Journal of Transportation Management

Volume 26 | Issue 2 Article 3

1-1-2016

An empirical assessment of logistics/supply chainmanagement in two Latin American countriesJohn E. SpillanUniversity of North Carolina at Pembroke, jspillan@uncp.edu

Michael A. McGinnisPenn State New Kensington Campus, mam47@psu.edu

Ali KaraPenn State York Campus, axk19@psu.edu

César Antúnez de MayoloUniversidad del Pacífico, c.antunezdemayolof@up.edu.pe

Gustavo JaraUniversidad de Piura, gustavo.jara@udep.pe

Follow this and additional works at: https://digitalcommons.wayne.edu/jotm

Part of the Operations and Supply Chain Management Commons, and the TransportationCommons

This Article is brought to you for free and open access by DigitalCommons@WayneState. It has been accepted for inclusion in Journal ofTransportation Management by an authorized editor of DigitalCommons@WayneState.

Recommended CitationSpillan, John E., McGinnis, Michael A., Kara, Ali, de Mayolo, César Antúnez, & Jara, Gustavo. (2016). An empirical assessment oflogistics/supply chain management in two Latin American countries. Journal of Transportation Management, 26(2), 7-27. doi:10.22237/jotm/1451606520

Fall/Winter 20167

AN EMPIRICAL ASSESSMENT OF LOGISTICS/SUPPLY CHAIN MANAGEMENT INTWO LATIN AMERICAN COUNTRIES

John E. SpillanUniversity of North Carolina at Pembroke

Michael A. McGinnisThe Pennsylvania State University

New Kensington Campus

Ali KaraThe Pennsylvania State University

York Campus

César Antúnez de Mayolo Universidad del Pacífico

Gustavo JaraUniversidad de Piura

ABSTRACT

The Bowersox Daugherty (1987) logistics strategy typology (Process Strategy, Market Strategy, andInformation Strategy) is an important conceptual framework for studying logistics/supply chainmanagement strategy and its role on logistics/supply chain management outcomes. The purpose ofthis research is to empirically apply the typology in Peru and compare the findings with the previousresearch conducted in Guatemala. The three Bowersox/Daugherty dimensions are used to define theconstruct Overall Logistic Strategy (OLS), and then, the OLS was used to measure OrganizationalCompetitiveness (COMP) through two intervening variables LCE (Logistics CoordinationEffectiveness) and CSC (Customer Service Commitment). The results indicate that generally thelogistics strategy in Peru is fundamentally similar to Guatemala’s. In other words, the direction ofthe relationships among the conceptualized constructs tested in the SEM model was significant andexplained a sizable variation in COMP in both countries. This provided additional support for therobustness of the structural model in different cultural environments. However, some differences areapparent. First, the importance of the three independent variables and three dependent variablesappear to be greater to the Peruvian respondents than Guatemalan respondents. Second, on closerinspection Peruvian logistics data indicates relatively greater emphasis on information, coordination,customer service, and relatively less emphasis on cost efficiency, than Guatemalan managers.Managerial insights and suggestions for future research and discussed.

Journal of Transportation Management8

INTRODUCTIONLogistics management is the process ofmanaging material, service, information andcapital flows from the source, through the firmand to the customer (Logisticsworld, 2015). It isa critical part of an organization’s corporatestrategy (Heskett, 1977). One conceptualframework used in studying logistics/supplychain management is the Bowersox/Daugherty(1987) typology, which has been the basis forlongitudinal research in the United States and aseries of international markets. Collectivelythese studies have demonstrated that theBowersox and Daugherty typology is applicableover time in the United States and in severalother countries with different culturalbackgrounds and economic development levels.As such, these recent empirical studies addressthe concerns of Luo, Van Hoek, and Ross (2001)who stated that cross-cultural logistics/supplychain management research has lagged incomparison to other business disciplines. Theauthors believe that the analysis contained inthese studies validate the Bowersox/Daughertytypology as an effective model for the study oflogistics/supply chain management acrosscultures.

Considering the speed of the globalization, afirm’s ability to manage logistics in cross-country environments has become an importantsuccess factor. Although, globalization offerssignificant opportunities for multi-nationalcorporations (MNCs) to shift theirmanufacturing and distribution around theworld, especially in the developing andemerging markets, global manufacturingstrategies may not be effective if not supportedby successful logistics strategies. Therefore, westrongly believe that cross-cultural/cross-countrylogistics studies have significant potential toenrich our understanding of logistics systemsand strategies applicable in different nationalenvironments. These studies provide in depthlogistics knowledge, which can have importantinternational logistics management implicationsin helping managers to identify similarities, and

would encourage similar strategies, or identifysignificant differences.

Kohn, McGinnis, and Kara’s (2011) recent studyreported the role of overall logistics strategy(OLS) on logistics coordination effectiveness,customer service effectiveness, andorganizational competitive responsiveness.Using multi-year data collected in the U.S., theirfindings demonstrated that the Bowersox/Daugherty dimensions had a significant impacton the company’s competitiveness through thelinks of logistics coordination and customerservice. The purpose of this study is to explorewhether the Bowersox/Daugherty typology isuseful for examining logistics strategies in twodissimilar Spanish language countries in LatinAmerica, namely Peru and Guatemala.

