Sample effect – how conclusive is a comparison with the previous year?

In 2009, the Farm Accountancy Data Network collected accountancy data for just under 3400 reference farms. Comparison with the previous year’s results is extremely important when analysing data, although the present study shows that percentage changes compared to the previous year can depend considerably on the particular composition of the sample. For example, in 2009, the agricultural income of all the reference farms in the mountain region fell by only 0.9 % compared with 2008, while comparable farms showed a corresponding reduction of 5.3 %.