The Effects of Information Technology in Retailer Performance and Survival: The Case of Store-Based Retailers

This research aims to empirically examine the effects of Information Technology (IT) in the performance and survival of store-based retailers. In particular, we investigate how two types of IT resources—Overall IT investment and Human IT resources—respectively influence retailer performance and survival. We draw on the resource-based view, dynamic capabilities, and IT business value research to develop the hypotheses. Our study employs a longitudinal panel dataset at the firm level obtained from secondary sources from 2010 to 2021. Our sample includes 95 store-based public retailers in the U.S. and has 654 firm-year observations in total. We employ the dynamic panel model and binary logistic regression method to investigate the effects of IT resources on retailer performance and survival, respectively. Our empirical results indicate that while overall IT investment has a positive effect on a retailer’s survival, human IT resources positively influence performance. In addition, stronger performance leads to greater survival likelihood. The findings provide important theoretical, managerial and policy implications concerning the business value of IT in the retailing context. Plain Language Summary This research aims to empirically examine the effects of Information Technology (IT) in the performance and survival of store-based retailers. In particular, we investigate how two types of IT resources—Overall IT investment and Human IT resources—respectively influence retailer performance and survival. We draw on the resource-based view, dynamic capabilities, and IT business value research to develop the hypotheses. Our study employs a longitudinal panel dataset at the firm level obtained from secondary sources from 2010 to 2021. Our sample includes 95 store-based public retailers in the U.S. and has 654 firm-year observations in total. We employ the dynamic panel model and binary logistic regression method to investigate the effects of IT resources on retailer performance and survival, respectively. Our empirical results indicate that while overall IT investment has a positive effect on a retailer’s survival, human IT resources positively influence performance. In addition, stronger performance leads to greater survival likelihood. The findings provide important theoretical, managerial, and policy implications concerning the business value of IT in the retailing context.


Introduction
In today's business environment, firms face an increased intensity (and turbulence) of competition in the face of ever-changing market conditions and consumer expectations.The retailing industry is a prime example of this challenge: retailers have to deliver a high customer experience (e.g., consumers' ability to engage retailers and purchase through multiple channels) while trying to control their costs and maximize their resources (Grewal et al., 2021).This need has become more evident during the Covid-19 pandemic, which has accelerated the speed that retailers have to adapt.Many of the changes are driven by or related to technologies such as artificial intelligence (AI), machine learning (ML), big data, virtual reality (VR), augmented reality (AR), Internet of Things (IOT), telecommunications, etc. (Grewal et al., 2021;Shankar et al., 2021).Consumers use a wide variety of technologies to shop online such as wearables, smart speakers and payment technologies (digital wallet).For traditional brick and mortar retailers, this has created the necessity to provide seamless, synchronized omnichannel retailing strategies that optimize all customer touchpoints (Jindal et al., 2021).In order to compete effectively with digital native ecommerce firms, they need to transform their business model and constantly pivot to technologybased solutions.The fast changing retail market conditions require substantial investments in IT resources and capabilities (Cao & Li, 2015;Larke et al., 2018;Ye et al., 2018) and human know-how (Lapoule & Colla, 2016).Researchers have established the importance of aligning IT strategy with business strategy, especially in a competitive marketplace (Ariyachandra & Frolick, 2008;Henderson & Venkatraman, 1999;Luftman, 2003).It is imperative for retailers to understand how to make strategic decisions about IT resources to maximize the returns and attain competitive advantages.
A growing number of retailers have been outsourcing their IT duties and operations for third-party guidance and services (Lou et al., 2020;Perdikaki et al., 2015).This decision is motivated by several factors.First, faced with a broad spectrum of fast-evolving technologies, firms may not have the necessary capabilities to meet the emerging IT needs.Second, IT outsourcing can reduce costs and increase efficiency.This enables retailers to focus on essential customer-centric activities.To successfully transform to omnichannel retailing, allocating a limited amount of investments to in-house human IT resources could be an efficient approach for store-based retailers to improve profitability and increase their longterm survival rate.However, to effectively deploy IT resources and decide when to outsource IT functions, retailers need to have certain levels of IT skills and expertise.Hence, it is critical to understand the role of human IT resources in performance and survival.
