Efficiency of the European Union farm types: Scenarios with and without the 2013 CAP measures

1.1 Literature review

Agriculture is a complex sector that depends on several factors and is called to respond to various challenges [13]. This often hinders the efficient use of the resources in the farms [14]. The education of the farmers, the location and size of the farms are variables, which sometimes, are found to influence agricultural efficiency [15] as well as the availability of land in specific contexts as those from the overseas islands [16]. The particularities of the agricultural systems and the characteristics of the farmers have been shown to impact farm efficiency [17].

Agricultural policies have implications on the use of resources and consequent outputs [16], including promotion of sustainable development [18], environmental efficiency [19], and organic farming which may bring interesting contributions [20]. The CAP measures have been designed to promote farm sustainability, reducing the environmental impacts [21]. The environmental impacts on agriculture in the EU is a concern for all stakeholders [22]. The relationship of agricultural sector with the environment is particularly relevant in the context of climate change [23] and its environmental consequences [24], where it is intended to meet the food needs and reduce the environmental impact [25]. Farm management through adjusted plans for better practices is another approach that contributes to more sustainable farming systems [26]. In addition, the limited number of studies in the EU context about the farm types hampers the analysis for this specific framework. In these plans, strategies of specialization/diversification [27], depending on the specific context [28], and dimensions associated with the farm design [29] should be considered in the decision-making processes.

In terms of methodology, to carry out an efficiency assessment at the farm level, the DEA is often considered [30] to define production frontiers [31] through benchmarks with the most efficient practices [32] and, in this way, identify the more profitable production patterns [33]. The DEA is considered to assess the relationship of several farm dimensions with efficiency, including the impacts from the CAP measures [31]. In some cases, the DEA is adopted jointly with other approaches to better address the specificities of the agricultural sector [34].

In general, researchers combining DEA with other methodologies aim to assess the environmental impacts [35] from farm activities [36], or at finding alternative activities [37] to deal with the consequences of climate change [38]. Another aim to identify new techniques of efficiency assessment is related to the needs to promote more eco-efficiency and sustainable intensification [39]. In particular, it is intended to make the systems more compatible with the increased requests for food [40] and the requirements to reduce the environmental implications [41].

Factor-cluster analysis is an example of approaches frequently used together with DEA [42] to assess the efficiency of agricultural systems [43]. Other examples are the policy analysis matrix [44], slack-based measures [45], life-cycle assessment [46], partial least squares structural equation modelling [47], and tobit models [48].

Usually, efficient use of energy in the farms deserves great attention [49], mainly because of its importance for agricultural performance [50] and its interrelationship with other dimensions [51]. The farms may also be an important source of renewable energies that may be produced with different resources and technologies [52,53]. Water is another farming resource, where efficient management is a determinant [54] for sustainable and integrated developments.

Using data at farm level from the European Union Farm Accountancy Data Network (FADN) database to assess the efficiency of the farming systems in the EU contexts through DEA approaches is rather rare, highlighting the relevance of this study. There are studies that consider some European cases [55], but there are not many studies considering the EU context as a whole.

3.1 Data analysis

The details shown in Table 1 were obtained by calculating the slope of a trend line that fits the different observations of each variable over the period 2004–2018, for different farm types. To calculate the slope, it was considered as the coefficient of linear regressions. The results for these regressions were considered only to assess trends over the considered period, because regressions to find the marginal impacts call for time-series estimation approaches. This table shows that over the last two decades (2004–2018) there has been a trend in several farm types for increasing crop productivities of the area (total crops output (euros/ha)), as well as the specific crop costs per hectare. This means that the increase in crop production was supported by increase in the specific inputs costs (seeds, fertilizers, and crop protection products). Similar patterns were followed by the livestock activities, with increasing trends for the total livestock output per livestock unit (LU) and respective specific livestock costs (feed) per LU. These findings should be considered by the policymakers in future CAP reforms, in relation to sustainability and environmental concerns. Some exceptions to the increasing trends were found in the case of COP and orchards of fruit farming systems (in the productivities of the LU) and olive systems (in the specific livestock costs per LU). The competitiveness of the farms (farm net value added (euros/annual work unit, AWU)) also presents an increasing trend for several farm types, with the exception of COP, sheep and goat farms. The farm net value added is the output minus intermediate consumption, minus depreciations and plus balance current subsidies and taxes. This indicator is used to assess the remuneration of the fixed factors of production (work, land, and capital). AWU is the annual work unit.

