What Attracts Foreign Direct Investment Inflows in the United States

This article was downloaded by: [Northeastern University]On: 13 November 2014, At: 05:34Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

The International Trade JournalPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/uitj20

What Attracts Foreign Direct Investment Inflows in theUnited StatesKostas Axarloglou aa Athens Laboratory of Business Administration and Democritus , University of Thrace ,GreecePublished online: 23 Feb 2007.

To cite this article: Kostas Axarloglou (2005) What Attracts Foreign Direct Investment Inflows in the United States, TheInternational Trade Journal, 19:3, 285-308, DOI: 10.1080/08853900591007438

To link to this article: http://dx.doi.org/10.1080/08853900591007438

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/uitj20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/08853900591007438

http://dx.doi.org/10.1080/08853900591007438

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

ITJ 19(3) #68362

WHAT ATTRACTS FOREIGN

DIRECT INVESTMENT

INFLOWS IN THE

UNITED STATES

Kostas Axarloglou

The present study evaluates the relative impact of industryand state specific economic factors on inward FDI in several USstates that compete for the same inward FDI. It appears thatrelative labor productivity, relative spending on education, andrelative crime rate are important in inter-state competition forthe same inward FDI. Also, when the contest in attracting inwardFDI comes down to two states, relative tax incentives also becomeimportant in attracting FDI inflows.

* * * * *

I. INTRODUCTION

What are the economic conditions that make a local economyan attractive destination for FDI inflows? How important arethese conditions in the presence of increasing competition amonglocal economies in attracting these inflows? Especially for FDI

Kostas Axarloglou is Assistant Professor of Business Economicsat the Athens Laboratory of Business Administration and Dem-ocritus University of Thrace, Greece.

Jane Cloran and Nikolaos Kalios provided invaluable research as-sistance. Financial support from the Board of Research, BabsonCollege is gratefully acknowledged.

ISSN: 0885-3908 print/1521-0545 online. DOI: 10.1080/08853900591007438THE INTERNATIONAL TRADE JOURNAL, Volume XIX, No. 3, Fall 2005 285

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

286 THE INTERNATIONAL TRADE JOURNAL

flows in the United States, what are the state specific economicconditions that make a US state an attractive destination for FDIinflows relative to other states? These questions are addressedempirically by using state-specific FDI and economic data for asample of United States states that received the most FDI inflowsbetween 1974 and 1994.

In the relevant literature, there are studies that analyze theeffects of various economic variables in attracting FDI inflows.1

However, some of these studies are based on instances of FDI(Coughlin et al., 1991; Friedman et al., 1992) in the host countriesor states, and therefore, the empirical results measure only theimpact of various control variables on the likelihood of an FDIto happen.

Some other studies (Friedman et al., 1992) are based mostlyon aggregated data washing out thus interesting patterns due toaggregation. Moreover, some of these studies (Coughlin et al.,1991) depend on data, in a panel form (across host countriesor states), that span in most cases a short period of time whichappears to be quite incomplete (Hatzius, 2000). Our study avoidsthese drawbacks by using panel data that span a quite longperiod of time (21 years, from 1974 to 1994).

Finally, several studies focus on the conditions that attractinward FDI in the United States.2 However, very few of thesestudies deal with the relative effectiveness of state policies inattracting FDI from competing states, and when they do so, theyusually test their hypotheses indirectly.3 Instead, in the presentstudy, we test explicitly the relative effectiveness of industry and

1See for example, Culem (1988), Coughlin et al. (1990, 1991), Friedman et al.(1992), Head et al. (1995, 1999), Billington (1999), Hatzius (2000), and Axarloglou(2004).

2See Coughlin et al. (1990, 1991), Friedman et al. (1996), and Head et al.(1995, 1999).

3See Head et al. (1995, 1999).

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

K. Axarloglou: What Attracts Foreign . . . 287

state specific economic conditions and state policies in attractingFDI inflows, when states are in direct competition with otherstates in attracting the same inward FDI.

In section 2 a discussion motivates our empirical hypothesesand precedes our data description (section 3) and our empiricalresults (section 4). We conclude in section 5.

