1Cattle farming has played an important role in the Brazilian economy for over 50 years. In addition to generating revenue for producers, it operates in supplying the domestic market and generates foreign currency credits due to its spillovers into the international market (TEIXEIRA; HESPANHOL, 2014). According to data from the Center of Advanced Studies in Applied Economics (2021), Brazil is the world’s second-largest producer of beef and the world's largest exporter – representing 24% of the total exported around the globe.
2Regarding dairy cattle activity, the Southern region of Brazil stands out with its three states among the five most productive in Brazil. According to the Brazilian Agricultural Research Corporation (2022), the state of Paraná is the second with the highest milk production in the country, while the other two Southern states Rio Grande do Sul and Santa Catarina stand in third and fifth, respectively. The Southern region’s contribution to the country’s production is approximately 35%. About the local concentration of milk production, the Northwestern mesoregion of Rio Grande do Sul is the most prominent location in the country, producing a volume of 2.9 billion liters. The West of Santa Catarina and the Southwest of Paraná are two other notable mesoregions, producing volumes of 2.4 billion liters and 1 billion liters, respectively.
- 1 Does not make distinction between beef and milk activities.
3General cattle1 farming’s employability in the Southern region is important for the local economy. Data from the Agricultural Census (2017) show that Southern Brazil had 671,451 people engaged in such activity, while soy production – the main activity in Brazilian agriculture – counted 395,506 people employed. This figure corresponds to 69.77% more workers employed in cattle farming than in soy cultivation. Compared with employability in other livestock activities, cattle farming in South Brazil employed 262.12% more workers than poultry and pig farming combined, as these activities employed 185,422 people.
4The analysis of livestock productivity is essential for several reasons. Among them are food supply, labor employment, economic fomentation, and better allocation of resources. Regarding the latter, Guimarães (2021) explains that the increase in livestock productivity generates a “land-saving” effect and thus allows agricultural production or vegetation maintenance on these lands.
5Due to cattle farming’s importance in supplying the country’s domestic market and for exports and income generation, this study investigates whether there were changes in high productivity clusters of beef and dairy farming activities in Southern Brazil between 2006 and 2017. We also verified whether characteristic factors of livestock activity positively or negatively influence the productivity of beef and dairy cattle. This provides evidence for the discussion of food policies on foods derived from livestock farming, making it possible to generalize to regions that are players in this trade.
6We apply spatial regressions to data from the 2006 and 2017 Agricultural Censuses and Municipal Livestock Survey editions. We based the choice of explanatory variables that somehow influence livestock productivity on the literature on the subject. These are factors related to technical orientation, area, agroindustry, expenditures, labor, and financing. In general, the literature on livestock productivity indicates the possibility of positive and negative effects of these variables on dairy and beef cattle productivity. However, a few studies are showing that these effects occur locally as well.
7This research presents seven more sections besides this introduction. Section 2 addresses historical issues of cattle farming in South Brazil; Section 3 offers the literature review on the determinants of cattle farming productivity; Section 4 describes the methods applied in this study; Section 5 analyzes the data collected; Section 6 presents the results; Section 7 shows food policy implications; Section 8 highlights our final remarks.
8Cattle farming in South Brazil has been documented since the 17th century. It was stimulated mainly by the Jesuit priests settling in the country’s extreme South (TEIXEIRA; HESPANHOL, 2014). Due to the group’s great political power – based primarily on their catechetical actions – the Jesuits were ordered to leave South America, and their now vacant establishments made space for the first farms in the region.
9Farmers that pulled through the 18th century were then able to take advantage of the substantial number of animals left by the Jesuits (KRONE; MENASCHE, 2016). By the end of the century, the migration of cultures from other regions marked the emerging of the charqueadas in the Southern state of Rio Grande do Sul: a practice that changed the dynamics of cattle farming and elevated the trading conditions of establishments in certain parts of the state (MARQUES, 1990).
10Access to land ownership was essential to cattle farming and producing food of animal origin. According to Fernandes and Miguel (2016), in the 19th century, the few opportunities for land ownership were available through leasing, inheritance, or donation of sesmarias – abandoned lots that once belonged to colonizers and were given for occupation.
