How do Firms in Different Sectors Organize their Supply Chains? Evidence from Transaction-Level Import Data

Board of Governors of the Federal Reserve System International Finance Discussion Papers ISSN 1073-2500 (Print) ISSN 2767-4509 (Online) Number 1405 February 2025 How do Firms in Different Sectors Organize their Supply Chains? Evidence from Transaction-Level Import Data Sebastian Heise, Justin R. Pierce, Georg Schaur, and Peter K. Schott Please cite this paper as: Heise, Sebastian, Justin R. Pierce, Georg Schaur, and Peter K. Schott (2025). “How do Firms in Different Sectors Organize their Supply Chains? Evidence from Transaction-Level Import Data,” International Finance Discussion Papers 1405. Washington: Board of Governors of the Federal Reserve System, https://doi.org/10.17016/IFDP.2025.1405. NOTE: International Finance Discussion Papers (IFDPs) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the International Finance Discussion Papers Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers. Recent IFDPs are available on the Web at www.federalreserve.gov/pubs/ifdp/. This paper can be downloaded without charge from the Social Science Research Network electronic library at www.ssrn.com.

How do Firms in Different Sectors Organize their Supply Chains? Evidence from Transaction-Level Import Data∗ Sebastian Heise † Justin R. Pierce ‡ Georg Schaur § Peter K. Schott ¶ January 31, 2025 Abstract Heise et al. (2021) develop a model-based empirical measure – sellers per shipment (SPS) – to characterize how firms organize supply chains in response to a quality control problem. High SPS indicates spot-market purchasing with costly inspections, while low SPS suggests long-term relationships where buyers pay an incentive premium to prevent cheating. Here, we document intuitive variation in US importers’ SPS across sectors, and that show shipping characteristics such as average price, quantity shipped and shipment frequency are in each sector consistent with the model of sourcing developed in Heise et al. (2021), providing further confidence in the measure. JEL Codes: F13, F14, F15, F23 Keywords: Supply Chain, Uncertainty, Trade War, Procurement ∗The views and opinions expressed in this work do not necessarily represent the views of the Federal Reserve Bank of New York, the Census Bureau, the Board of Governors of the Federal Reserve System, or its research staff. The Census Bureau has reviewed this data product to ensure appropriate access, use, and disclosure avoidance protection of the confidential source data used to produce this product. This research was performed at a Federal Statistical Research Data Center under FSRDC Project Number 1883 (CBDRB-FY21-P1883-R9019, CBDRB-FY22-P1883-R9643). †Federal Reserve Bank of New York; sebastian.heise@ny.frb.org. ‡Board of Governors of the Federal Reserve System; justin.r.pierce@frb.gov §University of Tennessee; gschaur@utk.edu ¶Yale School of Management & CEPR & NBER; peter.schott@yale.edu 1

In recent years, tariffs and the threat of future trade wars have forced firms to reconsider how they source goods from abroad. The academic and public discourse has often focused on how these risks might affect where multinational firms locate their foreign affiliates, with “nearshoring,” “friendshoring,” and “reshoring” suggested as possibilities. In earlier work (Heise et al., 2021), we highlight the importance of another element of firms’ international sourcing affected by the risk of trade wars: firms’ organization of supply chains, and specifically their choice of procurement system.1 By procurement system, we mean the order frequency, order size, price paid, and inspectionregimethatbuyerfirmsusewhenpurchasinggoodsfromaseller. Aseminal paper on the choice of such procurement systems is Taylor and Wiggins, 1997, which showsthatabuyercanensurethatsuppliersprovidehigh-qualitygoodseitherthrough spot-market purchases with costly inspections—which Taylor and Wiggins, 1997 call the “American” system—or by paying an incentive premium as part of a long-term buyer-seller relationship, which they call the “Japanese” system. The model predicts that the “American” system involves large and infrequent orders at low prices, while “Japanese” procurement is associated with small and frequent purchases at higher pricesduetotheincentivepremium. Heiseetal.,2021extendtheTaylorandWiggins, 1997 framework to international procurement and show that a higher likelihood of trade wars is associated with less “Japanese” sourcing. They also show how to test the model’s implications empirically and use transaction-level U.S. import data to provide the first evidence consistent with the mechanisms in Taylor and Wiggins, 1997. Heise et al., 2021 introduce a model-based empirical measure that can be used to classify firms’ procurement systems: the ratio of the number of sellers to the number of shipments (SPS). The measure leverages the model’s prediction that firms purchasing under the “American” system will source goods from many foreign sellers, while those engaged in long-term relationships will purchase from fewer or even a single seller. Heise et al., 2021 show that after using SPS to classify firms’ imports by procurement system, their order patterns are consistent with other key implications of the model. In particular, those procuring goods from relatively fewer suppliers place smaller shipments at higher frequency and pay higher unit values, consistent with the “Japanese” system. This paper complements the findings in our earlier work by providing a detailed 1A more recent version of that paper is Heise et al. (2024). 2