The goal of this study is to empirically investigate the effects of IT resources in retailer performance and their survival in the case of store-based retailers.Drawing on the resource-based view (RBV) (Barney, 1991), dynamic capabilities (DC) (Teece et al., 1997), and IT business value (Clemons & Row, 1991; L. M. Hitt & Brynjolfsson, 1996;Melville et al., 2004), we propose distinct effects of two IT resources: overall IT investment and human IT resources.We assemble a longitudinal firm-level dataset of public store-based retailers in the U.S. over 12 years (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020)(2021) from secondary sources.The final sample consists of 95 and has 654 firm-year observations.For the empirical analysis, we employ the dynamic panel model and binary logistic regression method to examine the effects of IT resources in retailer performance and survival, respectively.
The impact of IT on firm performance has been studied extensively in IT or Operation Management fields over the last few decades (Bui et al., 2019;Cardona et al., 2013;Tallon et al., 2019).In the retailing area, studies have focused on how technologies influence retailer performance; such as, product and service delivery (Pantano & Vannucci, 2019); customer service (Setia et al., 2013); converting offline customers to loyal online customers (Mylonakis, 2005); decreasing consumer friction or pain points (Gauri et al., 2021); consumers' willingness to shop online (K€ uhn & Petzer, 2018;Pavur et al., 2016); mobile retailing adoption (Tyrva¨inen & Karjaluoto, 2019).While most studies have found a positive and significant relationship between IT and firm performance, empirical evidence is still somewhat inconsistent (cf., Kohli & Devaraj, 2003).Research indicates that the relationship between IT and performance depends on factors such as industry type and mediating mechanisms (Chae et al., 2018;Cho & Taskin, 2022), measurement and analysis methodologies (Brynjolfsson, 1993), and time lags in measuring payoff (Devaraj & Kohli, 2000).In addition, most studies only look at short-term performance metrics such as sales, return on investments, profitability, competitive position, competitive actions, and other objective financial measures (e.g., Chakravarty et al., 2013).Little research has looked at whether IT can sustain or even improve future firm survivability.Furthermore, there is a dearth of studies on the role of Human IT resource (Melville et al., 2004).One notable exception is the study of Chakravarty et al. (2013).They found that human IT resources can facilitate a retailer's agility to perform.
Our study makes contributions to the literature on the value of IT resources in several ways.First, our research adds empirical evidence on the benefits of IT by looking at the short-term performance metric and the long-term survival rate as outcomes of IT resources.This perspective enables us to obtain a deeper understanding of the distinct effects of IT resources.In addition, we highlight the role of human IT resource in the retailing industry, which provides new empirical evidence in the extant literature.Finally, most research on the IT value has mainly employed cross-sectional analyses using survey or case method (Tallon et al., 2019).The time series dataset assembled from secondary sources enables us to gain comparability on retailer IT investments and performance over time and control for the effects of unobserved firm-specific factors such as organizational resources.The characteristic of panel data is especially important for the survival analysis as it can help mitigate estimation bias.
In the following sections, we will first present different theoretical perspectives that ground our conceptual framework and hypotheses.This is followed by the methodology and the model and estimation results.Last, we discuss the managerial and academic implications from our research and provide conclusions.

Theoretical Foundation
According to RBV, a firm is said to achieve sustainable competitive advantage (SCA) ''when it is implementing a value creating strategy not simultaneously being implemented by any current or potential competitors'' (Barney, 1991) through the efficient deployment of input resources such as human resources, marketing, operations, IT in the development of inimitable capabilities to achieve certain output objectives (M. A. Hitt et al., 2016).SCA is achieved when resources are valuable, rare, imperfectly imitable, and when the firm has the ability to exploit resources (Nason & Wiklund, 2018).Amit and Schoemaker (1993) define capabilities as ''a firm's capacity to deploy resources, usually in combination, using organizational processes, to effect a desired end.''Capabilities (both tangible and intangible) allow firms to take advantage of other resources within their control to improve their productivity.Both resources (tangible such as physical hardware and intangible such as reputation and staff skills) and capabilities are important antecedents of firm performance (Kozlenkova et al., 2014).Researchers in various disciplines have looked at the impact of individual resources and capabilities on performance-such as human resources/human capital (Gerhart & Feng, 2021;Wright et al., 2001), marketing (Morgan, 2012;Vorhies & Morgan, 2005), operations management (Krasnikov & Jayachandran, 2008), and IT (Altschuller et al., 2010;Cardona et al., 2013).Other researchers have also looked at how internal resources such as marketing and operations management can be combined to achieve SCA (Hausman et al., 2002;T. H. Ho & Tang, 2009); and the interaction effect of corporate social responsibility and IT-enabled innovation on firm performance (Jung et al., 2023).