The authors postulate that a two-country/cross-cultural study of Guatemala and Peru wouldfurnish an intriguing example of how logisticssystems are assessed in two nations through thelens of one common measurement instrument.Furthermore, such a study would provide astrong validation of the dimensionality and thestructural relations identified in the recent Kohn,McGinnis, and Kara (2011) study. We emphasizethat the differences in each country’s geographicsize, population size, labor force make-up,infrastructure, and economic systems provide anexcellent platform for evaluating the validity ofthe research instrument, as well as providinginsights into logistics strategies and outcomes inthese heterogeneous countries.This current research adopts a perspective thatthe Bowersox and Daugherty typology providesa strong conceptual framework consistent acrosscountries with regards to salient dimensions oflogistics/supply chain management strategy.These dimensions should be coordinated atmany levels of the organization to achievecompetitive responsiveness regardless of thecountry’s environment. Through this researchthe authors hope to discover the applicability oflogistics/supply chain management strategy andunderstand the role logistics management

Fall/Winter 2016 9

strategy plays in maintaining and enhancingcompetitive advantage responsiveness in cross-country environments. Using a confirmatoryfactor analysis and a structural equation model,we assess the validity of three dimensions of theBowersox and Daugherty typology and theirsimultaneous relationship to logisticscoordination, customer service effectiveness,and overall organizational competitiveresponsiveness.

This paper is organized into seven sections. Thefirst two sections contain the introduction andliterature review and they provide an overviewof the conceptual framework for the study andbriefly compare selected characteristics of Peruand Guatemala. Sections three and four containthe research methodology and data analysis. Thefifth section discusses the similarities anddifferences in logistics/supply chainmanagement between the two Latin Americancountries. The sixth section presents adiscussion of the results and conclusions. Thefinal section provides implications for logistics/supply chain management practitioners,teachers, and researchers.

LITERATURE REVIEW AND ANOVERVIEW

OF PERU AND GUATEMALA

Literature ReviewIn 1987, Bowersox and Daugherty completed acomprehensive study of logistics integration.Their research focused on three distinctlydifferent logistic management strategy types thatfirms have used in their decision-making. Theyare summarized as follows:

The objective of Process Strategy is tomanage flows to gain control overactivities that “give rise to cost”. Incurrent terminology they are referred toas “cost drivers”.

The objective of Market Strategy is toreduce the complexity faced bycustomers. For example, this strategymay try to provide a single point of

contact for customers that sourcemultiple products from differentdivisions, or facilities, of the same firm.

The objective of Information Strategy isto coordinate information flowsthroughout the channel of distribution tofacilitate cooperation and coordinationamong channel (supply chain in today’svocabulary) members.

Three studies (McGinnis and Kohn, 1993, Kohnand McGinnis, 1997b, and McGinnis and Kohn(2002) have tested the three components of theBowersox/Daugherty typology in large U.S.manufacturing firms. The researchers found thatprocess and market strategies were emphasizedwhen logistics strategies were intense. They alsodetermined that both strategies existed atmoderate levels when firms used a balancedstrategy approach. Additionally, they found thatthese strategies were present only at low levelswhen firms used an unfocused strategy. Thesestudies indicate that the three dimensions(logistics process strategy, market strategy andinformation strategy) together, and referred to asOverall Logistics Strategy (OLS), provide abasis for assessing logistics/supply chainmanagement effects on firm competitiveness.One significant contribution of this research wasthat the three dimensions of logistics strategywould be more likely to be blended than usedseparately as Bowersox and Daughtery (1987)originally indicated.

Clinton and Closs’s (1997) research using a sampleof 818 U.S. and Canadian firms to assess thesignificance of the Bowersox/Daughtery typologyconcluded that there was a clear overlap of thethree strategies (process, market, information). Thisis to be expected because logistics performs thesame activities regardless of the overall logisticsstrategy. In addition, Spillan, Kohn, and McGinnis(2011) concluded that the strategies of small andlarge U.S. manufacturing firms vary in degree ratherthan type. Market, Process, and Informationstrategies were present in both small and large firms.Moreover, the authors concluded that the logistics

Journal of Transportation Management10

strategy outcomes of small and large firms weresimilar. It was concluded that the Bowersox/Daugherty typology was applicable to United Statesmanufacturing firms regardless of size.

