in the previously estimated Leontief Inverse Matrix²⁰, while demand multipliers are obtained by summing the corresponding columns.

Model

This study defined the livestock sector to include the following economic branches: 1,121 cattle farming, 1,122 pig farming, 1,123 poultry farming, and 1,124 sheep and goat farming. Branches 1,125 aquaculture and 1,129 other animal production are excluded from the analysis, based on the North American Industry Classification System (NAICS) for Mexico²¹.

Technical coefficients (A), derived by dividing each sector’s intermediate consumption (IC_ij) by its gross production value (GPV_ij), i.e., IC_ij/ GPV_ij²².

Where: A= technical coefficients; = Intermediate consumption of branch _i demanded by branch _j; = Gross production value of branches _i and _j.

The Leontief matrix (L) was derived by first constructing the identity matrix (I)²³, and then subtracting the technical coefficient matrix (A), as expressed by: L = I - A.

The determinant of the Leontief matrix was subsequently calculated using the expression M= X(L), where M is the Leontief matrix determinant, X is a scale factor greater than zero (X>0), and L represents the Leontief matrix. The inverse Leontief matrix [I−A]⁻¹ was computed to uncover the reciprocal influences within the industrial structure and to determine both the direct and indirect requirements of each sector²².

To compute the inverse matrix, cofactor matrices were first calculated, which is obtained by multiplying (-1)_i+j by the determinant of a submatrix known as a minor, denoted by [M_ij]. Each minor was obtained by removing the corresponding row and column from the original square matrix L of order n, producing submatrices of order n−1. This process was repeated for each element of matrix L. Done this calculation the cofactor matrix was constructed²⁴. The inverse Leontief matrix (L) provides supply and demand multipliers, as well as economic chaining across the national economic structure²⁵.

Supply and demand multipliers

The calculation of multipliers can be classified into two types: backward chaining and forward linkages²⁶. The variables of each matrix were evaluated and calculated separately.

(I-A)^-1 = E. Where is the multiplier; I=Identity matrix; A=matrix of intermediate input coefficients; E= direct and indirect requirements.

Productivity Indexes

Productivity indexes (Pi) were calculated for each livestock chain^27,28 as follows:

Pi_t= (OSLB/TOALB) / (VSLB/TVALB). Where: Pi_t = productivity index for a specific time; OSLB= output of the specific livestock branch; TOALB= total output of all livestock branches; VSLB= value of the specific livestock branch; TVALB= total value of all livestock branches. Its meaning refers to those values greater than 1 represent productivity above the sector average, while a lower value refers to productivity below the sector average, becoming an analytical tool for evaluating sectors performance.

Results

Supply and demand multipliers

In terms of input demand, there is evidence of a decrease in its multiplier, suggesting that each peso invested in the livestock sector now generates fewer additional investments across the economy. This is partly due to a shift in input sourcing, with new inputs being required from sectors that were not traditionally linked to livestock production.

This trend becomes evident when observing the growth rates of the demand multipliers, which show a decrease of approximately 8 % for cattle and pig farming, 7 % for poultry farming, and 4 % for sheep and goat farming (Table 1).

This same trend is also observable from the supply side (Table 2), particularly affecting the cattle and poultry branches. In contrast, the pig, sheep, and goat branches have increased their capacity to generate new investments by expanding the amount of intermediate goods they provide to other non-livestock related branches.

These results highlight two distinct market dynamics: those primarily oriented towards external markets and those whose focus is the domestic market.

Productivity indexes

It is observed that, since 2008 to 2018, on average, cattle production has shown an index of 0.38. This suggests that its productivity has been lower than the livestock sector average, despite having the highest prices. In contrast, the average index value for the rest of the livestock branches showed a better performance. Table 3 summarizes the Pi for the period analyzed. Cattle production was the only branch of Mexico’s livestock economy to show a decline, while the remaining categories recorded Pi above the average.

Discussion

Supply multiplier

Shifts in multipliers are consistent with a reorientation toward the domestic market, driven by adverse external conditions and policies that sustained national demand, and with the well-known sensitivity of input-output models to their underlying assumptions. Structural changes can deepen inequality and social fragmentation, potentially amplifying socio‑economic and political polarization^29,30.

Livestock performance has been influenced by animal disease outbreaks, macroeconomic and microeconomic fluctuations, climatic variability, and the expansion of multifunctional production systems¹⁷. For example, African swine fever and porcine epidemic diarrhea have caused substantial production and economic losses³¹. While inputs for processing decreased, demand for feed inputs rose²⁸. ASF alone generated losses of US$108 billion in 2018³². Porcine epidemic diarrhea (PED) had worldwide repercussions and Mexico was no exception. Estimations rose US$ 2’264,143.89³³. In parallel, repeated outbreaks of avian influenza (LPAI and HPAI) in Mexico caused substantial productivity setbacks in 2012-2013. These shocks propagate through costs, input use, and productivity^34,35,36. The outbreak caused the death of over 22 million birds and an estimated US$ 55.6 million in losses³⁴.

In sheep, imports displaced potential domestic output; in pigs, medium and large farms exhibit both competitiveness and comparative advantages under shadow‑pricing assessments. Looking ahead, climate change threatens the availability of key feed inputs (maize, sorghum, soybean), animal health, water access, and pest pressure^37,38.

Mexican agriculture is heterogeneous, encompassing many small-scale producers with limited landholdings alongside a smaller group with substantial capital¹². The network of pig mobilization and slaughterhouses in the central region shows high cohesion and concentration of supply^39-41.

Public funding has primarily targeted rural social benefits rather than enhancing productive capabilities⁴², limiting the potential expansion of supply-side linkages.

