Resumen. Este trabajo investiga los efectos de la concentración de mercado upstream y downstream sobre el margen precio-costo de los minoristas usando datos bimestrales durante el período 1989-1992 desagregados por tipo de minorista y producto. Adicionalmente a la concentración horizontal, la diferenciación y los factores de costo, el análisis incluye el poder del comprador entre los determinantes de la rentabilidad de los minoristas, como resultado de la negociación vertical. Usando un modelo de efectos fijos en primeras diferencias, encontramos evidencia de actividad de negociación entre las grandes cadenas y los productores de alimentos. Nuestro análisis de la competencia de precios al nivel minorista también revela una cierta interdependencia en las decisiones de fijación de precios de las organizaciones minoristas más grandes y liderazgo en precios por parte de las grandes tiendas independientes.
Palabras clave: márgenes precio-costo, negociación vertical, competencia de precios, comercio de comestibles italiano, datos de panel.
Clasificación JEL: L13, L81, L42, C78, C23.
Abstract. This work investigates the effect of upstream and downstream market concentration on retailers' price-cost margins using bimonthly data over the period 1989-1992 disaggregated by retailer type and product. In addition to horizontal concentration, differentiation, and cost factors, the analysis includes buyer power amongst the determinants of retailers' profitability, as a result of vertical bargaining. Using a fixed effects model in first differences we find evidence of bargaining activity between large chains and food manufacturers. Our analysis of price competition at the retail level also reveals some interdependence in the pricing decisions of the largest retail organisations and price leadership by large independent shops.
Key words: price-cost margins, vertical bargaining, price competition, Italian grocery trade, panel data.
JEL classification: L13, L81, L42, C78, C23.
The European grocery trade has been subject to an intensive process of consolidation and modernisation in the last twenty years, although at a different pace in different countries. Since the mid-1980s Northern European countries (UK, Scandinavia, France, and Germany) have been developing a modern retail distribution system, dominated by a few large companies with a relatively small number of large-size outlets. The modernisation of food distribution in these countries has been accompanied by a shift in the balance of power from manufacturers to retailers (Tordjman, 1994). One of the most evident effects of this has been an increased ability of the most powerful retailers to impose restrictive vertical arrangements to manufacturers (Shaffer, 1991; Dobson and Waterson, 1997; Clarke, Davies, Dobson, and Waterson, 2002). The behaviour of grocery retailers and the increasing level of concentration in these countries have raised competition authorities' concerns, both at the national and European level (e.g., Kesko/Tuko merger case in Finland and Competition Commission, 2000, in the UK).
In Southern European countries (Greece, Italy, Spain, and Portugal) the modernisation process in grocery retailing has progressed at a slower pace. The retail distribution systems in these countries are still characterised by low levels of concentration and by a significant presence of traditional outlets. The Italian grocery trade, in particular, has been dominated by a large number of small local shops, but has been subject to some relevant changes in the last twenty years, which are reflected in a 40% reduction in the number of outlets between 1970 and 1990, and by a further 50% reduction between 1991 and 2000. Furthermore, between 1980 and 2000 a ten-fold increase was observed in the number of large supermarkets and hypermarkets (Euromonitor, 1980-2003).
This work investigates profitability in the Italian grocery trade, taking into account the nature of vertical relationships between firms in the retailing and manufacturing sectors. The aim of the work is to investigate the nature of grocery retailers' pricing behaviour and to identify any evidence of a shift in the balance of power in favour of retailers, similar to that observed in other European countries, despite the relatively low concentration levels in the Italian market.
In addition to the factors traditionally considered when investigating the determinants of price-cost margins, we shall allow for the possibility that the retailers' bargaining power vis-á-vis manufacturers affects their profit margins. Anecdotal evidence for the Italian grocery retailing (1) indicates that the large and centrally managed retail organisations conduct their bargaining directly with their suppliers over quantity traded and/or price levels. To reflect this type of vertical interaction our empirical analysis allows for the possibility of a bargaining process, rather than assuming arm's length pricing between vertically related industries. The empirical evidence leads us to conclude that the Italian grocery trade is moving towards the adoption of the practices observed in the US and Northern European distribution systems.