Researchers have introduced dynamic capabilities (DC) as an extension to RBV (Teece et al., 1997) to explain how and why certain firms can achieve SCA in turbulent times.While RBV focuses on SCA, DC view emphasizes more on how firms ''integrate, build, and reconfigure internal and external competencies to address rapidly changing environments'' (Teece et al., 1997) while maintaining capability standards to ensure competitive survival.The goal of DC is to explain firm-level success and failure.Those that are successful (or survive) are able to develop internal technological, organizational, and managerial processes that result in agilities to create a competitive advantage for the firm.Both RBV and DC emphasize on efficiency and effectiveness of resource utilization to achieve competitive advantage (Barney, 1991;Teece et al., 1997).
Over the past few decades, researchers have looked at the role of IT in enabling or hindering an organization's ability to sense and respond to changes in the external environment and its ability to assemble the necessary resources to take advantage of market opportunities (organizational agility) to affect performance (see Tallon et al., 2019for a review, Karimi-Alaghehband & Rivard, 2019).Researchers have looked at agility at different levels-corporate, business unit, process or work group.For example, (Sambamurthy et al., 2003) look at agility at the process-level where a firm's capability is related to interactions with its customers, orchestrating internal operations, and leveraging its partnerships with business partners.Other researchers have looked at performance outcomes attributed to agility; for example, Lee et al. (2015) show that IT ambidexterity enhances organizational agility by facilitating operational ambidexterity.Generally, the literature suggests that agility can be an end in itself (first order impact at the process-level) or a means to an end (a second order impact of higher firm performance at the firm level) (Tallon et al., 2019).
Building on the RBV, IT business value research looks at the impact of IT on firm performance (Clemons & Row, 1991;L. M. Hitt & Brynjolfsson, 1996;Melville et al., 2004).It posits that IT-related resources such as technological and human IT resources are combined into unique resource capabilities that enhance firm performance and lead to SCA (Bhatt & Grover, 2005;Melville et al., 2004;Ross et al., 1996).These IT resources, with the complement of organizational resources, create economic value for a firm by ''conferring operational efficiencies that vary in magnitude and type depending upon the organizational and technological context'' (Melville et al., 2004).Organizational resources include policies and rules, structure, practices, and culture that may improve business processes, which may ultimately drive business performance.Business process (such as inbound logistics, sales, distribution, customer service) is ''the specific ordering of work activities across time and space, with a beginning, an end, and clearly identified inputs and outputs'' (Davenport, 1993).It provides a context for the application of IT to improve processes that lead to organizational performance.Performance is categorized into (1) business process performance, and (2) organizational or overall performance.While the former focuses on measures related to operational efficiency, the latter focuses on aggregate IT impacts across firm activities.Examples of business process performance metrics include on-time shipping, customer service, and inventory management.Examples of overall performance include sales growth, profitability, market value, productivity, competitive advantage.The IT business value literature has identified three general types of IT resources that impact performance: (1) technology resources, (2) human IT resources, and (3) organizational IT resources (Bharadwaj, 2000;Mata et al., 1995).These three complementary resources, when combined appropriately, result in superior firm performance (Melville et al., 2004;Pavlou & El Sawy, 2006).The IT business value literature also recognizes the importance of both external competitive environment (such as industry characteristics and trading partner resources) and macro environment (such as country characteristics) (Melville et al., 2004).
In assessing organizational performance, researchers often look at efficiency and effectiveness.To achieve sustained SCA, organizations need to learn how to do the right things (effectiveness) while doing things right (efficiency) (F.N. Ho & Huang, 2020).To achieve efficiency, the focus is on employing metrics such as cost reduction and productivity improvements in assessing a given business process.Effectiveness refers to the achievement of objectives relative to external environment; for example, the attainment of competitiveness advantage.The retail industry is a prime example of this challenge: retailers have to deliver a seamless shopping experience while trying to maximize their IT investments to achieve SCA.IT enables a retailer to improve intermediate process efficiency at organizational-wide level to achieve efficiency and competitive impacts.In a review of prior IT business value literature on the impact of IT on performance, Melville et al. (2004) reveal that (1) IT acts an intermediate business process, (2) IT has a complementary role, whether as a mediator or moderator, (3) the external environment plays a role in IT value generation, and (4) the importance of disaggregating the IT construct into meaningful subcomponents.
Researchers have looked at various classifications of IT resources.For example, some researchers identify three types of IT: IT spending, IT strategy, and IT management/capability (Dehning & Richardson, 2002).Ross et al. (1996) classify IT into three asset classes: human, technology, and relationship.Melville et al. (2004) operationalize IT resources as: physical, human, and organizational.All three components are complementary in achieving organizational agility/performance by acting as either as an enabler or facilitator of IT competencies (Chakravarty et al., 2013).Melville et al. (2004) propose that IT resources-both technology and human-create economic value for a firm by conferring operational efficiencies that lead to enhanced performance (such as sales growth and profitability).