II. DISCUSSION

A multinational company establishes a foreign subsidiary inorder to take advantage of profitable opportunities in the foreignmarket. Obviously, a rather large group of factors influencesthe company’s decision to move to a foreign market. Someof them are related to the sales opportunities the companyforesees in supplying the foreign market, while some others areassociated with the lower production cost the company expectsby producing in the foreign market. Also, the stage and structureof the foreign labor market in which the company operates isimportant. Finally, a set of fiscal and economic incentives anddisincentives available in the foreign market play an importantrole in the company’s FDI location decision.

In the literature, there is significant evidence that the sizeof regional markets in the United States is important whendomestic companies choose the location for their plants.4 At thesame time, several researchers have shown that foreign companiesare also driven by market size in choosing the location for theirsubsidiaries in the United States.5

However, especially in industries with rather low transporta-tion cost, cost related issues might be more important in location

4See Bartik (1989) and Wheat (1986).5For example, Friedman et al. (1992), Woodward (1992), and Coughlin et al.

(1991).

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


decisions for foreign subsidiaries within the United States. Mostresearchers—such as Luger and Shetty (1985), Coughlin et al.(1991), Friedman et al. (1992), and Billington (1999), amongothers—used average real wages in manufacturing to proxy la-bor cost, and find that usually higher real wages tend to deterinward FDI. On the other hand, very few studies include laborproductivity among the factors that attract inward FDI, sincethey mostly depend on aggregate data that usually hides im-portant productivity variations either across industries or acrosslocations. Friedman et al. (1992) found that higher labor produc-tivity attracts Japanese FDI inflows in the United States but notnecessarily European FDI inflows, while Hatzius (2000) showedthat high labor productivity increases FDI inflows in Germanyand the UK.

Coughlin et al. (1991) and Friedman et al. (1992) foundthat high state unemployment in the United States attractsFDI inflows, arguing that state unemployment usually indicatesavailability of local labor necessary for establishing new plantsin the United States.6 However, Friedman et al. (1996) haveshown that FDI for Mergers and Acquisitions does not respondto unemployment since labor availability is not that importantfor an acquired by a foreign multinational domestic plant. Onthe other hand, Woodward (1992) found that Japanese start-upsare usually deterred by high state unemployment. Overall, theseresults must be interpreted with some caution since they dependon fairly aggregated unemployment data.

The effects of unionization on inward FDI are also mixed.Coughlin et al. (1991) and Friedman et al. (1992) showed thatunionization attracts FDI inflows, while Woodward (1992) foundexactly the opposite. However, these results might be due toaggregation bias in FDI data, since Friedman et al. (1996)

6As Friedman et al. (1996) have shown.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


showed that unionization attracts FDI inflows for Mergers andAcquisitions but have an insignificant impact on FDI inflows forNew Plants.

Industry agglomeration effects appear to be very importantand robust in attracting FDI inflows.7 Multinational companiestend to establish their foreign subsidiaries in locations thatare characterized by significant concentration of plants of thesame industry or same nationality with their own subsidiaries.Obviously, due to industry agglomeration, companies expect tobenefit from the local availability of a specialized labor force,the proximity of suppliers and a better transportation andcommunications network. Finally, agglomeration effects are alsoimportant for Swedish (Braunerhjelm and Svensson, 1996) andJapanese (Woodward, 1992) FDI inflows.

In recent years, state governments in the United Stateshave intensified their efforts to attract FDI inflows by offeringeconomic and fiscal incentives to foreign multinationals.8 Headet al. (1999) showed that state promotional policies tend to offseteach other, and thus have a small impact on the geographiclocation of Japanese FDI in the United States. On the contrary,Coughlin et al. (1991) and Friedman et al. (1992) found thatstate promotional spending attracts FDI inflows. At the sametime, the effects of state taxes on FDI inflows are at best mixed.Friedman et al. (1992) showed that state taxes usually deterinward FDI while Woodward (1992) and Coughlin et al. (1991)found mixed effects of state taxes on FDI inflows. Coughlinet al. (1990) attributed these effects to the fact that only sometypes of taxes appear to be important in deterring inward FDI

7See Woodward (1992), Barrell and Pain (1998), and Head et al. (1995), amongothers.

8“Mercedes Plant Said to be Set for Alabama,” Wall Street Journal,09/29/1993.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


(such as corporate income taxes or unit taxation), while othertaxes (such as property taxes) do not seem to have any impact onFDI inflows. On the other hand, Friedman et al. (1996) showedthat state and local taxes influence only some types of FDI.Consequently, aggregation bias either in terms of various taxes ortypes of FDI flows seems to be responsible for the mixed effectstaxes appear to have on inward FDI.