In contrast, the total productivity (total output/total input), in general, presents a decreasing trend, with the exception (because they are fewer in the table, but it does not mean that they are less important in the European context) of COP, other fieldcrops, wine, cattle, granivores (pigs, poultry, for example) and mixed livestock. The total output/total input is defined by the database as [56] (sales and use of crop and livestock + change in stocks of products + change in valuation of livestock − purchases of livestock + various nonexceptional products)/(specific costs + overheads + depreciation + external factors). The decreasing trends in the total productivity ratio verified for the majority of the systems reveal that, for example, the total output increased less than the total input. The stocking density (LU/ha) was also decreased with exception for milk and mixed livestock. The number of farm types with decreasing trends for the total LUs is greater than that of the total utilised agricultural area, explaining, in part, the decreasing trends for the stocking density. The labour input also presents a decreasing trend in the majority of the systems, especially in case of livestock systems (exception for granivores).

As expected, farms with horticulture are those that use most labour input, given the characteristics of horticultural farm types. However, there are signs that smart agriculture practices may improve the automatisation of the sector and reduce the dependency on labour [2]. In turn, for several systems analysed, 2007 and 2008 were the years with most labour used by the EU farms (Figure 1). The average agricultural area utilised decreased significantly after 2006, with some recovery in 2018. The farms of the COP systems are large (Figure 2). The number of LUs also decreased after 2006; however, the recovery was more consistent. The granivore systems are those with, on average, more LUs (Figure 3). Despite the increase in the LUs after 2007, the stocking density decreased since this year, with some recovery in the last years (Figure 4). The information presented in these figures are for farm types where sometimes crops and livestock productions are combined. On the one hand, these data are for representative farms obtained through weighting approaches. It seems that the CAP reform of 2003 implemented since 2005 had an impact here, considering the findings described before. On the other hand, it seems that the CAP reform of 2013 overcame these implications, especially in the recent years, where there was some significant recovery in the variables analysed.

Figure 1

Labour input (h), per farm, for several farm types, for the period 2004–2018.

Figure 2

Total Utilised Agricultural Area (ha), per farm, for several farm types, for the period 2004–2018.

Figure 3

Total livestock units (LU), per farm, for the farm types, for the period 2004–2018.

Figure 4

Stocking density, per farm, for several farm types, for the period 2004–2018.

The total productivity remained, in general, without significant changes (Figure 5). Crop productivity of the area increased, in general, over the period, with farms with horticulture systems having, on average, higher productivities (Figure 6). A similar pattern was found for the trends, across the diverse farming types and over the considered period of the livestock productivity (Figure 7), specific crop costs (Figure 8), specific livestock costs (Figure 9), and competitiveness of the farms (Figure 10). The statistical information presented in these figures highlight that the CAP reforms of 2003, for the total decoupling of the subsidies, and 2013, with some adjustments relatively to 2003, had their impacts on the structures of the EU farms, but in inverse directions. The 2003 CAP reform allowed farmers to progressively receive subsidies from the first Pillar decoupled from production and activities, albeit with some constraints. The 2013 CAP reform adjustments seem to promote the recovery of some variables. In addition, the financial crisis of 2008 also had implications for the competitiveness of farms. These findings highlight the implications of the CAP reforms and international events on the European agricultural sector evolution.

Figure 5

Total output/total input, per farm, for the several farm types, for the period 2004–2018.

Figure 6

Total crops output (€/ha), per farm, for the several farm types, for the period 2004–2018.

Figure 7

Total livestock output (€/livestock units), per farm, for the several farm types, for the period 2004–2018.

Figure 8

Specific crop costs (€/ha), per farm, for several farm types, for the period 2004–2018.

Figure 9

Specific livestock costs (€), per farm, for several farm types, for the period 2004–2018.

Figure 10

Farm net value added (€/AWU), per farm, for several farm types, for the period 2004–2018.

