Long-term structural change in EU agriculture

The reduction in the number of farms and farmers in the EU has attracted increasing attention in this EU legislative period. It has been a distinctive message of the Commissioner for Agriculture Janusz Wojciechowski, who emphasised the ongoing decline in the number of farms in his confirmation hearing before the European Parliament in October 2019 and has regularly highlighted the issue, most recently in his address to the DG AGRI Agricultural Outlook conference last December. It has also been a consistent theme in debates in the European Parliament and in statements by farm group representatives.

I have previously posted on structural change in EU agriculture. For example, in this post, I discuss changes in the number of holdings between 2005 and 2020 and ask whether there is evidence that the rate of decline in the number of holdings has changed over time. My answer was that no clear answer emerged from the relatively short period examined in that post.

In a recent invited talk to the first symposium organised by the European Network of Agricultural Journalists on the future of food and farming in Europe, I presented some data on longer-term trends in farm structure. I thought the figures are sufficiently interesting to write up as the blog post presented here.

Trends in the number of holdings from 1960 to 2020

The raw data on the number of holdings in each country over a 60-year period between 1960 and 2020 are shown in Table 1. I have deliberately confined the sample of countries to those in Western Europe, given that farm numbers in central and eastern Europe over this period have been heavily influenced by central planning and its subsequent collapse after 1989. The data are sourced mainly from FAOSTAT apart from some missing data compiled from other sources.

The trend in the number of holdings over time can be influenced by changes in the definition of what is considered an agricultural holding. Nonetheless, there can be no mistaking the sharp decline in the number of farms in all countries over this period. France, for example, had almost 1.9 million farms in 1960 compared to less than 400.000 today.

Table 1. Number of farms in selected countries, 1960-2020

Sources and notes: The basic source is FAOSTAT Agricultural Census data. Eurostat farm structure data only go back as far as 1990 and there is reasonably good agreement between FAOSTAT and Eurostat data after that date. Reported Census data do not necessarily correspond exactly to the years in the table, in which case the Census data have been assigned to the nearest decadal year. Missing data for Finland in 1960 have been constructed by averaging the number of holdings in 1950 and 1970. Missing FAOSTAT data for Germany in 1990 are sourced from the Eurostat domain ef_ov_kvaa. Missing data for Sweden in 1990 are sourced from the Swedish Yearbook of Agricultural Statistics 2020. The total number of holdings reported in the Yearbook only includes those with more than 2 ha of arable land, but the 1999 figure using the same definition is close to the 2000 Census figure reported in FAOSTAT. Missing data for the UK in 2020 are sourced from Defra, Agricultural Statistics in the UK. Only England among the UK regions limits its farm structure survey to ‘commercial farms’ as defined in the EU Farm Structure Survey Regulation EC 1166/2008. As neither FAOSTAT nor Eurostat includes a 2020 figure for the UK, the 2020 total is obtained by extrapolating the 2010 figure by the change in the number of holdings in England between 2010 and 2020. Both the figures for total UK holdings and holdings in England show a slight increase between these two years. Readers are cautioned that changes in the number of holdings over time can reflect changes in the definition of what is an agricultural holding.

There are clear differences in the size of the agricultural sector in each country. We can make comparisons between countries by converting the absolute numbers in Table 1 to indices where 1960 takes the value 100 (see Figure 1 below). This figure makes clear that, although farm numbers have declined in each country, there are differences in the extent of the reduction. With Greece (the brown line at the top of the diagram) a bit of an outlier, the scale of the reduction over the entire period varies from around 60% (the UK and Ireland) to almost 90% (Finland and Belgium, with the reservation that the 1960 starting point for Finland is an estimate).

Figure 1. Structural change in selected Western European countries, 1960-2020

I have divided the period of six decades into the two periods 1960-1990 and 1990-2020. The first thirty-year period was the period of the original CAP policy with high market support prices reinforced by high import tariffs and the use of export subsidies – what I call the ‘command CAP’ or often referred to as the ‘Fortress Europe’ policy. The MacSharry reform in 1992, implemented from 1994 onwards, introduced a more market-oriented policy in which support prices were lowered to safety net levels, producer prices were more aligned with world market prices, and farm incomes were instead supported by direct payments. I have called these last three decades the ‘Market CAP’ in Figure 1.  It is interesting to ask whether the rates of decline in the number of farms differ between the two periods.