In the present study, by using a rich data set on FDI inflowsin several US states and all 2-digit SIC manufacturing industriesin each state for 21 years, we evaluate the relative effectiveness(in attracting FDI inflows) of various industry and state specificfactors among the United States states that compete for the sameamount of FDI inflows.

III. DATA

The empirical analysis depends on FDI inflows data in theUnited States states from 1974 to 1994 that have been recorded inthe publication, Foreign Direct Investment in the United States.9

The data report FDI transactions in the United States of foreigncompanies that posses at least ten percent of the ownership oftheir US subsidiary where the investment is placed.

The sample includes the dollar value of FDI inflows (FDI ijt)in all twenty 2-digit SIC manufacturing industries (index i forindividual industries) in each of the ten United States states(index j for individual states) in the sample that received themost FDI inflows between 1974 and 1994 (index t for time) basedon the ITA data set.10 The description of the rest of the dataappears in the Appendix.

9These data were maintained by the International Trade Administration(ITA), the US Department of Commerce, and were discontinued after 1994.

10The ten states in the sample were the top states in total FDI received forthe period between 1974 and 1994. These states together received approximately63% of all FDI inflows in US manufacturing that recorded by the ITA from 1974to 1994.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Tables I, II, and III present a break down of FDI inflowsacross states and industries. The state of New York received24% of the total FDI in the sample (Table III) and alongwith California and Ohio (13% of the total each) account forapproximately half of the total FDI that arrived in the tenstates of the sample. On the other hand, Chemicals (SIC 28) andElectrical Machinery (SIC 36) top the list of industries in terms ofFDI inflows (with 30% and 13% of the total respectively; Table I).Finally, the industry allocation of FDI inflows within the samestate is not homogeneous (Table II). In several states a singleindustry is the dominant recipient of the state’s FDI inflows,like Chemicals in New York and Pennsylvania (43% and 64% ofthe state’s total respectively), and Stone/Clay/Glass (SIC 32) inFlorida (40% of the state’s total).

IV. EMPIRICAL RESULTS

Anecdotal evidence points out that United States states com-pete among themselves through fiscal incentives and promotionalcampaigns in order to attract FDI inflows in so-called “locationtournaments” (Wheeler and Mody, 1992).11 Consequently, giventhe value of US inward FDI in each manufacturing industry, itis important to study the factors that influence its allocation tovarious states. For that, a dependent variable is constructed asthe share of industry and state specific FDI inflows out of theFDI received by the same industry in all ten states in our sample(RFDI ijt). Also, the independent variables in estimations shouldproxy relevant factors such as industry and state specific wages,labor productivity and unit labor cost, industrial agglomeration,the local labor market conditions, the state promotional spendingon transportation infrastructure and education, the state taxa-tion, and, finally, the state’s specific crime rate.

11See, for example, the Wall Street Journal article “Mercedes Plant Said toBe Set for Alabama,” 9/29/1993.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

