The long-term ineffectiveness of transfers: the trillion € waste
Since the beginning of OECD calculations in 1986, EU agricultural policies have transferred about €2.5 trillion from taxpayers and consumers to farmers. If an interest rate of 5% is assumed (a conservative figure compared to the social returns from investments in education and research), the present value of these transfers rises to about €5 trillion. If the agricultural subsidies and tariffs between 1957 and 1985 were added, for which no OECD data exists, we would get two-digit trillion numbers. But this money does not raise farm wages in the long run.
I have calculated these figures by taking the OECD Total Support Estimate (‘the annual monetary value of all gross transfers from taxpayers and consumers arising from policy measures that support agriculture’) and subtracting those major payment categories that have more to do with public goods than farm incomes. These categories are On-farm services, Payments based on non-commodity criteria, Research and development, Agricultural schools and Inspection services.
The reason for why transfers to farmers do not increase farm incomes is labor mobility: the higher farm incomes are, the less labor leaves and the more labor enters the agricultural sector. In the short run, labor markets are not efficient and farmers lose agriculture-specific skills and investments if they quit the sector. Thus the amount of agricultural labor is not perfectly responsive to farm incomes, and transfers raise farm incomes.
But in the long run, especially as existing farmers retire and young people decide on whether to begin farming, transfers to farmers simply change the amount of agricultural labor but not its remuneration. This means that without the CAP, we would have fewer farmers today – and also fewer poor farmers.
The Single Farm Payment is a special case because it is mostly not a subsidy for agricultural labor but for the owners of historic entitlements and land. But to the extent that is actually a subsidy for agricultural labor (due to the labor-intensive requirements of cross-compliance), the same analysis applies: more labor is employed in agriculture but not at better rates.
One can argue that agriculture needs a high share of low-skill labor and that a larger farm sector thus drives up low-skill wages across the economy. But given the small size of agricultural employment and the increasing skill-intensity of agriculture, this effect can only be minor.
One could also claim that the CAP has increased skill-intensity, and thus wages, in agriculture due to higher capital investments that require better skilled farmers. But this would mean that more low-skilled labor needs to find employment in other sectors and that the economy-wide poverty rate is not directly diminished.
There are certainly second-order effects of all kinds. But the prime long-term effect of transfers that affect farm labor (rather than land and capital input prices) is to increase the amount of agricultural labor but not agricultural wages.
(I had a methodological hesitation about applying an interest rate to transfers from consumers to producers. But consumer welfare would have remained the same if the tariff that caused the transfer had been replaced by a consumption tax, and the money thus raised for the government could have been invested in alternative uses - thus justifying an interest rate to account for opportunity costs. Please send me a note if you disagree with this procedure.)