Conceptual Framework and Hypotheses
Consistent with RBV, DC and the IT business value framework, we propose a conceptual model in Figure 1 that depicts the relationship between IT resources, firm performance and survival.Following the definitions used by Melville et al. (2004), we disaggregate the IT construct into two separate components and propose distinct effects in our conceptual framework.The first component, overall IT investment, encompasses physical capital resources and business applications.The second IT resource, human IT resources, refers to the number of IT employees within the enterprise.Henceforth, ''Overall IT Investment'' refers to total IT expenditures in software, hardware, telecommunications, and services relative to firm size; ''Human IT resources'' is used interchangeably with the number of IT employees relative to firm size.
We posit that higher levels of overall IT investment should lead to stronger IT capabilities and resources.Research supports the positive effect of IT-enabled capabilities on firm performance at the firm level (Grant & Yeo, 2018;Melville et al., 2004;Peppard et al., 2007) or in multi-country industry settings (Cardona et al., 2013); and, that in aggregate, overall IT resources create economic value through operational efficiencies (Kohli & Devaraj, 2003).For store-based retailers, they need to invest in IT technologies in order to address the growing ecommerce demand.For example, research has found that technologies can convert offline customers to loyal online customers (Mylonakis, 2005); improve customers' overall experience, and increase their level of engagement with the firm (Shah & Murtaza, 2005); develop a more intimate understanding of their customers resulting in more customized marketing (Cudmore & Patton, 2007); decrease consumer friction or pain points (Gauri et al., 2021).The customer value from IT directly impacts firm performance (Gellweiler & Krishnamurthi, 2022).Therefore, we propose that: H 1 : Overall IT investment has a direct positive impact on a retailer's sales growth.
Researchers have argued that human IT resources are valuable resources that consistently confer a SCA (Bharadwaj, 2000;Mata et al., 1995).Managerial expertise, efforts, and knowledge (i.e., human capital) are needed to facilitate the implementation of important business processes.They include technical expertise to integrate multiple systems, software application development, system maintenance, cybersecurity; and, managerial expertise to allocate resources efficiently and effectively, lead and motivate IT teams, identify appropriate opportunities, etc.Researchers such as Collis (1994) argue that firms with superior managerial capabilities can achieve organizational objectives through the manufacturing, offering, and delivery of products and services that meet customer needs.There is empirical support linking human capital and firm performance (Acquaah, 2012;Alnoor, 2020;Ilmudeen et al., 2023).Researchers have found that IT knowledge and training have a positive impact on performance (Liebach L€ uneborg & Flohr Nielsen, 2003;Powell & Dent-micallef, 1997;Tippins & Sohi, 2003).Song et al.'s (2021) study reveals that human IT capital is positively related to the level of a retailer's digitalization, and digitalization is positively related to the retailer's supply chain integration.Therefore, we propose that: H 2 : Human IT resources have a direct positive impact on a retailer's sales growth.
The effective use (''doing the right things'') of IT resources denotes whether an organization achieves its objectives relative to the external environment; that is, effecting a unique value-creating strategy relative to competitors (Barney, 1991).While the RBV provides an understanding on the link between IT resources and performance; however, this link may only confer a temporary advantage that may weaken over time either as the IT resources become commoditized (diminished rareness) or become imitable.This is true in the IT space as the rapid speed and adoption of innovation makes it difficult to maintain a sustained inimitable advantage.
In the case of store-based retailers, we argue that IT capabilities (in the form of efficient utilization) enable their agility to sense opportunities for competitive actions and marshal the necessary resources to seize those marketing opportunities, which impacts retail survival (Chakravarty et al., 2013).During the Covid-19 pandemic, retailers have to make swift changes by adopting digital technologies at an increased pace (Guthrie et al., 2021;Klein & Todesco, 2021).This allows retailers to achieve operational efficiencies such as marketing, logistics and supply chain, and organizational management (Ye et al., 2018).Robertson et al. (2022) show that small and medium-sized retail enterprises (SMEs) that are digitally mature exhibit higher levels of organizational resilience, that is, a set of adaptive capabilities that enable an organization to recognize, learn and cope with unexpected events (Williams et al., 2017;Yao & Fabbe-Costes, 2018).Other researchers have found that technology can enhance a retailer market value through improved product offerings and services (Herhausen et al., 2015); the ability to offer a fast and user-friendly consumer shopping experience (Hagberg et al., 2016;Verhoef et al., 2015); supply chain integration to monitor inventory, order delivery and fulfillment (Bernon et al., 2016;de Leeuw et al., 2016;Ishfaq et al., 2016).Therefore, we propose that: H 3 : Overall IT investment has a direct positive impact on a retailer's survival.