Tab

leI

FDIInflowsAcrossIndu

stries

andStates

Massa-

New

New

Pen

nsyl-

Industry

California

Con

necticu

tFlorida

Illinois

chusetts

Jersey

York

Ohio

vania

Texas

Total

2047

59.7

1208

1287

.527

50.3

63.7

57.7

1844

.823

7.7

340.8

474.9

1302

5.1

210

00

0.9

041

50

00

5.2

421.1

2274

015

220

.534

217.4

32.5

42.5

043

7.9

2330

.40

030

013

216.9

228

2.9

052

1.2

2433

0.1

00

00

04.8

100

034

4.9

2515

00

8.8

0.9

019

.238

100

91.9

2670

5.6

013

078

4.6

1843

.41.7

869.8

147.8

694.8

4052

17.7

2748

7.5

705.6

179.5

1585

.340

8.2

502.1

9131

.181

017

0.3

340

1431

9.6

2817

48.01

4809

.310

1.3

2816

.514

16.4

4896

1951

3.4

2025

.513

461.3

4829

.355

617.01

2994

10.3

076

950

219

12.4

8291

.642

047

09.4

1541

2.7

3016

3.1

260.5

67.5

44.3

129

113.1

271.3

5283

.911

380

.265

25.9

310

00

063

.21.9

170

7.8

089

.932

724.8

701.1

1921

.24.3

2004

135.6

395.7

643.9

673.4

2251

.194

55.1

3318

65.4

209

407

88.8

71.2

216.7

1126

.817

59.9

1704

.656

8.8

8018

.234

123.7

175.3

45.6

354.5

122.5

248.5

1329

.181

5.6

325.7

6436

04.5

3534

33.9

2967

.888

.818

03.8

1183

.287

2.4

538.1

369.3

910.2

505.1

1267

2.6

3661

53.4

1552

.542

9.7

2894

.915

83.8

821

7461

.415

01.3

804.2

1122

.524

324.7

3713

63.8

25.1

28.6

731

03.3

423.5

2166

966.6

29.3

5737

.238

2172

.960

.474

349.5

1004

.964

9.3

1349

.224

8.7

352.5

840.5

7101

.939

26.9

14.5

036

30.6

2043

.419

8.8

3.3

41.7

026

92.2

Total

2511

9.21

1268

9.4

4775

.715

381.5

9965

.511

243.7

4494

0.7

2461

321

042.3

1586

0.3

1856

31.3

Not

es:EachentryshowsthecurrentdollaramountofFDIinflowsinaspecificindustryandUSstatefor1974–1994.

292

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

Tab

leII

Indu

stry

Com

position

ofFDIInflowsforEachIndividu

alState

Industry

California

Con

necticu

tFlorida

Illinois

Massachusetts

New

Jersey

New

York

Ohio

Pen

nsylvan

iaTexas

200.18

950.09

520.26

960.17

880.00

640.00

510.04

110.00

970.01

620.02

9921

00

00.00

005

00.03

690

00

0.00

0322

0.00

290

0.00

310.00

010.00

210.00

300.00

480.00

130.00

200

230.00

120

00.00

190

0.00

120.00

480.00

930.00

010

240.01

310

00

00

0.00

010.00

040

025

0.00

060

00.00

060.00

009

00.00

040.00

150.00

050

260.02

810

0.02

720.05

100.18

490.00

010.01

930.00

600.03

300.00

2527

0.01

940.05

560.03

760.10

310.04

090.04

470.20

320.03

290.00

810.02

1428

0.06

960.37

900.02

120.18

310.14

210.43

540.43

420.08

230.63

970.30

4529

0.03

750.00

002

00.04

990.00

500.01

950.00

030.33

690.01

990.29

6930

0.00

650.02

050.01

410.00

290.01

290.01

010.00

600.21

470.00

540.00

5031

00

00

0.00

630.00

020.00

040

0.00

040

320.02

890.05

530.40

230.00

030.20

110.01

210.00

880.02

620.03

200.14

1933

0.07

430.01

650.08

520.00

580.00

710.01

930.02

510.07

150.08

100.03

5934

0.00

490.01

380.00

950.02

300.01

230.02

210.02

960.03

310.01

550.00

4035

0.13

670.23

390.01

860.11

730.11

870.07

760.01

190.01

500.04

330.03

1836

0.24

490.12

230.08

990.18

820.15

890.07

300.16

600.06

100.03

820.07

0837

0.05

430.00

190.00

600.04

750

0.00

030.00

940.08

800.04

590.00

1838

0.08

650.00

480.01

550.02

270.10

080.05

770.03

000.01

010.01

680.05

2939

0.00

110.00

110

0.02

360.00

006

0.18

170.00

440.00

010.00

200

Not

es:EachentryshowstheFDIinflowsinaspecificindustryandstateasashareoftotalinflowsinthesameUSstatefor

1974–1994.