Recent studies have explored the value andsuitability of the Bowersox/Daugherty typologyin different cultures/countries (McGinnis,Harcar, Kara, and Spillan (2011); McGinnis,Spillan, Kara, and King, D., 2012; and Spillan,

McGinnis, Kara, and Yi (2013). These studieswere conducted in China, Guatemala, Ghana,and Turkey. In each case confirmatory factoranalysis was used to assess the validity ofOverall Logistics Strategy (OLS) usingStructural Equation Modeling (SEM) to test thevalidity of the overall model of OLS-LCE(Logistics Coordination Effectiveness)-CSC(Customer Service Commitment)-COMP(Organizational Competitiveness). In two of

Fall/Winter 201611

these countries, China and Ghana, OLS wassupported, but support for the overall model wasmixed for the Guatemalan data and statisticallyinsignificant for the Turkish data. McGinnis, Spillan,

Kara, and King, (2012) analyzed empirical datacollected in Ghana and found that the OLS-LCE-CSC-COMP model was supported. Finally,Spillan, McGinnis, Kara, and Yi, (2013) compared

Journal of Transportation Management12

Chinese and United States data and found the boththe OLS and the OLS-LCE-CSC-COMP weresupported.

Peru and Guatemala ComparisonThe following narrative briefly compares Peruand Guatemala on selected dimensions ofgeography, population, economics,infrastructure, and culture. A summary of thesedimensions is presented as Tables 1 and 2.

Peru and Guatemala share a similar colonialhistory. Both countries had established cultures(Peru primarily Andean and Guatemala primarilyMaya) until their conquests by Spain in the 16th

century. Both gained their independence in the19th century (cia.gov). Both have struggled withvarious forms of governance sinceindependence.

Otherwise, the two countries differ. As shown inTable 1, compared to Guatemala, Peru is nearlytwelve times as large geographically, has aboutdouble the population, has a higher percentageof urban population, has a workforce that ismore agricultural and industrial, has a GrossDomestic Product about four times the size ofGuatemala’s, and has a varied climate (an aridlowland coastal region, the central high sierra ofthe Andes, the dense forest of the Amazon, withtropical lands bordering Colombia and Brazil)while Guatemala’s is tropical. Finally, Peru’spublic sector is somewhat less corrupt thanGuatemala’s (www/transparency.org, 2014). An

examination of the two cultures using the HofstedeCultural Dimensions (www.gert-hofstede.com,2014) revealed that, except for Power Distance(less concentration of authority in Guatemala) bothcountries are similar in Uncertainty Avoidance,Individualism/Collectivism, and Masculinity/Femininity.

Overall, the two countries are similar in havingbeen Spanish colonies for about three centuries,share the Spanish language, do not differ greatlyin terms of culture, and differ modestly in termsof public sector corruption. However, the twocountries differ in geographical size, populationsize, size of GDP, level urbanization, work forcemake up, climate and infrastructure.

From a logistical point of view, we can also viewthe relationship of Guatemala and Peru throughthe lens of the logistics performance index. Thisindex scores countries on their logisticsperformance according to six factors. Thesefactors are important in evaluating theeffectiveness of each country in terms of theiroverall logistical performance annually. The sixfactors include customs, infrastructure,international shipments, logistical competence,tracking and tracing, and timeliness. Bothcountries have very similar scoring records forthe year ending 2014. The Logistics PerformanceIndex in Table 3 summarizes a comparison oflogistical performance sores. Very little variationexists between Guatemala and Peru.

Fall/Winter 2016 13

The authors believe that Peru and Guatemala wouldprovide a good basis for comparing logistics/supplychain management strategies between two countriesin a region that shares characteristics in the areas ofhistory and culture but differ in many ways asdescribed above.

Objectives of the Study:One gap in this stream of cross-cultural logisticsstrategy research has been a lack of comparisonsbetween countries in one geographical-culturalarea. The authors were able to gatherinformation in Peru, which could then becompared with previously gathered data fromGuatemala. If the results from the two countrieswere similar then the authors thought that theywould have more confidence in generalizing theBowersox/Daugherty typology to the Latin-American region. Conversely, if the results fromPeru and Guatemala were dissimilar then itwould be concluded that the Bowersox/Daugherty model was not robust in that region.

Therefore, our interest in this study is to explorewhether the Bowersox/Daugherty typology is auseful instrument for examining logisticsstrategies in two dissimilar Spanish languagecountries located in Latin America. The authorspostulate that a two-country study of Guatemalaand Peru would furnish an intriguing example ofhow logistics systems are assessed in twonations through the lens of one commonmeasurement instrument. Furthermore, such astudy would provide a strong validation of thedimensionality and the structural relationsidentified in the recent Kohn, McGinnis, andKara (2011) study. We emphasize that thedifferences in each country’s geographic size,population size, labor force make-up,infrastructure, and economic system provides anexcellent platform for evaluating the validity ofthe research instrument, as well as providinginsights into logistics strategies and outcomes inthese heterogeneous countries.

METHODOLOGY

Measures and Questionnaire DevelopmentTo conceptualize the factors of our researchmodel, we used two sets of scales adapted fromthe McGinnis, Kohn, and Spillan (2010) study.In the first set the overall logistics strategy of thecompanies was measured on three dimensions;process strategy, market strategy andinformation strategy. The second set focused onthree dependent variables; logistics coordinationeffectiveness, customer service effectiveness, andcompany/division competitiveness. Respondentswere requested to determine their level ofagreement with three statements for process,market and information strategies for theircompany /division, for three statementsregarding logistics coordination effectiveness,customer service effectiveness, and for fourstatements regarding company/divisioncompetitiveness on a five point -type scale (1 =definitely agree, 5=definitely disagree).