All the IT-enabled technologies in retailing require IT human skills and capabilities (Y.Chen et al., 2018;Lapoule & Colla, 2016).Research has shown some of the operational barriers for omnichannel retailers include the failure to deploy the required human resources to train current staff and provide needed skills; and lack of management engagement (Brynjolfsson et al., 2013;Chopra, 2016;Picot-Coupey et al., 2016;Ye et al., 2018;Zhang et al., 2010).For example, Powell and Dentmicallef (1997) show that in addition to IT capital investment, some retailers ''have gained advantages by using ITs to leverage intangible, complementary human and business resources such as flexible culture, strategic planning-IT integration, and supplier relationships.''Researchers have looked extensively at how new technologies such as RFID, GPS, AI impact retailers' strategy, operations, and survival (see Varadarajan et al., 2010 for a review).For example, S. Chen (2003) finds that new e-technologies (such as cell phones) can generate opportunities through the creation of new e-business models, which are key factors in determining the survival and failure of retailers.Strong in-house Human IT resources should allow store-based retailers to develop capabilities to successfully transform their business models and incorporate the fast-evolving technologies into their business processes and strategies.Therefore, we propose that: H 4 : Human IT resources have a direct positive impact on a retailer's survival.
We posit that retail performance serves as a strong precursor of long-term survival.In an empirical study conducted in the context of multichannel retailing, Oh et al. (2012) find that higher levels of exploitative and explorative competences can lead to stronger performance.In the transformation process to omnichannel retailing, we posit that store-based retailers must possess strong capabilities to efficiently and effectively allocate and deploy different IT resources in order to serve existing customers and deliver innovative services to attract new customers.For example, Ye et al. (2018) identify one of the important strategic drivers in the transition to omnichannel retailing pertain to supply chain and logistics management.Store-based retailers have to invest in centralized IT platforms and systems to achieve supply chain efficiency, traceability and scalability and to build integrated channel fulfillment processes.We argue that strong ambidextrous capabilities should enhance organization agility.To survive in a turbulent environment characterized by increasing ecommerce and to compete effectively with digital native firms such as Amazon, store-based retailers need to possess great agility to quickly change in the business processes of marketing, supply chain and organization management (Ye et al., 2018).Therefore, we expect that store-based retailers with stronger performance will have a greater (lower) survival (failure) likelihood in the long term.
H 5 : Sales growth has a direct positive (negative) relationship with the likelihood of a retailer's survival (failure).

Data
We choose to focus on public store-based retailers in the United States because of our interest in the phenomenon of traditional firms' survival in the digital economy and the availability of financial performance data.We assemble a longitudinal firm-level dataset from multiple secondary sources including Standard & Poor's Compustat database, IT budget and spend dataset from Aberdeen Group (https://www.aberdeen.com),Annual Retail Trade Survey by the U.S. Census Bureau, and Digital Commerce 360's U.S. Top 500 Database (https://www.digitalcommerce360.com).Following the sample selection approach used by similar studies (Cao & Li, 2015;Feng & Fay, 2020;Shi et al., 2018), we choose the U.S. Compustat database as the sample frame and select public retailers whose financial records are available from 2010 to 2021.The financial data are then merged with variables obtained from the other sources.After eliminating cases with missing data across all variables, the final sample contains 95 retailers and 654 years-firm observations covering six retail sectors according to the Global Industry Classification Standard: Textiles, Apparel & Luxury Goods (252030), Internet & Direct Marketing Retail (255020), Multiline Retail (255030), Specialty Retail (255040), Food & Staples Retailing (301010) and Health Care Providers & Services (351020).The list of the sample companies used for the empirical analysis is provided in Appendix A.