293

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

Tab

leIII

StateCom

position

ofFDIInflowsforEachIndividu

alIndu

stry

Industry

California

Con

necticu

tFlorida

Illinois

Massachusetts

New

Jersey

New

York

Ohio

Pen

nsylvan

iaTexas

200.36

540.09

270.09

880.21

110.00

490.00

440.14

160.01

820.02

620.03

6521

00

00.00

210

0.98

550

00

0.01

2322

0.16

890

0.03

430.00

460.04

680.07

760.49

650.07

420.09

700

230.05

830

00.05

760

0.02

490.41

620.43

750.00

560

240.95

710

00

00

0.01

390.02

890

025

0.16

320

00.09

580.00

980

0.20

890.41

350.10

880

260.13

520

0.02

490.15

040.35

330.00

030.16

670.02

830.13

320.00

7727

0.03

400.04

930.01

250.11

070.02

850.03

510.63

770.05

660.01

190.02

3728

0.03

140.08

650.00

180.05

060.02

550.08

800.35

080.03

640.24

200.08

6829

0.06

110.00

001

00.04

990.00

320.01

420.00

080.53

790.02

730.30

5530

0.02

490.03

990.01

030.00

680.01

980.01

730.04

160.80

970.01

730.01

2331

00

00

0.70

300.02

110.18

910

0.08

680

320.07

670.07

420.20

320.00

050.21

190.01

430.04

180.06

810.07

120.23

8133

0.23

260.02

610.05

080.01

110.00

880.02

700.14

050.21

950.21

260.07

0934

0.03

430.04

860.01

260.09

830.03

390.06

890.36

870.22

630.09

040.01

7835

0.27

090.23

420.00

700.14

230.09

340.06

880.04

250.02

910.07

180.03

9936

0.25

290.06

380.01

770.11

900.06

510.03

380.30

670.06

170.03

310.04

6137

0.23

770.00

440.00

490.12

740

0.00

060.07

380.37

750.16

850.00

5138

0.30

590.00

850.01

040.04

920.14

150.09

140.18

990.03

500.04

960.11

8339

0.00

990.00

540

0.13

480.00

020.75

900.07

380.00

120.01

550

Total

0.13

530.06

840.02

570.08

290.05

370.06

060.24

210.13

260.11

330.08

54

Not

es:EachentryshowstheFDIinflowsinaspecificindustryandstateasashareoftotalinflowsinthesameindustryfor

1974–1994.

294

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


As Coughlin et al. (1991) and Friedman et al. (1992) havefound, the condition of the local labor market influences FDIinflows. Specifically, high state specific wages in manufacturing,a proxy for local labor cost, deter FDI inflows while stateunemployment, an indication of the availability of labor, seemsto attract FDI. However, Bernard and Jensen (1999) showedthat labor markets are not well integrated across regions andindustries. Consequently, and to avoid any aggregation bias,we follow Goldberg and Tracy (1999) in constructing a moreaccurate proxy for the tightness of the industry and state specificlabor market. State and industry employment (Lijt) is regressedon a linear time trend preserving the residuals (LRES ijt). Thesame is done for employment in US manufacturing (with theresiduals being (LRESM t). Then the relative tightness in theindustry and state specific labor market is defined as LDijt =LRES ijt − LRESEM t, with an increased (LDijt) reflecting arelatively tight local labor market.

On the other hand, real wages are not a good proxy forlabor cost since high wages do not necessarily imply high unitlabor cost. In fact, and under certain conditions, high wagesare the outcome of high labor productivity. We measure laborproductivity (APLijt) as the industry and state specific averageproduct of labor, given by APLijt = Qijt

EMP ijt, with (Qijt) and

(EMP ijt) being, respectively, the industry and state specificreal gross product and employment. Certainly, a more accurateproxy for labor output, instead of (Qijt), would be the industryand state specific value added, data that are available formanufacturing industries in the entire United States but not inindividual states. Consequently, we proxy the industry and statespecific value added (V Aijt) by multiplying the share of valueadded out of total production in each 2-digit SIC manufacturingindustry in the United States and (Qijt). Then we calculate

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


an alternative proxy for average labor product as APLijt =V Aijt

EMP ijt.12 In the estimations to follow we use both proxies for

labor productivity. Finally, in our estimations we also includea proxy for unit labor cost given by COST ijt = Wijt

APLijtand

calculated using both proxies for labor productivity.Several researchers (Woodward, 1992; Barrel and Pain, 1999;

Hubert and Pain, 2002; and Head et al., 1995, 1999; amongothers) have shown that agglomeration effects are very strongin FDI location decisions. We also control for these effects usingthe share of industry and state specific employment out of thesame industry’s employment in all the ten states in the sample(AGGLijt).13

Finally, in estimations we control for a series of state specificfactors that might be important in influencing inward FDI. Inparticular, the state specific per capita real disposable income(DIjt) measures the level of economic prosperity of a givenstate and proxies the size of the local market. It is expectedthat, ceteris paribus, more developed states attract FDI inflowseither because market demand is stronger in these states or sincelabor productivity is usually high.14 Also, the real per capitastate spending on higher education (EDU jt), and on buildingand maintaining the state’s highways (HWAY jt) and airports(AIRjt), are expected to attract more FDI inflows. On the otherhand, the per capita state taxes (TAX jt) and the state crimerate (CRM jt), that proxy social instability and poverty, usuallydeter FDI inflows.