Data CollectionTo collect data in Peru, the authors used theMcGinnis and Kohn survey. Articles based onthis instrument are found in McGinnis and Kohn(1993), Kohn and McGinnis (1997a), and latercited work. A bilingual associate translated theinstrument into Spanish. Back translation wascompleted to check any discrepancy in additionto potential translation errors. One of the co-authors trained 27 students by explaining tothem the purpose of the survey, what its contentswere, how to complete the survey and how torespond to questions from the respondents. Afterthe training, the students conducted face-to-faceand e-mail interviews with representatives fromsmall companies located in nine major regionalcenters in Peru. The students interviewedcompany representatives from 300 companiesand received 138 usable responses. We believethat the respondents are a reasonable sample ofPeruvian businesses involved in business logistics.

In Guatemala, as reported by McGinnis, Spillan,and Virzi (2012), one of the co-authors workedthrough the Ministry of Economics to collectdata. Ministry of Economics staff was trained to

Journal of Transportation Management14

administer the survey. After the training wascomplete, the Ministry of Economics staffconducted face-to-face interviews withrepresentatives from midsize and largecompanies located in nine major regional centersin Guatemala, providing a sample across a largegeographic area and a substantial cross-sectionof the Guatemalan business sector.

The authors decided that the Peruvian andGuatemalan data were collected in a manner thatenables a defensible basis for a comparison oflogistics/supply chain management strategies inthe two countries. The three independentvariables and three dependent variables used inthis research are presented as Table 4. Includedin Table 4 are the items for each variable and thescale reliabilities in Peru and Guatemala.Previous research (Kohn and McGinnis, 1997b)has concluded that the six variables are validwhen studying logistics strategy using logisticsmanagers in manufacturing firms.

ANALYSIS AND RESULTS

The first step was to check the constructreliabilities. For purposes of comparison theresults from the Peru survey and the previouslygathered data for Guatemala (McGinnis, Spillan,and Virzi, 2012) are shown as Table 4. The alphacoefficients for reliability for the threeindependent variables (Process Strategy, MarketStrategy, and Information Strategy) were higherfor the Peru respondents. In the case of ProcessStrategy, the alpha for Peru was significantlyhigher (0.725) than for Guatemala (0.524). Thealphas for the dependent variables variedbetween the two countries. For LogisticsCoordination Effectiveness and CustomerService Commitment, Guatemala’s alpha washigher (0.733 versus 0.684 and 0.634 versus0.430 respectfully) while Peru’s alpha forCompany/Division was higher (0.752 versus0.532) than Guatemala’s. Overall, the authorsconcluded that the reliability of the six variableswas adequate for further analysis.

Although some of the reliability scores were belowthe suggested levels (0.70) in the literature, ingeneral we can make a case that these scores aresatisfactory for testing and validating the structurereported in Kohn, McGinnis, and Kara (2011).Alpha is not a good indicator of unidimentionalityand low levels of alpha can be attributed to thesample homogeneity (Bernardi 1994) and do notput the results in question. Usually 0.70 is desiredbut Schmitt (1996, p. 351) states that “…use of anycutoff value is shortsighted.” Accordingly, when ameasure has other desirable properties, the lowalpha scores may not be a major impediment toits use (Schmitt, 1996). In addition, ascoefficient values are relatively receptive to thenumber of items in the constructs, particularlywhen constructs have fewer than 10 items, as inthe case in this research, it is common to findcoefficient alphas around 0.50 (Pallant, 2007).For instance, almost all alphas reported in Rojas-Mendez and Davies (2005) study was below thecutoff suggested in the literature. The scaleitems used in our study have been previouslyused in several studies in the literature; havebeen considered as having sufficient contentvalidity (Kohn and McGinnis, 1997a), andpossessing adequate levels of reliability. Allconstructs have been previously described anddiscussed by Keller et al. (2002). Previousstudies that used these scales also reported lowalpha scores (Kohn, McGinnis, and Kara, 2011).

Based on the findings shown in Tables 3 and 4,the authors concluded that a comparison ofmodeling the Peru data using the Bowersox/Daugherty typology, and comparing those resultswith the previously modeled Guatemalan data(McGinnis, Harcar, Kara, and Spillan, 2011),would provide insights into differences andsimilarities of logistics/supply chainmanagement strategies between two LatinAmerican economies.

Table 5 provides further insights into the two datasets. First, the Kaiser-Meyer-Olkin measure ofsampling adequacy (KMO-MSA) (Kaiser, 1970)and Bartlett’s test for sphericity was conducted for

Fall/Winter 2016 15

Journal of Transportation Management16

the 2 data sets. In addition the mean scores for theconstructs in both countries were assessed. Thevalue of KMO-MSA was 0.845 for the Peruviansample and 0.900 for the Guatemalan sampleindicating the data were appropriate for factoranalysis. All KMO results were above 0.50, whichis the minimum cut off for factor analysis.Additionally all levels of significance for Bartlett’stest for sphericity are less than 0.000. KMO resultsalong with the Bartlett results indicate the data issuitable for factor analysis. Finally, the averagevalues for five of six variables of the Peru data werenumerically lower (stronger agreement) than forGuatemala, however, none of the averages of thesix variables differed by an amount that wassignificant (alpha = 0.05).