Using the U.S. Compustat database, we gather sample retailers' yearly financial information, such as the timing of the event (being delisted from Compustat), sales, assets, and Herfindahl-Hirschman Index (HHI).When a retailer is delisted from a stock exchange, it would be excluded as an independent, public entity in the Compustat database.This could occur when a retailer ceases operations, declares bankruptcy, is acquired or merges with others, does not meet minimum listing requirements, or seeks to get less expensive sources of capital by going private.If a retailer gets delisted, we record its last year of available observations in the Compustat database and mark the subsequent year as the time of the ''failure'' event.The ''failure'' variable is coded as one if a failure event is observed and zero otherwise.Retail performance is approximated by the logarithm of the annual sales growth percentage (Cao & Li, 2015).To capture the market competitiveness in a retailer's industry, we use the HHI, which is calculated by the sum of the squared market share of each retailer competing in the same industry.The HHI measure can range from close to 0 to 10,000, with lower values indicating a less concentrated market.In addition, industry dummies of sample retailers are included as control variables to account for specific industry effects.
IT spending data are obtained from a private data source, Aberdeen Group (acquired by Spiceworks in 2020).Aberdeen Group is a leading data provider for enterprise IT budgets.It collects information on over 110,000 enterprises in North America, and all of the individual locations that comprise those enterprises.Previous research has used their data to study the effects of IT on organizations (e.g., Bloom et al., 2014).For the purpose of the present study, we focus on the following five items Aberdeen covers: the number of IT employees, IT Hardware Budget, IT Software Budget, IT Services Budget, and Telecommunication Budget.The scale definitions are provided in Appendix B. The number of IT employees is measured on an ordinal scale from one to seven, where seven being the highest.The remaining four entities are measured on an ordinal scale from one to six, where six represents the highest category.In particular, the ordinal measure on the number of IT employees should reflect a firm's human IT resources and the other four items are expected to capture the extent of IT investments in various areas.The descriptive statistics on the five original items for the sample firms are provided in Table 1.
Prior to the empirical model estimation, we need to address the high correlations observed in these five IT resource items as this will cause multicollinearity issues.To solve this problem, we undertook a Principal Components Analysis (PCA) on the five variables and extracted two main components that capture the majority of variances (92%) in all the variables.The first principal component, ''Overall IT Spend'' relates strongly to all items except for the number of IT employees.It reflects a firm's overall IT investment in key areas.The second component, termed ''Human Resource,'' exhibits high correlation only with the number of IT employees.Subsequently, we chose to include the two main components in our empirical model.As these two components are also positively correlated with firm size, we transform the two components by dividing them by the logarithm of firm assets to mitigate the scale effect.
To account for industry dynamism, we choose a proxy variable which is the annual ecommerce sales growth rate during the sample period gathered from the Annual Retail Trade Survey by the U.S. Census Bureau (this variable could capture the impact of the Covid-19 pandemic).Furthermore, we augment the data by adding several variables gathered from a proprietary data source, Digital Commerce 360's U.S. Top 500 database.The first variable pertains to the classification of retailers.This variable allows us to verify and select retailers that are store-based into our empirical sample.In addition, we obtain the measure of a retailer's online experience and the firm age variable from this database.For each firm-year observation, the former variable is calculated as the difference between the year that a retailer started its ecommerce operation and the focal year and the latter is calculated as the difference between its founding year and the focal year.We apply the logarithm to both variables in our model.
The descriptive statistics on the key variables used in the model are provided in Table 2.The results indicate that multicollinearity is not an issue with the use of the two factors extracted from PCA.

Model
To examine the effects of retailers' IT spend and human IT resources on sales growth, we employ the dynamic panel data modeling approach.This method allows us to better account for endogeneity and capture the dynamics in the data.We regress the sales growth on its own lagged value, OverallIT and HumanResource, and other control variables including retailer age, retailer ecommerce experience, industry competition, and industry dynamism.The model specification is presented in Equation 1.
We lag both IT variables by one period to alleviate the reverse causality between the explanatory and dependent variables.Two endogeneity problems remain.First, unobserved firm-specific variables may influence firm growth and IT variables, thus confounding their relationship.Second, there are potential random shocks that affect both performance and other contemporaneous covariates.To alleviate the two problems, we adopt the estimation method suggested by Arellano and Bond (1991), which is commonly used in the literature (e.g., Ho, Wang, Ho-Dac & Vitell, 2019).First, we use the first differencing method to remove the unobserved firmspecific effects.Next, we employ the two-period lagged term of the growth rate to instrument for its first differenced term.The generalized two-step method of moments (GMM) with bias-corrected standard errors is used to estimate the model.Furthermore, we perform the Sargan test and the autocorrelation test to ensure the instrument variable is appropriate for the data at hand and there is no serial correlation.Both tests show nonsignificant results.
Next, we use a binary logistic model for the survival analysis because the failure/survival dependent variable is binary.The failure likelihood is modeled as a function of the retailer performance and the two IT variables.A random intercept term is used to account for unobserved firm heterogeneity.In addition, we include firm age, firm ecommerce experience, industry competition, industry dynamism, and industry dummies as control variables.The survival model is presented below.