analysis of the choice of procurement system by firms’ major sector of activity. Recently, there has been growing interest in using empirical measures such as SPS to examine and characterize relationships between buyer and seller firms. Much of this literature has focused on applications to specific goods or sectors. For example, Cajal-Grossi et al., 2023a use the SPS measure developed by Heise et al., 2021 to examine “relational” buyer-seller relationships in the Bangladeshi garment industry. Cajal-Grossi et al., 2023b then use the SPS measure to examine the effect of Covidrelated supply chain disruptions on procurement strategies in the garment sector for six developing countries. In this paper, we use confidential data for U.S. import transactions to provide descriptive statistics on the use of procurement systems for a broad range of sectors. We classify importers’ procurement systems using SPS and show that the finding of more-frequent, smaller, and higher-priced imports within long-term buyer-seller relationships predicted by the model in Heise et al., 2021 is remarkably stable across different sectors. 1 Which Sectors Use “Japanese” Sourcing? As in Heise et al., 2021, we characterize procurement systems using confidential transaction-level data from the the US Census Bureau’s (Census) Longitudinal Foreign Trade Transaction Database (LFTTD). Our dataset covers the years 1992 to 2016, and we restrict imports to those that are “arm’s length,” or between unrelated firms.2 When examining import behavior, our unit of observation is an importer m sourcing a good h from country c via mode of transportation z, which we refer to as a “quadruple.” This level of aggregation helps isolate obvious sources of variation in observed price and quantity. Following our earlier work, we classify buyer quadruples’ procurement systems using the ratio of the number of sellers used to the number of shipments received: Sellers mhcz SPS = . (1) mhcz Shipments mhcz 2Census considers firms to be related if either party owns a 6 percent or greater share of the other. 3

Heise et al., 2021 provides some statistics on the distribution of SPS across mhcz quadruples, and here we focus on heterogeneity across sectors. The first two columns of Table 1 provide measures of the mean SPS by major sector of the importing firm m, for two periods, 1995-2000 and 2002-2007.3 There is substantial variation in procurement systems across sectors. Transportation andWarehousing, the sector withthe highest ratio of sellers pershipment, has an SPS in both periods that is nearly twice as large as that for the sector with the lowest value of SPS, manufacturing. This finding suggests that manufacturers are substantially more engaged in longer-term relationships than transport and warehouse firms, with the latter engaged more in spot-market sourcing. While our SPS measure allows us to delineate which relationships appear more “Japanese” than others, it does not define a formal threshold. To provide some guidance for the importance of “Japanese” sourcing, Heise et al., 2021 define a quadruple as being engaged in “Japanese” sourcing (Jcz = 1) if SPS falls in the first mhcz mhcz quartile of the distribution of SPS within a country-mode bin in the 1995-2000 mhcz period. The third and fourth columns of Table 1 report the share of the value of U.S. imports accounted for by quadruples with Jcz = 1. Going forward, we refer to mhcz “Japanese” sourcing as J and to “American” sourcing as A. As shown in the table, J quadruples account for a disproportionately large share of import value in all sectors. But again, there is substantial variation across sectors, with the share of J trade for manufacturers in 2002-2007 over 25 percentage points higher than that for transportation and warehousing. Examining changes over time, the prevalence of J procurement has increased in most sectors, as evidenced both by declining SPS in columns 1 and 2 and an increasing share of import value associated with J quadruples in columns 3 and 4. Two exceptions to this upward trend are the high-wage services sectors of “Professional Services”and“FinanceandInsurance,” whichlikelydonotuseimportedgoodsintensively in their production functions. The largest shift toward longer-term buyer-seller relationships between the 1995-2000 and 2002-2007 periods occurs in the retail sector, which saw a 15 percentage point increase in the share of import value occurring under J procurement. 3The major sector of the firm is based on employment across sectors. 4