Confirmatory Factor Analysis

To confirm the underlying factor structure, theauthors conducted CFA on all datasets usingAMOS. We assessed the goodness of the fit ofthe models using various fit indices used inprevious studies, including the χ 2 statistic,normed fit index (NFI), non-normed fit index,(NNFI), comparative fit index (CFI) goodness offit index (GFI); standardized root mean, squareresidual (SRMR); and root mean square error ofapproximation (RMSEA). The two-stepapproach suggested by Anderson and Gerbing(1988) was used to first examine themeasurement model and then the structuralmodel. In the measurement model, thehypothesized relationship between the 9 logisticsstrategic orientations and the three first orderfactors were examined to understand how wellthe relationships fit the data. In the structural model,we examined the relationship between the three first

Fall/Winter 2016 17

order factors (PROCSTR, MKTGSTR, andINFSTR). The findings supported the underlyingfactor structure of the 19 items with correlatedfactors.

The results of the estimation of the first orderfactor model (Figure 1) revealed very strongresults for all datasets used as indicated byseveral different measures (χ2 GUATEMALA=48.65, and χ2 PERU= 43.81). As suggested byMcGinnis, Kohn, and Kara (2011), we allowed two

of the error terms to be correlated. The figures ofGFI and CFI, were all larger than or equal to for allthree countries (GFI GUATEMALA=0.944; CFIGUATEMALA=0.942; GFI PERU=0.937; CFIPERU=0.953).

The normalized chi-square (chi-square/degrees offreedom) of the CFA model was smaller than therecommended value of 3.0; the RMR was smallerthan 0.05, and the RMSEA were smaller than or

Journal of Transportation Management18

very close to 0.08 (RMSEA GUATEMALA=0.08and RMSEA PERU=0.078). Although χ2 value fortwo of the datasets were significant, due to thesensitivity of this measure, it was not considered amajor concern since the other fit indices showedstrong model fit. Accordingly, the results showedthat all loadings in the model were significant,leading us to conclude that the relationships betweenthe items and latent factors were confirmed by thethree datasets obtained from different countries.

Structural Equation ModelsThe structural model was used to test thehypotheses of all six factors tested in themeasurement model. The hypothesized structuralmodels for three datasets are shown in Figure 2and 3. Inspection of these exhibits revealed thatall linkages were significant and the directionsof relationships were as hypothesized for theGuatemala and Peru datasets. The model fits forboth datasets were good and above theacceptable levels mentioned in the literature (SeeFigure 2 and 3).

Overall, both Guatemala and Peru datasetssupported the hypothesized relationshipdirections and strength of the hypothesizedrelationships. Figures 2 and 3 also displaystandardized coefficients for the linkages, and r2

values for the variables. Finally, the values forChi-square (193.616 AND 166.511), p-value(0.000), GFI (0.866 and 0.875), CFI (0.910 and0.904), and RMSEA (0.08 and 0.072) indicate agood model fit for both datasets. As wediscussed earlier, the Overall Logistics Strategy(OLS) construct is a second-order construct andits three dimensions (MKTGSTR, INFOSTR,and PROCSTR) are first-order factors measuredby their respective indicators. Overall, bothGuatemala and Peru data supported thehypothesized relationship directions and strengthof the hypothesized relationships. The other threedata sets (1990, 1994, and 2008) supported thedirections of the hypothesized relationship directionsand provided faint to modest support of the strengthof the model’s relationships.

DISCUSSION AND CONCLUSIONS

While Peru and Guatemala share similarhistories regarding colonialism, and thenindependence from Spain; and generally sharesimilar cultures, there are substantial differencesregarding the two countries’ geographic size,size of economy, make-up of their populations,climate, and infrastructure. These differencessuggest that business practices, includinglogistics/supply chain management strategies,could differ substantially between the twocountries. However, the results presented in thisresearch suggest that the logistics/supply chainmanagement strategies of the two countriesshare more similarities than differences.

Overall, logistics/supply chain managementstrategies are not greatly affected by substantialgeographic, size of economy, population, andclimate differences between Peru andGuatemala. These findings are not inconsistentwith the findings of other cross-cultural researchcited earlier. If confirmed by subsequentresearch, the findings reported here suggest thatlogistics/supply chain management strategiesmay be similar in other Spanish speaking LatinAmerican countries.

The research reported in this manuscript offersopportunities for additional research in LatinAmerica and within other regions of the world.For example, little is known about logistics/supply chain management strategy amongcountries in South East Asia, the EuropeanUnion, Japan, and India. Perhaps furtherresearch would either further confirm the valueof the Bowersox/Daugherty typology or facilitatethe development of alternate frameworks thatwould be applicable across cultures andeconomies.