Results
The estimation results are provided in

Discussions
The objective of this research is to empirically investigate the effects of IT resources on retailer performance and their survival for store-based retailers.Our research reveals three main findings.First, overall IT investment has a direct, positive effect on retailers' long-term survival (H 3 ).This implies that IT expenditures (relative to store-based peers) appear to provide store-based retailers increased opportunities to remain competitive in the fast-changing retail environment.Put it differently, IT investments (e.g., hardware, software, network, and communications) may help store-based retailers develop organizational agility and adaptive capabilities to transform their business model and deliver a better  omnichannel experience for their customers, thereby increasing their long-term survival likelihood.
The second finding pertains to the direct, positive effect of human IT resources on retailer performance (H 2 ).This implies that it is beneficial and important for retailers to have a sufficient number of internal IT employees relative to the firm size in order to maintain robust short-term sales growth.Our results are consistent with previous research that found that human IT resources can facilitate a retailer's capabilities to perform (Chakravarty et al., 2013).
Third, our study found that retailer performance (specifically sales growth) has a positive effect on survival (H 5 ).The results indicate that increased IT resources reduce the failure likelihood, thus increasing the chance of survival.This finding, combined with the insignificant result for H 1 , reveals that while IT resources may not appear to pay off in the short run in terms of firm performance for retailers; however, they help a retailer's survival in the long term.Sheth (2007) found that most retailers fail when they are either unable or unwilling to adapt to changing ecosystem such as technology, competition, and consumer lifestyles.Our study supports this notion that in order for retailers to maintain a competitive advantage (and survival), they need to make cautious resource allocation decisions between different IT resources.
Our research indicates that the relationship between IT and performance is neutral.This finding is consistent with equivocal results from extant literature (Kohli & Devaraj, 2003).The relationship could be influenced by contingency factors such as industry type and mediating mechanisms (Chae et al., 2018;Cho & Taskin, 2022), measurement and analysis methodologies (Brynjolfsson, 1993), and time lags in measuring payoff (Devaraj & Kohli, 2000).Furthermore, we speculate that IT resources cannot endow store-based retailers with significant growth in sales.While RBV informs the linkage between business process (such as IT investment) and firm performance by conferring a temporary competitive advantage; however, imitation and diminished rareness over time weaken such advantages as all retailers invest heavily in IT during turbulent times (Hopper, 1990;Kohli, 2003).Hence, IT investment is used mostly as a defense in the case of store-based retailers.Moreover, the intense competition from pure-play ecommerce firms makes it difficult to translate the IT investment directly into stronger sales performance as store-based retailers need to simultaneously invest in their physical stores and may not have sufficient funds to overinvest in IT resources alone.
The present study finds no support for the direct relationship between Human IT resources and survival as hypothesized in H 4 .According to the DC theory, firms are more likely to survive in rapidly changing environments if they are able to quickly integrate and reconfigure internal and external competencies to address new challenges (Teece et al., 1997).We speculate that storebased retailers can rely on external human resource to enhance their IT expertise or knowledge needed for adaptation.For example, store-based retailers can use outside IT vendors to support their business needs in ecommerce technologies.This strategy can allow them to deploy their IT employees more efficiently by prioritizing important needs such as customer services, fulfillment performance and channel integration.Nonetheless, Human IT resources still have an indirect impact on survival through the pathway of firm performance.

Conclusions
Retailers, especially those that are store-based, are going through an enormous transition during a turbulent time that has been hastened by the Covid-19 pandemic.Some researchers have even labeled brick-and-mortar retailing as an ''endangered species'' (Sheth, 2021).Even though brick-and-mortar retailing will be here to stay; however, traditional retailers may not be able to survive.The key question is: if retailers are to survive in an era of rapid technological changes, how should they invest in IT resources to transform themselves to become successful omnichannel retailers?In our study, we have found evidence that supports the distinct roles of IT resources in retail performance and survival of traditional store-based retailers.