Demand multiplier

Strengthening the domestic market requires lower production costs, higher quality, and better farm incomes to compete with imports. Following NAFTA, rising imports displaced an estimated 12.13 million head of potential domestic swine production by 2017. Mexico’s heavy reliance on imports of pork and feed grains has contributed to structural transformation in livestock systems^29,43,44.

The international market has been a key driver of rapid growth in poultry production, reflecting rising demand linked to poultry’s price advantage over other animal products^45,46.

Developing capabilities in any sector calls for public intervention to create a more competitive environment along value chains⁴⁷. Despite these policies, production conditions showed no material improvement, though Mexico maintained a positive agri-food trade balance driven by agro-industrial products⁴⁸.

Price volatility directly affects competitiveness. Mexico’s dairy sector is projected to continue expanding through 2030 in both volume and value. After sharp declines in 2023, international butter and skim milk powder prices are expected to recover gradually as input costs rise. In Mexico, nominal fluid‑milk prices may rise through 2027, but real prices have fallen, pointing to competitiveness pressures^49,50,51. World milk production is projected to increase by 1.6 % annually over the next decade, reaching 1,085 million tonns by 2033, primarily driven by improved animal productivity⁴⁵.

Production indexes

In contrast with multiplier trends, most branches recorded positive growth in production volumes and values during 2008-2018: dairy cattle (1.26 and 4.88 % per year), beef cattle (1.74 and 9.76 %), pigs (2.62 and 7.63 %), poultry meat (2.61 and 6.73 %), eggs (2.08 and 7.01 %), and sheep (2.0 and 7.47 %). Goat production declined slightly in meat (−0.79 % per year) and milk (−0.09 %). Cumulative inflation of roughly 46 % helps explain the larger movements in values relative to volumes⁵².

During the 2007-2012 period, National Development Plan aimed to boost meat production by 20.75 %⁵³. Nonetheless, production grew by an average of only 7 % across the different branches, with the largest increase in sheep and a 4 % decline in goats. The 2013-2018 Sectoral Program for Agricultural, Fisheries, and Food Development likewise focused its intervention strategy on maintaining and accelerating growth to ensure a sufficient supply of animal-based proteins for a growing population, by implementing measures and tools that encourage livestock operations to enhance both productivity and production⁵⁴. However, with insufficient results.

A comparison of the Sectoral Program’s objectives with the actual outcomes shows an average 11 % increase in overall production, highlighted by a 19 % rise in poultry production during that administration. Although goat farming continued to lag, it recorded a 0.5 % increase, contrasting with the previous six-year period when it declined. However, productivity indexes remained largely unchanged from previous levels.

The productivity index for the beef sector decreased, which contradicts Rebollar-Rebollar, et al⁵⁵. According to their findings, Mexico’s beef industry showed no signs of slowing down following trade liberalization and the implementation of NAFTA; rather, it was seen as a catalyst for competitiveness and productivity.

The productivity index for beef declined, in contrast to evidence suggesting that trade liberalization has been associated with competitiveness gains in that market. More broadly, agricultural programs have not corrected structural asymmetries or closed technology adoption gaps; their emphasis on income transfers, with limited integration into productive upgrading strategies, has diminished productivity impacts⁵⁵.

Limitations and directions for future research

This study is subject to limitations inherent to national input output accounts and the Leontief framework. First, input output tables provide economy wide averages and cannot capture regional heterogeneity, technological diversity within branches, or firm level behavior. Second, the approach is static and relies on fixed technical coefficients, thereby abstracting from short run input substitution, price adjustment, capacity constraints, and dynamic learning. Third, the benchmark years used here (2008, 2012, 2018) offer only discrete snapshots and cannot fully reflect high frequency shocks or policy changes occurring between observations. Fourth, because the accounts are monetary, they do not directly represent key biophysical flows for livestock systems, such as nutrient balances, water use, land requirements, or emissions.

Future work could address these constraints by developing interregional or multiregional input output accounts to identify spatial linkages and leakages, integrating hybrid physical and monetary satellite accounts for feed, land, water, and emissions, applying price responsive or dynamic extensions such as price input output, econometric input output, or dynamic multiregional input output models to relax fixed coefficient assumptions and evaluate policy scenarios, and linking input output results with farm level microdata to validate multipliers and productivity patterns at the production unit level.

Conclusions and implications

Evidence from input output multipliers and productivity indices indicates a structural reconfiguration of Mexico’s livestock sector between 2008 and 2018: backward demand multipliers declined across all branches, implying weaker economy wide stimulus per peso of final demand, while forward supply multipliers fell in cattle and poultry but rose in pigs and in sheep and goats, suggesting a split between value chains under stronger cost pressures and import competition and those more tied to domestic intermediate demand. Productivity patterns are consistent with this divergence, as cattle underperformed while other branches improved, pointing to persistent gaps in technology adoption, biosecurity, and value chain coordination amid recurrent shocks. Policy should therefore move from generic support toward targeted capability building, prioritizing technological upgrading and biosecurity through results based incentives and risk management, reducing exposure to input cost volatility via domestic feed strategies and improved logistics and storage, expanding standards, certification, and cold chain investments to strengthen market access, and scaling inclusive extension for smallholders that integrates finance, technical assistance, and producer organization. Integrating environmental accounting for water, land, and emissions with circular strategies would help align competitiveness gains with sustainability objectives. Together, these measures could strengthen forward and backward linkages, raise total factor productivity, and increase domestic value added while improving competitiveness, and the use of interregional and hybrid accounts would further refine targeting by locating linkages and leakages and monitoring whether investments translate into more resilient, higher productivity systems.