The author’ summary of both countries fit theOLSLCECSCCOMP model that has beenpreviously tested longitudinally in the UnitedStates and cross culturally in Guatemala, Turkey,Ghana, and China. Two conclusions that can bedrawn from this research are (a) logistics/supply

Fall/Winter 2016 19

Journal of Transportation Management20

chain management strategy in Peru is comparable tothat found in previous research and (b) bothPeruvian and Guatemalan logistics/supply chainmanagement strategies both fit theOLSLCECSCCOMP model well.Additional comparisons reported in the Appendix Ashow similar, but not identical, patterns of logistics/supply chain management strategies in Peru andGuatemala. In both countries 40-45% of thelogistics/supply chain management strategies wereIntense, 42-47% of the strategies were Moderate,and 11-13% of the strategies were Passive. Theresults of this second research approach reinforcethe previously stated findings that Peruvian andGuatemalan logistics/supply chain strategies, whilenot identical, are similar.

When the authors compared the results ofPeruvian respondents to the Guatemalanrespondents the differences were exhibited intwo different ways. First, the means ofindependent and dependent variables weresomewhat lower (Scale: 1 = Strongly Agree to 5= Strongly Disagree), indicating that thePeruvian respondents placed greater importanceon all independent and dependent variables, onaverage, than did the Guatemalan respondents.The differences in this could be because of thetype of managers completing the survey or theperception of logistics that exist among therespondents that were interviewed. The authorsdecided that these differences did notsubstantially affect the results shown in Tables 3and 4. Second, Process Strategy - PROCSTR(focus on controlling costs) was generallyconsidered to be less important (higher average)than Market Strategy – MKTGSTR(management of logistics activities to reducecomplexity faced by customers) and InformationStrategy – INFOSTR (focus on managingactivities to achieve greater inter-organizationalcoordination and collaboration throughout thechannel). This contrasts with the findings ofPeruvian logistics managers where PROCSTRwas generally more important than MKTGSTR,and MKTGSTR was less important thanINFOSTR. A possible explanation for the

difference in the relative order may be due to theperception of supply chain managementoperations and support services among Peruvianmanagers when compared with Guatemalanmanagers. Greater emphasis might be placed onhard measures of performance (PROCSTR).However, the supplemental analysis shown inAppendix A reinforces the authors’ conclusionthat logistics/supply chain strategies in the twocountries are similar.

Overall, the study of logistics strategy in Perusuggests that the approach is fundamentallysimilar to Guatemala’s. In other words, thedirection of the relationships among theconceptualized constructs tested in the SEMmodel were significant and explained a sizablevariation in COMP in both countries, whichprovided additional support for the robustness ofthe structural model in different culturalenvironments. However, some differences areapparent. First, the importance of the threeindependent variables and three dependentvariables appear to be stronger to the Perurespondents than Guatemalan respondents.Second, on closer inspection Peruvian logisticsdata places relatively greater emphasis oninformation (INFOSTR), coordination (LCE),customer service (CSC), and relatively lessemphasis on cost efficiency (PROCSTR) and(MKTGSTR), than Guatemalan managers.Possible reasons include (a) informationtechnology and communication along with fewercompetitors may reduce the need to emphasizecost control, and (b) more sophisticatedinformation systems can facilitate bettercommunication, coordination, and customerresponsiveness in more sophisticatedcommunication economies. The authors believethat (a) may be the determining reason, since thePeruvian economy ranks 61 on the GlobalCompetitiveness Index, while Guatemala ranks 86on the same study (World Economic Forum, 2013).

RESEARCH IMPLICATIONS

The results of the analyses and country comparisonsin this manuscript provide insight into logistics

Fall/Winter 2016 21

strategy in two similar cultures but differenteconomies. A comparison of the results from thePeru and Guatemala data suggest that logistics/supply chain management strategies do not differsubstantially. This enabled the authors to makesome generalizations regarding Peruvian andGuatemalan logistics/supply chain managementstrategies.

First, because the two economies aresubstantially different, the Bowersox/Daughertytypology appears to be an appropriate frameworkfor comparative logistics research. Second, therelationships among the independent variables(PROCSTR, MKTGSTR, and INFOSTR) andthe dependent variables (LCE, CSC, and COMP)were similar.

Differences between the findings in Peru andGuatemala studies may be due to size of theeconomy, size of population and manager’sperceptions of logistics and supply chaindifferences. This suggests that futurecomparative logistics research should include anunderstanding of other contributing factors suchas size of economy and management perceptiondifferences.

For logistics/supply chain management faculty,this research suggests that logistics frameworks,such as the Bowersox/Daugherty typologyshould not be considered as absolute. Rather,logistics frameworks should be considered asconcepts that are likely to vary somewhat withthe size of the economy, the nature of theeconomy (agricultural, industrial, post-industrial), and the culture of the population.

For logistics practitioners, these findings suggest thatlogistics strategies should consider whether anethnocentric (do things the way we do it in ourcountry), polycentric (tailor the logistics systems tobe unique for each country where business istransacted), or geocentric (a logistics system thatblends the needs of each country where business isconducted) approach is appropriate. Each of theseapproaches may be appropriate in different

situations. The crucial aspect is to consider thesethree options, and their respective advantages anddisadvantages.