This study contributes to the literature on the value of IT resources.Using a longitudinal dataset that spans more than 10 years, we provide empirical evidence on the distinct short-term and long-term effects of disaggregate IT resources in the case of store-based retailers.In particular, we extend the performance metric to the long-term survival likelihood, which is not widely investigated in the literature.As previous research utilizes mostly crosssectional survey data or case studies, our study expands the scope of the extant literature concerning the value of IT resources.This study also provides unique insights into how different IT resources affect performance and survival respectively in a rapidly changing business environment; by doing so, it broadens the understanding of the strategic implications of IT resources.The second implication pertains to the role of human IT resources.Extant literature has identified the effects of human resource in service (e.g., Oh et al., 2012) and in sales force (e.g., Lapoule & Colla, 2016) on firm performance.Other disciplines such as industrial and organizational psychology have extensively looked at the impact of human capital on performance (see Crook et al., 2011 for a review).However, research on the impact of human IT resource has been scarce (Melville et al., 2004).Our research highlights the important role of human IT resources in the retailing industry.In particular, we show that internal human resources in the IT domain could result in performance advantages.
For practitioners, our research offers important insights.Our study found that overall IT investment does not have a significant effect on sales growth (H 1 ).While IT human resources have been found to affect sales growth directly, there is no direct effect on retailer survival (H 4 ).The differential impacts of the two resources identify important issues for store-based retailers.They need to make choices about how much to invest in technologies and human resources to compete and grow sales, and being mindful of the sustainability of their IT strategies; that is, how to realize the returns of their IT investment and being able to survive in the foreseeable future.When attempting to maximize the utilization of their investments and assets, retailers may be faced with the challenges of strategically allocating funds between technologies and human resources.To this end, our findings highlight the importance of the decision a storebased retailer has to make with respect to IT human resources in the process of digital transformation.Having a strong, lean team of in-house IT employees may help increase retail performance in terms of growth.However, given that it is not a critical resource for survival, retailers need to be cognizant of the costs and benefits involved with having internal IT staff, and carefully consider outsourcing non-essential services to external vendors.
Our research also raises an important implication for antitrust policy makers.Our analysis shows that storebased retailers with significant IT resources, after the endogenous effects of firm size being controlled for, are more likely to perform better and survive in the long term.This implies that large retailers such as Walmart have SCA from their IT resources over small store-based retailers.This advantage may be more pronounced if we include pure online retailers such as Amazon.Recent anecdotal evidence suggests that after the pandemic, technology investments from large retailers such as Walmart and Target continue to rise despite the economic slowdown and declining earnings whereas other retailers are cutting back expenses to defend profits (Mullaney, 2022).In other words, large store-based retailers are also imitating the business strategy of Amazonforegoing short-term returns for long-term dominance (Khan, 2018).Our research suggests that smaller retailers are faced with IT related barriers to survival.Other researchers have also noticed a potential threat of anticompetitive effects emerging in the digital commerce (Balto, 2000).Therefore, government agencies and regulators might consider leveling the competition and protect smaller retailers through enforcement of antitrust laws.
Our research is not without limitations.First, we use high-level, broad proxies for IT investment and IT Human resources.For example, we use number of IT employees as a surrogate for IT Human resources.Complementary organizational resources such as organizational structure, policies and rules, workplace practices, and culture are not included in our empirical analysis.In addition, our data lacks measures that indicate retailer adoption of technologies.Future research may wish to use more granular measures and/or different methodologies (e.g., syndicated data or natural or quasi experiments) to investigate how adoption of specific technologies or organization knowledge capabilities involving IT staff may impact business processes and value creating mechanisms.Second, our research focuses on public, store-based retailers.Future research may wish to broaden the sampling frame to include onlineonly retailers and small and medium sized, private retailers to gain a more comprehensive understanding of the phenomenon.
Appendix A: List of Store-Based Retailers

Table 1 .
Descriptive Statistics on IT Resource Items.
Table 3. H 1 tests a direct relationship between overall IT resources and firm performance.The result shows no support for H 1 as the coefficient of Overall IT Spend on performance (shown in Panel A) is nonsignificant.H 2 concerns the effect of Human IT Resource on performance.As the coefficient of Human IT resource (shown in Panel A) is positive and significant, H 2 is supported.H 3 tests the relationship between IT resource and firm survival.As expected, H 3 is supported as the coefficient of Overall IT Spend on failure (shown in Panel B) is negative and significant.H 4 tests the relationship between Human IT resource and firm survival.It is not supported as the coefficient of IT Human Resource on the failure variable (shown in PanelB) is nonsignificant.Finally, H 5 concerns the relationship between firm performance and survival.As expected, the coefficient of retail performance on failure (shown in Panel B) is negative and significant, which means that strong retail performance reduces the failure likelihood, thus increasing the chance of survival.Therefore, H 5 is supported.Most of the control variables do not seem to exhibit strong significant effects on either performance or survival.Figure2summarizes the results of the five hypothesis tests.

Table 2 .
Descriptive Statistics on Key Variables.