Efficiency determinants in retail stores: a Bayesian framework
Introduction
The level of competition in the retail industry has intensified in recent years, driven by several factors such as the decline in household wealth, rising unemployment, tight credit conditions, rapid globalization, unclear economic picture, drop in consumers’ confidence index, and increase in merger and acquisition activity. In fact, it is very common today to find several retail stores competing for a slice of market pie that rarely varies in size and often saturated. Consumers have also become extremely price conscious, especially as retailers constantly fight for each dollar by offering substitutable products at lower prices [1], [2]. Weekly specials including catalogues and additional marketing are also used intensely to attract consumers and drive sales. While the winner of this price war is the consumer, retailers have to survive with lower profit margin which necessitates a stricter control of productivity and a decrease of operational wastages, especially as the industry is traditionally known to be labour intensive.
In the retail literature, the need for higher productivity has also been highlighted as one of the key issues for future survival. Numerous studies have recently appeared, addressing the performance challenges of several retail markets, such as Portugal, the US, and France [1], [2], [3], [4]. The key message from most of these studies is that efficiency measurement is of vital concern at both the store level and the strategic level, since it helps retailers differentiate themselves from other competitors. Studies dealing with some store-specific characteristics that affect retail performance have also appeared in the literature.
Within the growing importance of performance measurements in retail, more advanced methodologies have also been introduced. In contrast to the traditional literature where simple ratios were the most adopted productivity methods, significant volume of work currently exists in the area of measuring production efficiency by estimating an “efficient frontier” that serves as a benchmark for evaluating performance. The most adopted methods for that purpose include the stochastic frontier (SF) and data envelopment analysis (DEA) methods [5], [6], [7], [8], [9]. The advantage of both these methods is that they allow the use of multiple inputs and outputs in the measurement of performance, making them thus more suitable in the retail context. Although the DEA method is a nonparametric, deterministic approach that defines a relationship between multiple spending inputs and outputs by building an efficient frontier, it has been critiqued for not providing fit statistics such as r-square or p-value that can be used for statistical inferences [10]. The SF method, on the other hand, uses a parametric approach by explicitly taking into account the stochastic properties of the data [1], [19].
The aim of this study is driven by all of the above, and the motivation is to extend the existing literature by offering more accurate insights into the performance determinants of retail stores. Specifically, the study focuses on analyzing on the supermarket industry, which is traditionally known to be very competitive. Most supermarkets operate on a high inventory turnover usually with low profit margins, which suggests that retail mark up for each individual product is very low. In most countries, supermarkets compete against national, regional, local and independent supermarkets, specialty food stores, club stores, drug stores, convenience stores, discount merchandisers, and other local retailers. They compete on the basis of price, store location, product mix or types of products and brands and services. The industry’s small mark up price format indicates that price is extremely significant. In other words, supermarkets rely on low mark up and high sales volumes as opposed to higher mark up and lower volume of sales. Is this context, efficiency analysis constitute a useful and interesting tool to improve the profitability of supermarket chains.
Most existing studies in the literature have generally focused on the estimation of efficiency without providing an in-depth analysis to the factors that lead to efficiency variations between retail stores. The sole focus on efficiency makes the study also limited to one sample or one geographic area of analysis. However, as this study focuses on identifying the sources of efficiency, the results can be more applicable to other retail sectors, or even to the same retail sector in other countries. The methodology used in this study also provides an innovation to the existing literature. For the first time, we use the Bayesian methodology that has several advantages over the maximum likelihood (ML) traditionally used to estimate the SF approach. For instance a key advantage of the Bayesian approach is that it allows the inclusion of “prior” information about parameters in inferences. With Bayesian, the results are also usually presented in terms of probability density function (pdfs), making it thus possible to make probability statements about the model parameters.
In testing our hypotheses we use a sample of Spanish retail supermarkets. There are several interesting characteristics of the Spanish retail market that allow us to test our desired hypotheses. In the next section we provide a brief overview of the Spanish retail market. This is followed by the literature review, methodology, data characteristics/hypotheses, discussions, and concluding remarks.
Section snippets
Contextual setting
In recent years, the Spanish supermarket industry has been characterized by a series of changes that have affected its structure and performance [11]. Among them, four critical factors deserve special attention.