11In this scenario, 19th-century family-led livestock strengthened its significance as an activity in the rural areas of South Brazil. Mainly in the region called Campanha Gaúcha, Southwest of the state of Rio Grande do Sul, the farming of small and medium cattle herds became the main production (FERNANDES, MIGUEL, 2016).
12Still in the 19th century, in Paraná, there were signs of cattle farming in the Campos Gerais region, as Canto et al. (2010) explained. In Santa Catarina, the occupation of Germans and Italians in Florianópolis led to the production of food and the farming of cattle purchased by the drovers (MATTEI; MATTEI, 2022).
13According to Trintin (2006), in the 20th century, due to the fragmentation of land in Rio Grande do Sul, there was a migration of gauchos to western Paraná. As a result, the cattle-raising culture became stronger in the region. The profile of the migrant families was characterized by resources that allowed them to quickly settle and begin subsistence activities. In addition to dairy cattle, they also raised poultry and swine.
14Several factors influenced the expansion of pasture areas in the state of Paraná. Among them was the severe frost of 1975, which negatively affected coffee cultivation—the main economic activity in the northern part of the state. In the following two decades, the pasture area in the state increased from 4.5 million to 5.3 million hectares (TRINTIN, 2006).
- 2 Creates the Veterinary Service, in the Ministry of Agriculture, Industry and Commerce, and approves (...)
15Eventually, elaborate public actions were set to encourage cattle farming. At the beginning of the 20th century, Decree nº 8.331/19102 established the Veterinarian Service of the Ministry of Agriculture, in addition to the creation of refrigerated parks. However, as pointed out by Medeiros Neto (1970), it was only in the 1960s that cattle farming was given more notoriety: new breeds were brought to Brazil and with that the production became more dynamic.
16Using data from the Agricultural Census, Fontoura and Pizzato (2009) observed the evolution of the cattle herd in Rio Grande do Sul and found that there was an increase of 58.74% in the number of cattle between 1960 and 1980. At the end of the 20th century, for 1990 and 2000, Fernandes, Bressan and Verneque (2004), identified an increase in milk production of 50.3% in the Southern Brazilian region. The increases by state were: 55.1% in Paraná, 54.3% in Santa Catarina and 44.8% in Rio Grande do Sul. Concerning bovine milk productivity, the “Liter of milk/Number of cows” ratio increased by 46.7% between 1990 and 2000. Increases were of 46.4% in Paraná, 50.6% in Santa Catarina, and 45.7% in Rio Grande do Sul.
17The industrial sector in Paraná benefited from the expansion of livestock farming in the state. The participation of livestock-related activities in the value added shows that, in 1998, animal slaughtering, preparation of preserves, milk cooling, and dairy products already accounted for 39.21% of the value added. Meanwhile, processing, milling, roasting, and manufacturing activities—mostly linked to agricultural rather than livestock production—accounted for 27.56% (TRINTIN, 2006).
18Therefore, it can be seen that, over time, livestock farming has consolidated itself as a structural component of the economy of Southern Brazil, shaping not only the region productive framework but also its cultural and social identity. The historical process, which began with Jesuit settlements, evolved into a complex and dynamic activity driven by migration flows, production needs, and technological advances. By the end of the 20th century, livestock farming was already deeply integrated into the regional industrial system, reinforcing its role in income generation and value addition across several areas of Southern Brazil.
19 As it is part of the agricultural activity, productivity in cattle farming may take advantage of elements that also influence agricultural productivity, such as area or access to information. Ostapechen and Gotardo (2019) analyzed the determinants of dairy productivity in Western Paraná and the results indicated a positive relationship between the technical guidance received by producers and the amount of milk produced per cow. On the other hand, there is an inverse effect on milk productivity when considering the level of illiteracy of the producer in charge.
20In a study of the determinants of dairy productivity in Paraná, Alves et al. (2020) corroborated the positive relationship between technical guidance and dairy productivity. They also found that the consumption of electricity boosts dairy productivity, as well as signs of a negative relationship between establishment area and productivity. The latter can be explained by the fact that most establishments dedicate most of the area to primary activities, which reduces space for dairy farming and its required technology.