Figure 2 presents the same data but this time the Y-axis measures the logarithm of the number of holdings in each country. Log scales show relative values rather than absolute ones. Thus, the height of each series ranks countries by their total number of farms, while the slope of the series now shows percentage growth or decline. The figure shows clearly that, for the three countries with the largest number of farms in 1960 (Italy, dark blue: Spain, light blue, and France, green), the pace of decline picked up in the second three-decade period associated with the ‘Market CAP’. This is also the case for Austria (dark orange) and Finland (purple).

However, if we look at the total sample of countries, in eight out of the 13 countries the rate of decline in numbers either remained the same or actually slowed during the period of the ‘Market CAP’ (In Table 2, I present the actual percentage changes in the different periods). In other words, these countries experienced a similar or even larger percentage reduction in the number of holdings during the period of ‘Command CAP’ than during the period of market liberalisation. The case of the UK is particularly interesting because that is the only country where we begin to see, in the last decade 2010-2020, an increase in the number of holdings. This is the result both of an increase in the number of large holdings but also of the very smallest holdings.

Figure 2. Comparative rates of structural change in selected EU countries, 1960-2020
Table 2. Percentage rates of change in number of holdings, 1960-2020. Rows marked in blue indicate countries where the rates of change in the later period were similar to or lower than the rates of change in the earlier period.


It is thus unclear whether, on balance, the rate of structural change in farming has increased in the later period compared to the earlier period or not. I also want to stress that the relationship between the rates of decline in the number of holdings and the different phases of the CAP is an association, and in any case it is not possible to conclude that there is a causal relationship. Factors other than the level and type of support under the CAP can determine the pace of structural change, such as the pace and nature of technological change, the rate of growth in non-farm incomes, the demographic status of the farmer population, or relative changes in the profitability of different farm enterprises with different land requirements. These factors may have acted with different intensities in the two CAP periods and contributed to the different rates of decline observed.  

The decline in the number of farms, facilitated by the number of farms without obvious successors, is clearly a factor behind the recent farm protests. Yet the continued labour productivity gap, and thus income gap, between farmers and non-farmers suggests that farm consolidation will continue. At the EU level, the 2020 agricultural census showed that almost two-thirds of the 9.1 million holdings in the EU are less than 5 ha in size. Many of these holdings are of a semi-subsistence nature and perform a valuable role in self-provisioning for the families involved. But it is unlikely that many of these have a future as professional farms and we must anticipate that many of these will cease to exist in the next generation. The tendency towards farm consolidation will be further amplified by the need for extensification to meet the Farm to Fork strategy targets, which will require larger farms to yield a comparable income.

It is not politically popular, but in planning for the future of European farming it makes sense to start from a realistic foundation. Rural areas will have fewer farms, and this will require strengthened rural development rather than agricultural policy interventions to maintain the viability and quality of life in these areas.

This post was written by Alan Matthews.

Update 18 April 2924: Textual changes made to clarify that the blue bars in Table 2 included countries where rates of decline in the number of farms were similar in both periods as well as where they were lower in the later period. Also, years were added to the X-axis in Figure 1.

Photo credit: Colin Paterson, Yorkshire Dales via Flickr under a Creative Commons licence.

Print Friendly, PDF & Email

2 Replies to “Long-term structural change in EU agriculture”

  1. Very interesting, Alan. I am intrigued by the “breaks” in the trendlines. My hunch would be that these are explained by specific national policies. Would be interesting to look into this a bit, I think.

  2. Hi Pieter, this seems to be a sensible suggestion. There was an interesting comment to my post on LinkedIn highlighting the sharp fall in numbers in Finland between 1990 and 2000 and which suggested this could be due to introduction of a pension scheme for farmers in the late 1980s. If someone had the time to follow up, it would be a very useful study.

Leave a Reply

Your email address will not be published. Required fields are marked *