First, the significant growth of self-service establishments in comparison to traditional stores [12]. Specifically, supermarket chains have become one of the main players in grocery retailing in almost every Spanish city [13] and, in recent years, have earned significant market shares
Literature review
There are two competing scientific methods to analyze efficiency: the stochastic frontier (SF) and the DEA. Several papers exist in the retailing literature using either of these methods or a combination of two. In Table 1, we summarize these papers presenting the models and the inputs and outputs used.
We can observe that most authors have used the DEA method and that there are only few papers using the econometric models. Sellers and Mas [19] estimated a production function of Spanish
Stochastic frontier: the Bayesian framework
The stochastic frontier model used in this study can be simply expressed aswhere is the natural logarithm of total cost for retail store i at time t, a vector of exogenous variables,a known functional form, a vector of unknown parameters which define the deterministic part of the frontier technology, and and are two random terms, one is symmetric around zero and represents measurement noise, and the other is non-negative and capture
Data
The sample is taken from Spanish self-service retail establishments in the general grocery retail sector, with retail selling areas of between 400 and 2500 square meters (supermarkets). The estimation of efficiency requires homogeneous units, so in order to guarantee the consistency of the companies analyzed, hypermarkets were excluded, because the assortment and services provided to consumers are quite different from supermarkets.
During the period considered there were a series of mergers and
Results
We estimate the Bayesian frontier model2 in this study using data on 74 retail stores over T=7 years. The functional form of the model is a restricted translog and can be expressed as follows3
Discussion and conclusions
Growing competitiveness among retail companies and the globalization of markets have given rise to an economic environment in which it is becoming increasingly difficult for companies to survive. In today’s extremely competitive environment, consumers are more demanding than ever as they constantly seek immediate access to products at lower prices. In this context, the analysis of efficiency has become an important issue in the retail sector [62]. Efficiency favors intermediary management,
References (71)
Has productivity of retail food stores really declined?
Journal of Retailing
(2003)- et al.
Retail productivity and scale economies at firm level: a DEA approach
Omega
(2003) - et al.
An empirical study of the performance impact of computerization in the retail industry
Omega
(1998) Statistical precision of DEA and Malmquist indices: a bootstrap application to Norwegian grain producers
Omega
(2009)- et al.
DEA game cross-efficiency approach to Olympic rankings
Omega
(2009) - et al.
Benchmarking firm performance from a multiple-stakeholder perspective with an application to Chinese banking
Omega
(2010) - et al.
Bayesian approach to statistical inference in stochastic DEA
Omega
(2010) - et al.
Measuring performance of two-stage network structures by DEA: a review and future perspective
Omega
(2010) - et al.
Retail productivity assessment using data envelopment analysis
Journal of Retailing
(1998) - et al.
Distribution services and economic power in a channel
Journal of Retailing
(1998)
Bayesian efficiency analysis through individual effects: hospital cost frontier
Journal of Econometrics
The measurement of labour productivity in wholesaling
International Journal of Research in Marketing
Process for evaluating retail store efficiency: a restricted DEA approach
International Journal of Research in Marketing
Measuring retail productivity. What really matters?