21Structural issues also influence productivity, especially since they are linked to the regional and historical aspects of a given region. In a study of dairy farming in the state of Santa Catarina, Mattei and Mattei (2022) explain that the West is the most productive region in the state. That establishes a causal relationship with the location of large dairy companies, stimulating the local market and its labor absorption.
22In terms of beef cattle productivity, the main productivity indicator consists of the kilogram carcass/hectare ratio, since it covers three stages of production: breeding, rearing, and fattening (ZIMMER et al., 1998). Other works use similar, but not identical, metrics. Stocco (2018) used the number of cattle heads per hectare ratio to analyze the economic impacts of reducing Brazil’s beef cattle productivity gap. The research suggests that once the productivity gap is reduced, there is optimization in the use of land followed by a positive impact on the Gross Domestic Product (GDP).
23Carvalho et al. (2006) explain that livestock productivity may also be related to pasture since the low level of conserved forage and inferior pasture quality contribute to reduced productivity, increased feed costs, and the vulnerability of systems under unstable weather conditions.
24Zen et al. (2018) used data from the 2007 and 2017 Confederation of Agriculture and Livestock (CNA) to identify that there was a 22.2% increase in the average productivity of national cattle farming when analyzing arrobas produced per hectare used. Additionally, they observed that improved pasture management, reduction of input costs, integration of livestock with the crop, and information management may be aligned with improvements in livestock productivity.
25Barioni et al. (2008) define livestock productivity by its stocking rate, which must be determined by climatic factors, access to inputs, economic context, and pasture management. However, the decision to invest in greater productivity must be correlated with the economic perspective, for the return on the investment in raising animal productivity tends to occur earlier than the return on investment in pastures.
26Moreover, livestock activities require employing labor in production, considering that cattle farming entails disease control through the application of vaccines and cattle management. The importance of labor is also included in the effective operating cost (LOPES, CARVALHO, 2002; ARAÚJO et al., 2012).
27Lastly, regarding access to credit: as the agricultural product of a given year is influenced by the agricultural product of the previous year and the stock of financing is also a function of credit in past periods, access to credit may foster higher production values through each credit injection (MELO et al., 2013).
28Table 1 presents the variables selected for this study, based on the literature on the determinants of livestock productivity. We use data from the 2006 and 2017 Agricultural Census and the Municipal Livestock Survey.
- 3 For readers unfamiliar with the spatial approach, the Moran’s Index (Moran’s I) measures the degree (...)
29The econometric approach of this study is spatial. First, we verify the spatial correlation3 between spatial units using Moran's I statistics. Then, we apply Exploratory Spatial Data Analysis (ESDA) techniques in local (Local Indicators of Spatial Autocorrelation – LISA) and global forms. Finally, we apply spatial regressions with Kernel function weights to confirm which explanatory variable affects dairy and beef cattle productivity.
30According to Almeida (2012), the Moran's I statistic is written as:
31where n are the number of municipalities, z are the values of the variable of interest, Wz are the values of the variable weighted by the weight matrix and S0 is the sum of the elements of the weight matrix.
- 4 Considering that area is one of the explanatory variables, we adapted the PPC variable to the total (...)
|
Table 1 – Description of variables
|
|
Dependent variables
|
Description
|
References
|
|
PPL
|
Dairy farming productivity is represented by a thousand liters of milk produced and the number of cows milked ratio.
|
Fernandes, Bressan, and Verneque (2004)
|
|
PPC
|
Beef cattle productivity is represented by the number of cattle slaughtered and the total number of establishments ratio.
|
Barioni et al. (2008)4
|
|
Independent variables
|
Description
|
References
|
|
Technical guidance
|
Number of livestock establishments that received technical guidance, related to the total.
|
Ostapechen and Gotardo (2019); Zen et al. (2018); Alves et al. (2020)
|
|
Area
|
Area of total agricultural establishments in hectares, related to the total.
|
Parré et al. (2011); Santos and Lopes (2014); Alves et al. (2020)
|
|
Agroindustry
|
Number of establishments that obtained revenue from agroindustry products, related to the total.