12Notice that the two proxies for labor productivity are strongly correlated(correlation of 0.730).

13In estimations we also used the share of industry and state specific realgross product out of the same industry’s real gross product in all ten states in thesample, as an alternative proxy for agglomeration effects. The empirical results donot change much, and therefore we do not report them in the article.

14The correlation between the state per capita real income and labor produc-tivity in our sample is 0.549.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


The Choice of the Host State

In evaluating the relative importance of the various economicfactors that attract FDI inflows in United States states, weestimate reduced form Eqs. (1) and (2), with the estimationresults reported in Table IV asModel 1 andModel 2, respectively.

(1) RFDI ijt = f(Wijt,APLijt,AGGLijt,LD ijt,DI jt,

EDU jt,AIRjt,HWAY jt,TAX jt,CRM jt)

(2) RFDI ijt = f(COST ijt,AGGLijt,LD ijt,DI jt,EDU jt,

AIRjt,HWAY jt,TAX jt,CRM jt)

Since our data is in a panel form, we use the Fixed andRandom Effects models in estimating our reduced form equa-tions. Then we test for the appropriate model and we report onlythose results. Equations (1) and (2) are estimated employing firstthe proxy for labor productivity that depends on industry andstate specific real gross product (Specification 1) and also the onethat depends on industry and state specific value added (Spec-ification 2). Finally, all regressors are in logs and therefore theestimated coefficients represent elasticities (Table IV).

As expected, real wages and unit labor cost appear to depressthe share of inward FDI flows allocated to a specific state, whilehigher labor productivity boosts this share. Also, in Specifica-tion 1 these effects are statistically significant. Approximately,a 10% increase in the industry and state specific real wagesand unit labor cost depress the share of FDI allocated in thesame state by 2.2% and 1.2%, respectively. On the other hand, a10% increase in the industry and state specific labor productivitycauses a 1.3% increase in the state’s share of inward FDI.

Industry agglomeration in specific states appears to have rel-atively strong, statistically significant and robust across various

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Table IVAttracting Factors for the Share of FDI Inflows

Specification 1 Specification 2IndependentVariables Model 1 Model 2 Model 1 Model 2

Wijt −0.225∗ −0.0728(−2.662) (−1.012)

APLijt 0.126∗ −0.0214(2.771) (−0.567)

COST ijt −0.122∗ 0.0099(−2.679) (0.269)

AGGLijt 0.449∗ 0.395∗ 0.411∗ 0.363∗(2.956) (2.692) (2.729) (2.486)

LDijt −0.0393∗ −0.0381∗ −0.0284∗ −0.0285∗(−3.135) (−3.093) (−2.342) (−2.377)

DI jt 0.0712 0.0768 0.110 0.107(0.738) (0.799) (1.126) (1.096)

EDU jt 0.297∗ 0.302∗ 0.299∗ 0.307∗(3.222) (3.306) (3.273) (3.388)

AIRjt −0.0015 0.0003 −0.0013 0.0004(−0.159) (0.038) (−0.141) (0.0533)

HWAY jt 0.0439 0.0432 0.0613 0.0602(0.907) (0.903) (1.290) (1.282)

TAX jt 0.0286 0.0271 0.0397 0.0370(0.706) (0.671) (0.975) (0.914)

CRM jt −0.162∗ −0.152∗ −0.133∗ −0.126∗(−2.820) (−2.662) (−2.330) (−2.216)

R2 0.048 0.048 0.041 0.040

Sample 907 907 907 907

Notes: The dependent variable is the share of industry and state specific FDI inflowsout of total FDI inflows in the respective industry and for all the states in thesample. T -statistics are reported in parentheses. A (∗) or (∗∗) next to a reportedcoefficient indicates its significance at 0.01 and 0.05 levels.

specifications, effects on the share of FDI inflows in a specificstate. A 10% increase in industry agglomeration in a given stateboosts the share of FDI that flows in the state by approximately4%. On the other hand, the tightness of the state and industryspecific labor market deters FDI inflows and these results arealso statistically significant and robust across all specifications.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Also, a 10% increase in a state’s per capita spending on educa-tion expands the state’s share of FDI inflows by approximately3%, while a 10% increase in a state’s crime rate depresses thestate’s share of the US inward FDI by 1.5%. Finally, the restof the independent variables do not seem to have a statisticallysignificant impact on the state’s share in the US FDI inflows.