Journal of Business Research
An empirical analysis of productivity in a Portuguese retail chain using the Malmquist index
Journal of Retailing and Consumer Services
Plural forms versus franchise and company owned systems: a DEA approach of hotel chain performance
Omega
Estimation and inference in two-stage, semi-parametric models of production processes
Journal of Econometrics
Efficiency in hypermarket retailing: a stochastic frontier model
International Review of Retail, Distribution and Consumer Services
Technical efficiency in the retail food industry: the influence of inventory investment, wage levels, and age of the firm
European Journal of Marketing
Transformaciones estructurales, precios y márgenes en el sector de distribución al por menor de alimentos [Structural transformations, price and margins in grocery retailing]
Economic Bulletin Bank of Spain
Evolución de la distribución comercial y de los hábitos de compra: del dualismo al polimorfismo [Distribution evolution and purchase habits]
Información Comercial Española
Concentración y competencia en los canales de distribución alimenticios [Concentration and competition in grocery retailing]
Papeles de Economía Española
Concentración en el sector de distribución comercial en España [Concentration in Spanish grocery retailing]
Distribución y Consumo
Intensificación de la competencia intertipo entre supermercados e hipermercados [Increasing intertype competition among supermarkets and hypermarkets]
Distribución y Consumo
An empirical analysis of productivity growth in grocery retailing: evidence from Spain
International Journal of Services Industry Management
Different approaches to the evaluation of performance in retailing
International Review of Retail, Distribution and Consumer Research
An estimable model of supermarket behaviour: prices, distribution services and some effects of competition
Empirica
Return to scale in retail trade
Review of Income and Wealth
Productivity in the distribution trades: the shopper and the economies of massed reserves
The distribution sector and development process: are there patterns? Yes
Economic Inquiry
Bayesian inference on technology and cost efficiency of bank branches
Financial Markets and Institutions
Posterior analysis of stochastic frontier models using Gibbs sampling
Computational Statistics
An exploratory analysis of the correlates of labor productivity in retailing
Journal of Retailing
60 ( 3)
Distribution in Great Britain and North America
Productivity growth in the retail trade
Oxford Bulletin of Economics and Statistics
Cited by (35)
Cross-benchmarking for performance evaluation: Looking across best practices of different peer groups using DEA
2020, Omega (United Kingdom)Citation Excerpt :This raises the need to extend cross-benchmarking for setting more realistically achievable targets, for example by using context-dependent DEA as in Ramón et al. [31]. Moreover, DEA suffers from statistical limitations, as pointed out in Assaf et al. [4], who use an innovative Bayesian frontier methodology to analyze the performance determinants of retail stores. Cross-benchmarking has been developed here for use when information on preferences is not available.
Spillover effects of investment in big data analytics in B2B relationships: What is the role of human capital?
2020, Industrial Marketing ManagementCitation Excerpt :De Jorge-Moreno (2008) analyses the impact on Spanish hypermarkets' technical efficiency of the devolution of retail planning and regulation to regional governments. Assaf, Barros, and Sellers-Rubio (2011) analyzes the impact of vertical integration, age, geographic expansion, and low price strategy on the cost efficiency of retail stores. The data involve a sample of Spanish retail stores that operate in a highly competitive and dynamic environment.
The Effect of Cross-Channel Integration on Cost Efficiency
2019, Journal of Interactive MarketingCitation Excerpt :Our focal outcome is cost efficiency in the retail context. Efficiency is defined as the ratio of outputs to inputs (Assaf, Barros, and Sellers-Rubio 2011; Barros 2006; Ingene 1982; Mishra and Ansari 2013). Moreover, cost efficiency determines how well a firm allocates and uses its resources to produce given outputs (Krasnikov, Jayachandran, and Kumar 2009).
Data-driven analytics for benchmarking and optimizing the performance of automotive dealerships
2019, International Journal of Production EconomicsCitation Excerpt :DEA and its extensions are certainly applicable for activities such as dealer/outlet termination (Vyt and Cliquet, 2017). Despite its compelling use cases and several attempts to overcome some of the limitations of DEA (Liu et al., 2018), it remains that DEA suffers from its inability to fit models that can be used for statistical inferences (Assaf et al., 2011). For these reasons, we instead take the statistical regression approach for deriving improved and actionable recommendations for retail performance management.
Psychological Distress is increasing among customer-facing retail employees: Evidence from 1997 to 2015
2018, Journal of Business ResearchCitation Excerpt :The preponderance of evidence on the changes in retailing confirms that rapidly increasing competition exists between brick-and-mortar and digital businesses with the basis of competition often being lowest price to the consumer (Bachrach, Ogilvie, Rapp, & Calamusa, 2016; Colwell & Ramsland, 2003). One result is narrowing profit margins that force many traditional retailers to vacate mall locations (Assaf, Barros, & Sellers-Rubio, 2011; Wakefield & Baker, 1998), open value-priced outlet stores, reduce store square footage, and decrease staffing levels (D'Innocenzio, 2017). Consequently, a wide range of sources and perspectives provides that CFREs are required to accept broader responsibilities because of decreased numbers of coworkers and the changing demands of the dynamic business environment.
Multi-period efficiency measurement in data envelopment analysis: The case of Taiwanese commercial banks
2014, Omega (United Kingdom)Citation Excerpt :When this approach was initially proposed, it was used for not-for-profit organizations where the input and output factors do not have market values. Later, it was found that it is also applicable to profit-driven companies such as banks [4–7], manufacturing companies [8], hospitals [9], and retail stores [10]. DEA is now a standard technique for performance measurement.