|
Mattei and Mattei (2022)
|
|
Expenditures
|
Value of expenses incurred by livestock establishments in thousand reais (R$), related to the total.
|
Barioni et al. (2008); Zen et al. (2018)
|
|
Labor
|
The number of people employed in livestock establishments, related to the total.
|
Lopes and Carvalho (2002); Araújo et al. (2012)
|
|
Financing
|
Number of livestock establishments that obtained financing from a credit program, related to the total.
|
Melo et al. (2013)
|
|
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
|
32Nevertheless, the global form of Moran's I is not enough to capture local effects in spatial correlations. Anselin (1995), therefore, suggests applying LISA analysis, since it provides the identification of the local spatial correlation.
33We obtain Moran's I statistic for local analysis in univariate form using the following equation:
34In the case of spatial regressions, we initially apply the Ordinary Least Squares (OLS) method with spatial weights and then use Lagrange multiplier statistics to identify which model is the most appropriate: Spatial Lag (SAR) or Spatial Error Model (SEM).
35The SAR model is written as:
36The SEM model takes the following form:
37Although spatial regressions might capture spatial dependence, it is relevant to apply the Geographically Weighted Regression (GWR) approach so that the variation of the effects of the explanatory variables on the dependent variable is weighted along the spatial units. Due to the various issues that can bias the results in an average response, it is important to carry out a local analysis to correctly capture the characteristic effects of each spatial unit (ALMEIDA, 2012).
38The GWR approach allows regressions to avoid possible errors related to the increasing distance between spatial units. With that in mind, we utilize alternative non-Euclidean metrics instead of traditional Euclidean metric techniques (FOTHERINGHAM et al., 2002). The GWR results are obtained through the following equation:
39Due to lack of regularity in the spacing between observations, we opted for an adaptive band – usually more appropriate under these conditions (CHARLTON et al., 2009).
40That said, the direct relationships between explanatory variables and livestock productivity can be generally obtained through the following equations:
41where TG, A, AI, E, L, and F represent technical orientation, area, agroindustry, expenditures, labor, and financing aspects. The remaining terms consist of the lags of the explanatory and dependent variables. It should be noted that to avoid possible endogeneity problems, only elements without spatial lag were inserted in the GWR model.
42Figure 1 presents the spatial distributions of our livestock productivity variables. It is possible to observe an increase in the number of municipalities within the groups presenting the highest values for both dairy and beef productivity.
Figure 1 - Spatial distribution of dairy (PPL) and beef cattle productivity (PPC)
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
43After eleven years, despite the intensification of high productivity of dairy farming in other places in the South region, the intensification of productivity in places such as the "Great Mercosur Border" – the area corresponding to the Southwest of Paraná, West of Santa Catarina and Northwest of Rio Grande do Sul – is noteworthy. The region has been showing increases in dairy production since the 1990s, although there is still a need for structural improvement so that the dairy industry in this location is competitive at an international level (BEBER et al., 2019).
44It is also possible to identify signs of high productivity of beef cattle farming in Northwestern Paraná and Western Rio Grande do Sul after eleven years. It should also be noted the emergence of high productivity concentration of beef cattle in places such as the South-central region of Paraná and Northeast of Rio Grande do Sul.
45Table 2 presents the mean values of the variables that compose both cattle farming productivities, for the municipalities of each state in the Southern region. There was an increase in municipal averages of milk production and slaughtered cattle and a reduction in municipal averages of milked cows and agricultural establishments. Comparing states, it appears that after eleven years, the state of Santa Catarina has the highest municipal average of milk production, while the state of Rio Grande do Sul has the highest municipal average of slaughtered cattle.
Table 2–Mean values of the variables that composethe PPL and PPC, per municipality
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023)
46The increase in the number of factors that compose the numerator of productivity measures, as there was a reduction in the number of elements that correspond to the denominator, indicates the increase in productivity of both livestock activities. Ferrazza and Castellani (2022), also found a decrease in the average number of agricultural establishments and cows milked whilst observing an increase in production after eleven years. According to the authors, this increase in productivity is a reflection of access to specialized technical information, animal genetic improvement programs, and increases in job formality within livestock activity.