For researchers, the Bowersox/Daughertytypology appears to be one framework that canbe useful when conducting comparative logisticsresearch. The authors believe that this typologycould be a useful tool for understanding logisticsstrategies in different countries. Further researchshould continue to assess the value of theBowersox/Daugherty typology for comparativelogistics research and examine differences, andthe cause of differences, of logistics strategiesbetween countries or economies.

REFERENCES

Bernardi, R.A. (1994), “Validating ResearchResults when Cronbach’s Alpha is below 0.70: AMethodological Procedure,” Educational andPsychological Measurement, 54 (3): 766-775.

Bowersox, D. J. and Daugherty, P.J. (1987),“Emerging Patterns of Logistical Organization,”Journal of Business Logistics, 8(1): 46-60.

CIA Factbook (2013). https://www.cia.gov/library/publications/the-world-factbook/geos/pe.html Accessed April 2014.

Clinton, S.R. and Closs, D.J. (1997), “LogisticsStrategy: Does it Exist?” Journal of BusinessLogistics, 18(1): 19-44.

Geert Hofstede: Cultural Insights (www.geert-hofstede.com). May 14, 2014.

Heskett, J. L. (1977) Logistics: Essential toStrategy, Harvard Business Review, 55 (6): 85-95.

Kohn, J. W. and McGinnis, M. A. (1997),“Advanced Logistics Organization Structures:Revisited,” Journal of Business Logistics, 18(2): 147-162.

Journal of Transportation Management22

Keller, S.B., Savitskie, K., Stank, T.P., Lynch, D.F.,and Ellinger, A.E. (2002), “A Summary and Analysisof MultiItem Scales Used in Logistics Research,”Journal of Business Logistics, 23 (2): 83-119.

Kohn, J.W. and McGinnis, M.A. (1997a),“Logistics Strategy: A Longitudinal Study,”Journal of Business Logistics, 18 (2): 1-14.

Kohn, J.W. and McGinnis, M.A. (1997b),“Advanced Logistics Organization Structures:Revisited, “Journal of Business Logistics, 18(2): 147-162.

Kohn, J.W., McGinnis, M.A., and Kara, A.(2011), “A Structural Equation ModelAssessment of Logistics Strategy,” The Journalof International Logistics Management, 22 (3):284-305.

Luo, W., Van Hoek, R.I. and Ross, H.H. (2001),“Cross-Cultural Logistics Research: A LiteratureReview and Propositions,” International Journalof Logistics Research and Applications, 4(1): 57-78.

Logisticworld, (2015). http://www.logisticsworld.com/logistics.htm

McGinnis, M.A. and Kohn, J.W. (1993),“Logistics Strategy, OrganizationalEnvironment, and Time Competitiveness,”Journal of Business Logistics, 14(2): 1-23.

McGinnis, M.A. and Kohn, J.W. (2002),“Logistics Strategy – Revisited,” Journal ofBusiness Logistics, 23 (2): 1-17.

McGinnis, M.A., Harcar, T., Kara, A., and Spillan,J.E. (2011), “Cross-Cultural Validation of theFactorial Structure of a Logistics Strategy Model: AThree-Country Study,” Journal of TransportationManagement, 22(2): 25-43.

McGinnis, M.A., Spillan, J.E., and Virzi, N. (2012),“An Empirical Study Comparing Guatemalan andUnited States Logistics Strategies,” TheInternational Journal of Logistics Management,23(1): 77-96.

McGinnis, M.A., Spillan, J.E., Kara, A., andKing, D.O. (2012), “A Comparison of LogisticsStrategy and Logistics Integration in the USAand Ghana,” Journal of TransportationManagement, 22(1): 27-43.

Pallant, F. (2007), SPSS Survival Manual: AStep-by-Step Guide to Data Analysis UsingSPSS, 3rd edition, Berkshire, England: OpenUniversity Press, Mc-Graw Hill Education. p.6.

Rojas-Méndez, J. I., Davies, G. (2005),“Avoiding Television Advertising: SomeExplanations from Time Allocation Theory,”Journal of Advertising Research, 45 (March):34-48.

Schmitt, N. (1996), “Uses and Abuses ofCoefficient Alpha,” Psychological Assessment, 8(4): 350-353.

Spillan, J.E., McGinnis, M.A., Kara, A., and LiuYi, G. (2013), “A Comparison of the Effect ofLogistics Strategy and Logistics Integration onFirm Competitiveness in the USA and China,”International Journal of Logistics Management,24(2): 153-179.

Transparency.Org- www.transparency.org, 2014.

World Economic Forum (2013) “GlobalCompetitiveness Report 2009-2010,” Geneva(2013).