3.2 Efficiency analysis

In this Subsection, a DEA was carried out considering an extended Cobb-Douglas [64] production function and the developments from Martinho [6,65]. The total output (total of output of crops, livestock, and of other output) was considered, while inputs were labour cost, total utilised area, total LUs, total inputs (total specific costs + overheads + depreciation + external factors), total intermediate consumption (specific costs + overheads) total specific costs (seeds and seedlings, fertilizers, crop protection products, other specific crop costs, feed for grazing stock and granivores, other specific livestock costs and specific forestry costs), total farming overheads (machinery and building current costs, energy, contract work and other direct inputs), total external factors (wages paid, rent paid, and interest paid), total assets, and gross investment on fixed assets. The main objective of considering these variables as input was to assess the main inefficiencies associated with several farming inputs and their respective costs. To deal with the structural breaks from the several shocks associated with the CAP reforms and the financial crisis, only data for the period 2014–2018 were considered.

Table 2 presents the statistical summary of the several variables related to the 15 farm types, and Table 3 reveals that the several farm types considered are efficient (with values of 100%) in the use of the diverse inputs, except for the following farming types: other fieldcrops, mixed crops and livestock, and total. The main inefficiencies in these farm types are associated with the total area utilised for agriculture where could be possible to reduce about 50% of the area without affecting the level of output. The other fieldcrop system should be benchmarked, especially with the wine farming types, mixed crops, and livestock with the mixed livestock and total with permanent crops combined (Table 4). The permanent crops combined appear here with an interesting balance among output, inputs, and respective costs. In these tables, the total farming type, with column head “total,” represents not the sum of the 14 different farming types, but the representative farm type for the all EU context.

The consideration of the EU subsidies from the CAP framework eliminates the inefficiencies verified for other fieldcrops and improve the efficiency of mixed crops, livestock and total farm types (Table 5). In other words, the subsidies given to farmers since 2014 imply the use of more inputs, at least the use of inputs with higher costs, with consequences on the sustainability. In addition, these subsidies make the milk system as a better benchmark for the total EU systems (Table 6). These are interesting findings that deserve special attention by several stakeholders, especially by the policymakers, considering that the current world’s endeavour is to achieve more output with less resources.

3.3 Productivity and output growth

Considering the similarities between the levels of technical efficiency for several farm types verified in Subsection 3.2, it is important to go further and understand whether this pattern is followed by the evolution of competitiveness (total factor productivity). In this perspective, an analysis of total factor productivity and total output evolution was carried out, its results are presented in Tables 8 and 9 and Figure 11 (summary statistics are presented in Table 7). These results are based on the growth rates for the period 2014–2018, including, and not, the current subsidies (excluding those on investments). The data considered were obtained from the FADN [56] database and were explored with DEAP [61] software, following the procedures proposed by Coelli [63] for the Malmquist Index (input oriented) to obtain total factor productivity changes.

Figure 11

Evolution of the total factor productivity (Malmquist index) and total output growth rates (including subsidies), over the period 2014–2018, for several EU farm types.

On an average, total factor productivity, for the period 2014–2018, decreased for several farm types, except for horticulture that grew by 1.4% (Table 8). The worse performances were verified for the granivores, milk, and mixed livestock. In addition, there are great similarities among the total factor productivity changes with and without subsidies. The only slight differences are verified for the following systems: olives, permanent crops combined, and mixed crops and livestock. In these cases, total factor productivity growth rates became slightly worse with the subsidies. These findings confirm the results obtained above from the efficiency analysis. In fact, the subsidies from the CAP measures increase the use of inputs (or at least increased the costs with the use of inputs), in some cases, and worsen the performance and competitiveness of the farms (or at least do not improve them), which in a context of environmental impact mitigation deserves special attention. On the other hand, the total output increased, on average over the last five years, except for the permanent crops combined systems. The higher growth rates were verified for the milk and mixed livestock farming types. The values of 2018 were determinant for the average higher changes of the mixed livestock system (Figure 11). In general, the growth rates for the total output are higher without subsidies than with subsidies. These values reveal that the total output grew at greater rates than the level of subsidies.

As shown in Table 9, presenting the results for the period considered, on average, over the different farming types a similar trend is observed with that described in Table 8. It is worth observing that the negative growth rates verified for the total factor productivity improved from −23.6% in 2015 to −11.5% in 2018. The total output growth rates followed an inverse pattern, with 2016 as the worse year for the output growth.