47Table 3 shows the values of Moran's I statistic for choosing the spatial weighting matrix. We opt for the Queen matrix because it presented the highest statistically significant value for Moran's I statistic in all estimates.
Table 3 – Choice of the spatial weighting matrix
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
48Table 4 indicates an increase in the number of municipalities inside the spatial correlation clusters of the high-high (HH) and low-low (LL) productivity types in both activities after eleven years. Among the groups that have statistical significance, the high-high and low-low groups concentrate on the largest number of municipalities. Comparing activities, we notice that dairy farming has a higher concentration of municipalities inside the clusters of direct spatial correlation of its productivity than beef cattle.
Table 4 - Composition of productivity spatial correlation clusters
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
Note: Queen Matrix.
49The direct spatial correlation group of milk productivity contained 34.04% of the municipalities in the South region in 2006. That participation increased to 42.46% in 2017. In the inverse spatial correlation group, that number decreased from 3.44% to 1.60% during the analyzed period. In the case of the spatial correlation of beef cattle productivity, the number of municipalities inside the direct spatial correlation group increased from 23.36% to 25.89% after eleven years, whereas the inverse spatial correlation group reduced the concentration of their municipalities from 2.68% to 1.17%.
50When comparing the numbers of municipalities present in the high-high and low-low clusters, we observe that after eleven years in dairy farming, the high-high cluster now has more municipalities than the low-low clusters: 258 and 252 municipalities, respectively. On the other hand, in beef cattle, the low-low cluster maintained its predominance in comparison to the high-high cluster: 215 and 93 municipalities, respectively.
51The local spatial correlations in the productivity variables are presented in Figure 2, reinforcing that there is an intensification of high-high clusters (in red) after eleven years. Regarding the productivity of dairy farming, there was an intensification of the concentration of the high-high cluster in the “Great Mercosur Border”. The municipalities in these clusters have a high level of dairy farming productivity and consist of neighbors with a high level of dairy farming productivity.
Figure 2 – Univariate LISA for dairy and beef productivity
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
Note: Queen Matrix.
52Finally, an intensification of beef cattle productivity is identified in Western Rio Grande do Sul and Northwestern Paraná when focusing on local spatial correlations. The municipalities within these clusters exhibit high levels of beef cattle productivity and are surrounded by others with similarly high productivity levels.
53After finding statistical significance in the Lagrange Multiplier tests (Robust LMρ) in the four approached scenarios (Tables 5 and 6), we concluded that the SAR-type models were the most adequate for our global spatial regressions.
54Tables 5 and 6 present the effects of the explanatory variables on dairy and beef cattle productivity variables. About our SAR model, we present the total effects (direct and indirect effects are detailed in Appendix A). Due to the non-normality of the residuals, we estimated the coefficients using the Generalized method of moments (GMM). The non-rejection of the Anselin-Kelejian test in the SAR model regressions indicates that our models captured the spatial autocorrelation appropriately. We also estimate an SLX model for the robustness of direct effects; however, Akaike Criterion values were smaller for the models that have a spatially lagged dependent variable.
55The regressions for the entire study area show that financing and technical guidance had a positive relationship with dairy farming productivity in both years (Table 5). The spatially lagged dependent variable (dairy productivity) also had a positive direct effect on the dependent variable. On the other hand, area and agroindustry had statistical significance only in 2006, while labor and expenditures became statistically significant in 2017.
56Analyzing beef cattle productivity (Table 6), we notice that aspects related to the area, technical orientation, and spatial lag of the dependent variable showed a positive direct effect in 2006 and 2017, although financing started to have a negative effect. Agroindustry and labor had statistical significance in 2017, while expenditures were statistically relevant in 2006.
57Nevertheless, we still estimate a GWR model on the basis that our Akaike Criterion (AIC) values for local models were lower than AIC values for global models (Table 7). Table 8 presents the Diff of Criterion and the local majority effects that reinforce whether the signs found in the global models are repeated locally.