Table 1: “Japanese” Relationships by Main Industry of the Importer Jcz =1 MeanSPS mhcz ShareofImportValue (1) (2) (3) (4) Industrycode(NAICS) 1995-2000 2002-2007 1995-2000 2002-2007 Manufacturing(31-33) 0.119 0.113 0.739 0.778 Agriculture(11) 0.123 0.106 0.584 0.630 Wholesale(42-43) 0.158 0.128 0.623 0.729 Other services 0.160 0.130 0.655 0.713 Professionalservices(54-55) 0.177 0.220 0.586 0.415 Mining,utilitiesandconstruction(21-23) 0.182 0.131 0.561 0.684 Financeandinsurance(52-53) 0.187 0.213 0.516 0.514 Retail(44-45) 0.208 0.157 0.532 0.688 Information(51) 0.211 0.182 0.553 0.566 Adminsupport&wastemgmt(56) 0.213 0.195 0.312 0.423 TransportationandWarehousing(48-49) 0.216 0.210 0.487 0.511 Notes: SourcesareLFTTDandauthors’calculations. Columns1and2reporttheweightedaveragesellersper shipment(SPS )acrossbuyerquadrupleswithatleastfivetransactionsbymain6-digitNAICS mhcz industry-period. ToobtainthemainNAICS,wefindineachyeartheindustrywiththeimporter’slargestshareof employment,andthentakethemodalmainindustryacrosstheyearsinwhichthequadrupleisactive. We aggregateSPS acrossquadruplesusingimportvaluesasweights. Columns3and4reporttheshareofthe mhcz valueofUSimportsaccountedforbyquadrupleswithSPS inthefirstquartileofthedistributionofSPS mhcz mhcz withincountry-modeinthefirstperiod. Rowsofthetablearesortedbythecolumn(1). 2 Shipping Patterns Within Procurement System, by Sector Heise et al., 2021 examine whether quadruples—once categorized by SPS—engage in shipping patterns consistent with their model. Pooling observations across all sectors, they show that, indeed, quadruples with lower values of SPS receive more frequent and smaller shipments at lower prices, consistent with the J system. They therefore argue that SPS, reproduced in equation 1, provides a model-based continuous measure of the extent of J or A sourcing for a given quadruple. In this paper, given the recent interest in sector-level empirical applications of the SPS measure, we perform a similar analysis examining how shipping patterns vary by SPS, separately, by major sector of U.S. importing firms. To do so, we estimate the following equation from Heise et al., 2021: ln(Y ) = β ln(SPS )+β ln(QPW ) mhcz 1 mhcz 2 mhcz +β beg +β end (2) 3 mhcz 4 mhcz +λ +(cid:15) . hcz mhcz 5