No systematic analysis of pricing patterns in the Italian grocery retailing has previously been carried out, due to the limited availability of detailed pricing information in the public domain. This work relies on detailed information on the retail prices disaggregated by retailer type and product level, based on scanner data collected on a regular basis by ACNielsen Italy. The availability of such detailed information allows us to provide evidence about some of the predictions in recent theoretical contributions on countervailing power (von Ungern-Sternberg, 1996, and Dobson and Waterson, 1997).
The data set used contains information about structural and behavioural features of 6 types of retailers from large, centralised multiples to small, independent shops. It covers 36 product markets for 19 time periods from 1989 to 1992. Panel data estimation techniques are used to account for the effects of retailer and product specific unobservable factors and to address dynamic pricing issues.
The structure of the paper is as follows: section 2 contains a brief description of Italian grocery trade, section 3 presents the related literature and the hypotheses underlying the estimating equation. Section 4 discusses the data and the estimation method while section 5 contains a discussion of results. Section 6 concludes.
The Italian grocery trade in the European context
The Italian grocery trade has been traditionally characterised by a large number of small and medium-sized independent shops. In 1995 a food store served on average 500 inhabitants in Italy and Spain, compared with about 1000 in Germany and 1700 in France and the UK. In 1997 Italy had 150 000 independent food retail outlets, three times as many as Germany and ten times as many as France and the UK. The number of independent outlets had halved by 2001, so that Italy now has about twice as many outlets as Germany and five times as many as France and the UK. Independent retailers represented about 45% of the market in terms of revenue share in 1997, a much lower share than in 1990 when they represented one third of the market (see Euromonitor, 1992-2003).
Although not as dominant as in Northern European countries, in Italy the centralised retailing organisations (comprising large chains, voluntary unions, and buying groups) supplied about 50% of the market in 1996, which represents a substantial increase from the 25% share of 1990 sales. Over the same period the market share of food sales in hypermarkets also doubled from 6% to 13%. The number of hypermarkets in France, Germany, and the UK was 1000 or more in 1996, while Italy and Spain only had about 250. Their market share ranges from 35% in France and the UK to 25% in Germany. The substantial growth in large retail organisations in recent years seems to indicate that the Italian market might be moving towards the Northern European system. The evolution of the Italian grocery trade during the 1990s was characterised by the emerging role of large multiples, of voluntary and buying groups. While in 1989 their joint market share was about 30% by 1993 it had increased to 55%, with buying groups doubling their market share from 12% to 25%. This increased concentration on the buyers' side of the market is likely to have been reflected in an increased retailer bargaining power, affecting the profitability of the largest retail organisations.
Despite the evidence of a modernisation process taking place in the Italian grocery retailing system, little is known about the impact of the changes on the pricing policies of different retailing organisations and in particular about the impact of the increase in buyer power of centralised organisations and voluntary groups. This work aims to provide evidence of the effect of seller and buyer power on food retailers' margins, over a period of 3 years at the beginning of the modernisation process in the Italian retailing system.
Price-cost margins and bargaining power in the literature
The main determinants of firms' price-cost margins have been identified in the industrial organisation literature and used extensively as theoretical reference in empirical work (for a survey, see Martin, 1993). Few attempts have been made to include the effects of vertical relationships on profitability (see Waterson, 1980, as an exception).
Traditional models of successive monopoly or oligopoly assume arm's length pricing between vertically related industries, ignoring the potential effects of buyers' power. In standard oligopoly models with conjectural variations a firm's price-cost margin (Waterson, 1980) depends on the firm's market share in the relevant market (which should capture the firms' power as a seller), on a measure of implicit collusion at the industry...