Table 5 - Direct effects of spatial regressions, dairy cattle productivity (PPL) as dependent variable
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
Note: Queen Matrix
Table 6 - Direct effects of spatial regressions, beef cattle productivity (PPC) as the dependent variable
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
Note: Queen Matrix
|
Table 7 - Diagnosis of global and local regressions
|
|
yi
|
Year
|
Regression
|
AIC
|
R²
|
|
PPL
|
2006
|
Global
|
40881.63
|
0.26
|
|
Local
|
40621.15
|
0.47
|
|
2017
|
Global
|
39138.28
|
0.98
|
|
Local
|
38698.45
|
0.99
|
|
PPC
|
2006
|
Global
|
40469.41
|
0.31
|
|
Local
|
40243.32
|
0.49
|
|
2017
|
Global
|
41724.50
|
0.43
|
|
Local
|
41464.99
|
0.59
|
|
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
|
58Table 8– GWR’s Diff of criterion
Table 8– GWR’s Diff of criterion
Source: own construction, data from the Agricultural Census and the Municipal Livestock Survey (2023).
Note: NVF = No variability found. NS = Not significant
- 5 Considering that the samples have 1190 municipalities and that there are three types of local effec (...)
59The majority of local effects5 indicate the direction of the explanatory variables’ impact on the productivity of both activities in most municipalities in the Southern region. Results from the GWR model estimation reinforce that most municipalities had a positive relationship between technical guidance and dairy productivity. Area, in contrast, had a negative relationship with dairy productivity in most municipalities in the Southern region, but its local effect on beef cattle productivity was mostly positive. We also noticed that, in 2017, expenditures and financing positively influenced dairy farming productivity in most municipalities. The same happened with the relationship between labor and beef cattle productivity.
60The majority of positive local effects of technical guidance on dairy farming productivity corroborate the findings of Alves et al. (2020): access to technical knowledge by milk producers makes them more likely to adopt practices that avoid decreases in productivity. It should be noted that although the majority local effect of technical guidance on beef cattle productivity is not significant, there was still a relevant number of municipalities showing positive local effects, mainly in the state of Paraná (see Appendix D and E).
61In 2017, the area had a majority negative local effect on dairy farming productivity and a majority positive effect on beef cattle productivity. These effects can be explained by the fact that dairy farming is an activity that is often performed in reduced areas compared to other activities. According to Santos and Lopes (2014), confinement in milk production allows the activity to be conducted in reduced areas, although good infrastructure is required. In addition, issues related to agricultural establishment areas can lead dairy production to other activities (PARRÉ et al., 2011).
62Conversely, the majority of positive local effects of the area on beef cattle productivity highlight its importance to this industry. In accordance with Esteves et al. (1998), the “land” factor increases the profitability of beef cattle activity due to the increase in production – especially when aligned with an intensive pasture rotation system.
63Financial effects, such as costs and financing, had a majority positive local effect on dairy productivity, which reinforces Barioni et al. (2008)’s findings: considering that dairy productivity may be determined by optimizing the stocking rate, economic factors are essential to optimize the stocking rate of the livestock establishment.
64About financing and beef cattle productivity, our SAR model had already given indications that the relationship between the variables was no longer positive after eleven years (Table 5) since access to credit could be directing livestock from beef cattle farming to other activities. During this period, including in the state of Paraná, there was an increase in the use of financial resources for the succession of crops, such as soy and off-season corn. In the analysis of local effects, even though most municipalities showed a non-significant effect, the negative relationship between access to financing and beef cattle productivity (Appendix E) converses with the effect found in the SAR model. That was detected for locations like Western Paraná and the coast of Santa Catarina.
65However, we relate financing’s non-significant majority local effect with the fact that beef cattle farming is a conservative culture: producers prioritize the use of their financial resources rather than third-party access (MACEDO, 2006). Similarly, Silva Neto et al. (2011) analyzed the relationship between the regional distribution of credit resources and cattle herd growth. Their findings show that there might not exist a direct causal relationship between the two variables because, in certain regions, beef cattle farming may have reached a stage that no longer depends on rural financing.