Table 2: SPS and Procurement Attributes - Manufacturing mhcz (1) (2) (3) (4) Dep. var. ln(QPS ) ln(WBS ) ln(UV ) ln(length ) mhcz mhcz mhcz mhcz ln(SPS ) 0.500∗∗∗ 0.538∗∗∗ −0.181∗∗∗ −0.540∗∗∗ mhcz 0.014 0.014 0.022 0.012 log(QPW ) 0.769∗∗∗ −0.238∗∗∗ −0.367∗∗∗ −0.131∗∗∗ mhcz 0.018 0.018 0.022 0.008 Observations 560,000 560,000 560,000 560,000 Fixedeffects hcz hcz hcz hcz R-squared 0.950 0.712 0.816 0.434 Controls beg,end beg,end beg,end beg,end Notes: SourcesareLFTTDandauthors’calculations. Tablereportstheresultsofregressingnotedattributeof importerbyproductbycountrybymodeoftransport(mhcz)binsonbins’sellerspershipment(SPS )and mhcz totalquantityshippedperweek(QPW ). Industriesareassignedusingthemain6-digitNAICSindustryofthe mhcz importerbasedontotalemployment. QPS ,WBS ,UV ,andlength areaveragequantityper mhcz mhcz mhcz mhcz shipment,averageweeksbetweenshipment,averageunitvalue,andaveragerelationshiplength. Allregressions includeproductbycountrybymodeoftransport(hcz)fixedeffects,controlforthebeginningandendweekofthe quadruple,andexcludequadrupleswithlessthanfiveshipments. Standarderrors,adjustedforclusteringbycountry (c)andproduct(h)arereportedbelowcoefficientestimates. ***,**,and*representstatisticalsignificanceatthe1, 5and10percentlevels. The dependent variable consists of a set of shipping characteristics that the model in Heise et al., 2021 predicts will change based on the choice of procurement system. These shipping characteristics include average quantity per shipment (QPS ), mhcz weeks between shipments (WBS ), average unit value (UV ), and average mhcz mhcz length of the buyer(m)-seller(x) relationships within mhcz buyer quadruples. The key independent variable is SPS , the model-based measure of a quadruple’s promhcz curement system. Other controls include the quantity per week imported by the quadruple (as called for by the Heise et al., 2021 model), controls for the beginning and end period of a quadruple’s trading activity (to capture effects of trading in a given time period), and product by country by mode of transportation fixed effects (λ ). We estimate equation 2 separately for firms in three sectors that are intenhcz sively engaged in international trade: Manufacturing, Wholesale, and Retail. Results are presented in Tables 2 - 4. Beginning with Manufacturing (Table 2), we find that shipping characteristics are related to SPS in ways predicted by the model and are consistent with the results for the pooled sample in Heise et al., 2021. In particular, a higher SPS, which indicates a greater reliance on the spot market—and hence more A sourcing—is associated with larger shipment sizes, more time between shipments, a lower unit value, and shorter relationship lengths in the manufacturing sector. Examining results for the Wholesale and Retail sectors in Tables 3 and 4, respec- 6

Table 3: SPS and Procurement Attributes - Wholesale mhcz (1) (2) (3) (4) Dep. var. ln(QPS ) ln(WBS ) ln(UV ) ln(length ) mhcz mhcz mhcz mhcz ln(SPS ) 0.443∗∗∗ 0.475∗∗∗ −0.181∗∗∗ −0.571∗∗∗ mhcz 0.015 0.015 0.013 0.020 log(QPW ) 0.682∗∗∗ −0.328∗∗∗ −0.281∗∗∗ −0.167∗∗∗ mhcz 0.012 0.012 0.017 0.007 Observations 1,215,000 1,215,000 1,215,000 1,215,000 Fixedeffects hcz hcz hcz hcz R-squared 0.945 0.708 0.856 0.469 Controls beg,end beg,end beg,end beg,end Notes: SourcesareLFTTDandauthors’calculations. Tablereportstheresultsofregressingnotedattributeof importerbyproductbycountrybymodeoftransport(mhcz)binsonbins’sellerspershipment(SPS )and mhcz totalquantityshippedperweek(QPW ). Industriesareassignedusingthemain6-digitNAICSindustryofthe mhcz importerbasedontotalemployment. QPS ,WBS ,UV ,andlength areaveragequantityper mhcz mhcz mhcz mhcz shipment,averageweeksbetweenshipment,averageunitvalue,andaveragerelationshiplength. Allregressions includeproductbycountrybymodeoftransport(hcz)fixedeffects,controlforthebeginningandendweekofthe quadruple,andexcludequadrupleswithlessthanfiveshipments. Standarderrors,adjustedforclusteringbycountry (c)andproduct(h)arereportedbelowcoefficientestimates. ***,**,and*representstatisticalsignificanceatthe1, 5and10percentlevels. tively, indicates similar relationships between SPS and all four shipping characteristics, as indicated by the identical sign and significance of coefficients on the SPS variable and their highly similar magnitudes across sectors. In other words, while firms differ substantially across sectors in their choice of procurement system, the effect of changing procurement systems on shipping characteristics is remarkably robust across sectors. These results also illustrate that the results in Heise et al., 2021 are not driven by relationships for a single sector or group of sectors. 3 Conclusion This paper builds on Heise et al., 2021 by providing new analysis on U.S. firms’ choice of procurement systems by major sector. We provide descriptive statistics on the extent of long-term “Japanese” type procurement, showing substantial variation across sectors, with manufacturers most likely to use such systems. We also show— after classifying trade by procurement system—that buyers’ shipment characteristics align with those predicted by the model in Heise et al., 2021. This finding is robust across all sectors examined. Our results complement the findings in our earlier paper and the subsequent analysis by Cajal-Grossi et al., 2023a applying our SPS measure to the garment industry. 7

