The idea of introducing price floors in agricultural sectors has recently resurfaced in public debate. Although price floors were a flagship measure of the Common Agricultural Policy (CAP) in the 1970s and 1980s, they were a source of inefficiency and left a bad taste in people’s mouths. This paper shows, however, that a price floor for raw materials can promote efficiency in sectors where farmers face buyers with Monopsony power – that is, buyers capable of exerting downward pressure on prices.
Presentation
Key Results
- Whilst a price floor imposed on a competitive market is inevitably ineffective, a price floor on raw materials can be effective in sectors where farmers face Monopsony power.
- In the cow’s milk sector (not certified organic or PDO), we show that French processors exercise Monopsony power when purchasing raw milk, buying it at a price that was – on average over the period 2003–2018 – 16 per cent below its marginal contribution to their profits.
- In such sectors, a price floor indexed to international agricultural commodity prices and taking into account processors’ production costs would lead to better remuneration for farmers and a reduction in processors’ margins on raw material purchases.
- The introduction of an effective price floor could, however, destabilise a sector in the short term and increase concentration within its processing sector in the long term, making its effect on consumer prices uncertain.
- Support for agricultural incomes provided by a price floor alone is limited by international competition. It can be supplemented by measures to support agricultural supply (subsidies, trade policy), the effectiveness of which it enhances by preventing the transfer of such support to actors further down the supply chain.
A price floor indexed to international prices would not allow for the smoothing of agricultural incomes; this could be achieved through the introduction of an insurance scheme.
Method and Data
The analysis is based on data at the level of dairy processing plants, where prices and quantities of raw milk – by department on the purchasing side and by product on the sales side – were recorded from 2003 to 2018. The data are provided by the Ministry of Agriculture (Annual Dairy Survey), FranceAgriMer (Monthly Dairy Survey), and the Ministry of Public Finance (FICUS, FARE, LIFI). Access to certain data used in this study was granted within the secure environments of the Centre for Secure Data Access – CASD (Ref. 10.34724/CASD).
The authors restrict the analysis to cow’s milk products that are not certified organic or PDO. Margins are estimated in two stages.
(1) Estimation of processing costs and margins
A so-called ‘production function’ approach, which is standard in the literature (De Loecker and Warzynski, 2012), enables us to estimate the marginal cost of processing raw milk into finished and industrial products for each firm. Combined with data on the fat and protein content of raw milk and each dairy product (Depeyrot, 2010), as well as prices and quantities, this method enables us to estimate the margins on variable costs for dairy processors.
(2) Separate identification of monopsony and monopoly margins
The existence of dairy ingredients then enables us to estimate separately the monopsony margins, known as ‘Markdowns’, and the monopoly margins, known as ‘Markups’. This identification is based on the fact that these ingredients are:
— substitutes for raw milk on the purchasing side, and alternative outlets for finished products on the sales side,
— traded at a price that firms regard as given.
The identification therefore relies on the arbitrage conditions faced by manufacturers when making decisions regarding the procurement of raw milk or ingredients on the one hand, and the production and sale of finished products or ingredients on the other. At equilibrium, firms using ingredients equalise the price of ingredients (observed) with the marginal costs of sourcing raw milk (comprising the price of milk, observed, and the opportunity cost inversely related to the markdown, unobserved). Similarly, firms selling ingredients equate the price of ingredients (observed) with the net marginal revenue of each finished product (comprising the price of the finished product, which is observed, and the opportunity cost or mark-up, which is unobserved). These arbitrage conditions allow markdowns and markups to be identified separately.
Rémi Avignon
Etienne Guigue
Partners
This paper was supported by the French National Research Agency (Investissements d’avenir programme, ANR-18-EURE-0005), the ERC Consolidator Grant 816638 awarded to Jan De Loecker, and INRAE.