66The mostly positive local influence of labor on beef cattle productivity also reinforces the variable’s importance for this culture. According to Araújo et al. (2012), labor in beef cattle activities is relevant because it affects production system management and, along with other inputs, composes the effective operating cost of beef cattle production.
67The increase in livestock productivity, and consequently, the availability of foods such as milk and meat, is conditioned on the actions of policymakers in this activity, according to Matte and Waquil (2018), the reduction of the vulnerability of livestock farmers in the market, and therefore as a consequence, the increase in productivity of this activity is the effect of government actions and extension and research agencies. As corroborated in this study, elements such as technical assistance, availability of land use, credit and labor are fundamental to increasing productivity and must be considered in the development of public policies in this activity.
68The provision of technical information to producers is necessary for the adequate allocation of inputs used and the solution of adverse conditions in this activity, such as the limited availability of area for livestock management. According to data from the Agricultural Census (2017), less than half of livestock establishments in the southern region of Brazil receive any type of technical guidance, corresponding to 43.85% of livestock establishments. While technical guidance provided by a government entity in the region corresponds to only 15.94% of livestock establishments, the remainder being provided by cooperatives, companies or private initiatives.
69In view of this, even though it is one of the most productive regions in Brazil in terms of livestock products, the intensification of public technical assistance policies would further encourage an increase in productivity in livestock activity, since there is scope for the advancement of technical assistance in livestock activity in this region.
70The availability of credit for dairy producers provides a greater degree of security for them when compared to a scenario of low credit liquidity. According to Silva et al. (2017), the rural credit system aimed at livestock farming in Brazil can be allocated to funding, investment and commercialization. At each of these stages, the producer may suffer from adverse and unexpected market conditions. In the costing phase, the producer is more vulnerable to production limitation issues, such as climate change. Producers' access to rural credit must be facilitated for their productive security, however, respecting certain criteria. The Central Bank of Brazil requires at least seven conditions for rural credit to be released, which include: the producer must prove its suitability, present a budget, plan or project and be inspected by the financier.
71Specifically, in the case of beef cattle productivity, another important factor found in this study is labor. Policies for formalizing labor in livestock farming must be promoted so that this element has ample conditions to work in beef cattle production. One of the causes of rural evasion in Brazil in recent decades was the emergence of better working conditions in urban areas. According to data from the National Household Sample Survey (2022), even with the evolution of labor formalization in rural areas of Brazil in recent decades, only 12.57% of livestock workers in the South region are in a formal labor market situation.
72Therefore, corroborated by the study and discussed in this section, policies to increase livestock productivity in Brazil, and consequently, the production of foods such as meat and milk, must consider and prioritize issues such as technical assistance to producers, availability of credit and formalization of labor in livestock farming. As observed, mainly in the technical assistance and labor variables, livestock farming in the southern region of Brazil has room for respective intensification.
73This study aimed to analyze how some of the productivity factors often discussed in the literature determine dairy and beef cattle productivity in Southern Brazil. Between 2006 and 2017, we identified the intensification of the productivity concentration for these two cultures in the region. There was a greater concentration of dairy productivity in the West of Santa Catarina and in the North of Rio Grande do Sul, while there was an increase in beef cattle productivity in the Northwest of Paraná and the West of Rio Grande do Sul.
74Our local spatial analysis allowed us to observe that certain variables had a type of global direct effect on productivity, but the majority of local effect was different. The majority of positive effects of technical guidance, expenditures, and financing on dairy productivity are in line with the literature on the subject, as well as the effect of labor on beef cattle productivity. The results referring to the area also show the difference in characteristics between both activities, since its majority local effect is negative on dairy productivity, but it is positive on beef cattle productivity.
75This study contributes to the literature on the following basis: identifying the intensification of different livestock productivities in South Brazil, corroborating the literature with global direct impacts of the variables on productivity, and finding different majority local effects in distinct parts of the Brazilian Southern region. Therefore, it highlights the need to identify local characteristics of dairy and beef cattle farming activities in case agricultural policymakers aim to increase even more the productivity of these activities. Such exercise of identification is of utmost importance for the allocation of regional policies in certain localities, such as access to financing, technical guidance and formalization of the job market for livestock farming.