Technical Efficiency of Producer Cooperatives versus Private Firms: A Longitudinal Empirical Study^*

* We gratefully acknowledge financial support from the National Science Centre in Poland (decision number DEC‐2013/11/D/HS4/00252), from the FEDER (UNC315‐EE‐3636), from the Spanish Ministry of Science and Technology (2011‐ECO 2010‐22105‐C03‐C3 and 2014‐ECO2013‐45864‐P), from the Community of Madrid and European Social Fund (S2015/HUM‐34), and from the Spanish Ministry of Economy and Competitiveness (ECO2015‐65874‐P).

Abstract

In this paper, we hypothesize that producer cooperatives are more efficient than private firms, even when the impact of the political and economic environment is controlled for, and irrespective of the orientation with which efficiency is assessed; that is, whether firms target the reduction of inputs or the expansion of outputs. The empirical application focuses on a longitudinal population of olive oil manufacturing firms operating in the south of Spain from 1944 to 1998. The analyses clearly show that producer cooperatives are more efficient than private firms.

Notes

* We gratefully acknowledge financial support from the National Science Centre in Poland (decision number DEC‐2013/11/D/HS4/00252), from the FEDER (UNC315‐EE‐3636), from the Spanish Ministry of Science and Technology (2011‐ECO 2010‐22105‐C03‐C3 and 2014‐ECO2013‐45864‐P), from the Community of Madrid and European Social Fund (S2015/HUM‐34), and from the Spanish Ministry of Economy and Competitiveness (ECO2015‐65874‐P).

1. There is also hyperbolic‐oriented efficiency, which focuses simultaneously on increasing outputs and minimizing inputs. However, in this study, in addition to assessing the efficiency differences between cooperatives and private firms, we are also interested in determining whether any dissimilarities are linked to the orientation assumed. Therefore, we focus on both input‐ and output‐oriented models.

2. It is important to clarify that among private firms there is one specific group composed of sole proprietorship companies with a single owner‐shareholder and the same characteristics as other private firms.

3. The literature develops a vast array of other technical efficiency measures. However, the measures applied in this article satisfy a set of desirable mathematical properties, and are relatively easy to compute and straightforward to interpret, which means that they have been the most popular in practice. First and foremost, however, statistical inference is available for these measures and can be estimated using DEA via bootstrap methods. Hence, from these perspectives the measures applied in this article are superior to others.

4. This test is more general than the Wilcoxon, as it tests whether the entire distribution is different.

5. In particular, we do not possess firm level data for the quantity of olive oil produced or for the quantity of olives utilized. Also, the expense information for items other than capital and labor was not available.

6. For this we used the aggregated data from the official statistics for the total amount of olive oil produced and the total capacity installed for all olive oil firms in the Spanish province of Jaén from 1994 till 1998. This data allowed us to compute the indicator that expressed the number of tonnes of olive oil produced from every tonne of installed capacity. Multiplying this indicator by the firm's capacity level from our dataset, we obtained an approximate value of olive oil production at the firm level. Then we totaled up this production for all firms for each year and obtained the approximate value of total olive oil production for all the firms in the sample. This calculated production turned out to be very similar to the aggregated production reported by the official statistics.

7. It is worth noting that it is not possible to compute differences in efficiencies between cooperatives and private firms for every single firm, as the number of firms classified as cooperatives in the sample is different from the number of firms classified as private firms. As a result, the test of stochastic dominance was run for the data on the mean values of the differences in efficiencies between cooperatives and private firms for different time periods.

Additional information

Notes on contributors

Magdalena Kapelko

Magdalena Kapelko is Assistant Professor in the Department of Logistics at the Wroclaw University of Economics.

José Moyano‐fuentes

José Moyano‐Fuentes is Full Professor in the Department of Business Organization, Marketing and Sociology at the Universidad de Jaén.

Manuel Núñez‐nickel

Manuel Núñez‐Nickel is Full Professor in the Department of Business Administration at the Universidad Carlos III de Madrid.