Fall/Winter 2016 23

BIOGRAPHIES

John E. Spillan is Professor of Management at the University of North Carolina at Pembroke, School ofBusiness. He received a M.B.A. degree from the College of Saint Rose in Albany, New York and a Ph.D.from the Warsaw School of Economics. His research interests center on Crisis Management, InternationalMarketing, Entrepreneurship and International Business with specific interest in Latin America and EasternEurope. E-mail: jspillan@uncp.edu

Michael A. McGinnis, CPSM, C.P.M. is Professor of Business at Penn State University NewKensington Campus. He holds B.S. and M.S. degrees from Michigan State University and a D.B.A.degree from the University of Maryland. His research areas are purchasing, logistics strategy,negotiations, and supply chain management. E-mail: mam47@psu.edu

Ali Kara is Professor of Marketing at Penn State University York Campus. He holds a doctoratefrom Florida International University, Miami, Florida and an MBA degree from the University ofBridgeport, Connecticut. He has published numerous articles in Journal of Marketing Research,Journal of Advertising, International Journal of Research in Marketing, European Journal ofOperations Research, Omega, Journal of Small Business Management, Industrial MarketingManagement, Journal of Global Marketing, and International Journal of Logistics Management. Inaddition, Ali has made several national/international conference presentations. He currently teachesseveral marketing courses at Penn State York.

Cesar Antunez de Mayolo is Adjunct Professor of Marketing at Graduate School of Business,Universidad del Pacífico. MBA from PAD Business School, Universidad de Piura, MSc (Abd) inComputer Science from Pontificia Universidad Católica del Peru, and industrial engineering fromUniversidad de Lima. Email: c.antunezdemayolof@up.edu.pe

Gustavo Jara is Professor of Service Operations Management at Universidad de Piura, MBA andPAD Business School, Universidad de Piura, and Metallurgical Engineering from Universidad deLima. Email: gustavo.jara@udep.pe

APPENDIX A

The purpose of the Appendix is to compare thecluster analysis of Peruvian logistics strategieswith a previous assessment of Guatemalanlogistics strategies.

Three independent variables were clusteranalyzed to ascertain whether Peruvian logisticsstrategies were homogenous, and if not in whatway were they heterogeneous. SPSS 16.0’sTwo Step Cluster was used in this step. Asshown in Table A-1, three logistics clusters,named Intense Logistics Strategy (N=57),

Moderate Logistics Strategy (N=65), and PassiveLogistics Strategy (N=16) were identified. Asshown in Table A-1, the means of Process,Market, and Information strategies (PROCSTR,MKTGSTR, and INFOSTR respectively) weresignificantly different, alpha<0.05, among thethree logistics strategy clusters. Post hoc testsdid not identify any pairing of independentvariables. Post hoc analysis did not identifypairing of dependent variables. Within Clusters1, 2, and 3 there were no pairs of PROCSTR,MKTGSTR, or INFOSTR that were significantat alpha<0.05 using the paired t-test of variables.Overall, the means of PROCSTR, MKTGSTR,

Journal of Transportation Management24

and INFOSTR were significantly different atalpha<0.05.

As a comparison, a similar analysis ofGuatemalan data was adapted from McGinnis,Spillan, and Virz (2012) and is presented asTable A-2. Using the same criteria for Intense,Moderate, and Passive Logistics Strategies, itwas observed that the percentages of Peru/

Guatemala respondents categorized as IntenseLogistics Strategy (41.3/44.1%), ModerateLogistics Strategy (47.1/42.5%), and PassiveLogistics Strategy (11.6/13.4%) were similar.The differences in percentages, ranging from1.8% to 4.6%, did not suggest an underlyingdifference between logistics/supply chainmanagement strategies between the twocountries

Fall/Winter 2016 25

Next, the means of dependent variablesLogistics Coordination Effectiveness (LCE),Customer Service Commitment (CSC), andCompany/Division Competitiveness (COMP)were tested for significant differences among thethree logistics strategy clusters. As shown in

Table A-3, LCE, CSC, and COMP were eachsignificantly different, alpha<0.05, among theclusters.

Post hoc analysis did not identify pairing ofdependent variables. Within Clusters 1, 2, and 3

Journal of Transportation Management26

there were no pairs of LCE, CSC, or COMP thatwere significant at alpha<0.05 using the paired t-test of variables. Overall the means of LCE,CSC, and COMP were significantly different atalpha<0.05. The following paragraphs discussthe findings based on the analysis. Aninspection of LCE, CSC, and COMP in the threeclusters for both countries found that the values

for Intensive Logistics Strategy differed verylittle. However, in all three strategies the dataindicated that CSC was substantially moreimportant (lower average values) in Peru withdifferences of LCE and COMP being slight.These results was consistent with the results ofprevious Guatemalan data shown in Table A-4

Fall/Winter 2016 27

Overall, Peruvian logistics can be summarizedas grouping into three distinct overall strategies.This result is not inconsistent with earlier in theUnited States (McGinnis, Kohn, and Spillan,2010), Guatemala (McGinnis, Spillan, and Virzi,2012), and China (Spillan, McGinnis, Kara, and

Liu Yi (2013). Based on the analysis presentedin this appendix the authors concluded thatlogistics/supply chain management strategies inPeru are not fundamentally different than thoseobserved in Guatemala and in other countriesstudied in previous similar research.