Table 4: SPS and Procurement Attributes - Retail mhcz (1) (2) (3) (4) Dep. var. ln(QPS ) ln(WBS ) ln(UV ) ln(length ) mhcz mhcz mhcz mhcz ln(SPS ) 0.424∗∗∗ 0.458∗∗∗ −0.120∗∗∗ −0.556∗∗∗ mhcz 0.030 0.031 0.023 0.022 log(QPW ) 0.643∗∗∗ −0.366∗∗∗ −0.195∗∗∗ −0.115∗∗∗ mhcz 0.007 0.007 0.012 0.008 Observations 525,000 525,000 525,000 525,000 Fixedeffects hcz hcz hcz hcz R-squared 0.945 0.708 0.856 0.955 Controls beg,end beg,end beg,end beg,end Notes: SourcesareLFTTDandauthors’calculations. Tablereportstheresultsofregressingnotedattributeof importerbyproductbycountrybymodeoftransport(mhcz)binsonbins’sellerspershipment(SPS )and mhcz totalquantityshippedperweek(QPW ). Industriesareassignedusingthemain6-digitNAICSindustryofthe mhcz importerbasedontotalemployment. QPS ,WBS ,UV ,andlength areaveragequantityper mhcz mhcz mhcz mhcz shipment,averageweeksbetweenshipment,averageunitvalue,andaveragerelationshiplength. Allregressions includeproductbycountrybymodeoftransport(hcz)fixedeffects,controlforthebeginningandendweekofthe quadruple,andexcludequadrupleswithlessthanfiveshipments. Standarderrors,adjustedforclusteringbycountry (c)andproduct(h)arereportedbelowcoefficientestimates. ***,**,and*representstatisticalsignificanceatthe1, 5and10percentlevels. References Cajal-Grossi,Julia,RoccoMacchiavello,andGuillermoNoguera(2023a).“Buyers’sourcingstrategiesandsuppliers’ markupsinbangladeshigarments”.The Quarterly Journal of Economics 138.4,2391–2450. Cajal-Grossi, Julia, Davide Del Prete, and Rocco Macchiavello (2023b). “Supply chain disruptions and sourcing strategies”. International Journal of Industrial Organization 90. The 49th Annual Conference of the European AssociationforResearchinIndustrialEconomics,Vienna,2022,103004.issn:0167-7187.doi:https://doi.org/ 10.1016/j.ijindorg.2023.103004. Heise, Sebastian, Justin R. Pierce, Georg Schaur, and Peter K. Schott (2021). “Tariff Rate Uncertainty and the StructureofSupplyChains”.CowlesInternationalTradeConference.url:https://www.sciencedirect.com/sc ience/article/pii/S002219961100122X. Heise,Sebastian,JustinRPierce,GeorgSchaur,andPeterKSchott(2024).Tariffrateuncertaintyandthestructure of supply chains.Tech.rep.NBERWorkingPaperNo.32138. Taylor, Curtis R. and Steven N. Wiggins (1997). “Competition or Compensation: Supplier Incentives under the AmericanandJapaneseSubcontractingSystems”.American Economic Review 87.4,598–618. 8