State Competition in Attracting FDI Inflows

As mentioned already, United States states offer significantincentives in attracting FDI inflows, expecting to draw economicbenefits from the operation of foreign subsidiaries in their terri-tory. At the same time, they compete with other states in termsof the incentives they offer to foreign companies so to be chosenas the location of foreign subsidiaries.15 However, recent debateon the topic focuses on the magnitude of the offered economicincentives relative to the expected benefits from attracting FDIinflows.16 At the same time, Head et al. (1999) found that stateeconomic incentives to Japanese companies have small effectson the geographic distribution of Japanese FDI inflows in theUnited States, since these incentives tend to offset each otheracross states. Similarly, Hubert and Pain (2002) found mixed ef-fects of the fiscal incentives offered by various European Unioncountry-members on the FDI inflows they receive. In the presentsection we make an effort to shed some light on the relative im-portance of several economic factors in attracting FDI inflows ina given state with respect to other competing states.

15See, for example, “Sates: Over There Over There,” Financial World(4/17/1990), or “U.S. Governor’s Travel Abroad ‘Opens the Door’ to IncreasedExports and Jobs for their States,” Business America (5/7/1990), or “Ante up:States’ Bidding War Over Mercedes Plant Made for Costly Chase—Alabama Wonthe Business, But Some Wonder if It Also Gave Away the Farm—Will Image NowImprove?” Wall Street Journal (11/24/1993).

16“Alabama’s Winning of Mercedes Plant Will Be Costly, with Major TaxBreaks,” Wall Street Journal, 9/30/1993.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Our data set records FDI inflows in ten US states that werethe top recipients of inward FDI for the period between 1974–1994. Therefore, it is expected that these same states usuallycompete in attracting FDI inflows. Consequently, the industryand state specific FDI inflows as a share of total inward FDI inthe same industry and in the remaining nine states in the sample(SFDI ijt) reflects the effectiveness of this state in attracting FDIinflows with respect to its competing states. Also, the relativestrength of various economic conditions in a state with respectto the conditions in competing states most likely determinesthe attractiveness of the state as a site for the location offoreign subsidiaries. (SWijt) is the relative industry and statespecific real wage rate with respect to a weighted average of thesame industry’s real wages in the remaining nine states in thesample.17 Similarly, all other independent variables are the ratiobetween the specific variable in a given state and the average ofthe same variable in the remaining nine states in the sample.18

Reduced from Eqs. (3) and (4) are estimated with the resultsreported in Table V.

(3) SFDI ijt = f(SW ijt,SAPLijt,SAGGLijt,SLD ijt,

SEDU jt,SAIRjt,SHWAY jt,

STAX jt,SCRM jt)

(4) SFDI ijt = f(SCOST ijt,SAGGLijt,SLD ijt,SEDU jt,

SAIRjt,SHWAY jt,STAX jt,SCRM jt)

17The weights are the relative share of real state gross product in each industryout of the real gross product in the same industry and year in the remaining ninestates together.

18For instance, (SAPLijt) is the share of (APLijt) in industry (i) and state(j) for year (t) out of the average labor productivity in the same industry and yearin the remaining nine states in the sample.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Table VAttracting Factors for FDI Inflows in a State Relative to the

Rest of the States in the Sample


SW ijt 0.198 0.239(0.832) (1.012)

SAPLijt 0.499∗ 0.406∗(2.138) (2.158)

SCOST ijt 0.0584 0.0008(1.475) (0.357)

SAGGLijt 0.109 0.201 0.0896 0.211(0.719) (1.445) (0.630) (1.519)

SLDijt −0.0043 −0.0019 −0.0005 −0.0016(−0.386) (−0.213) (−0.075) (−0.179)

SEDU jt 1.942∗ 2.019∗ 2.022∗ 2.033∗(2.847) (2.968) (2.989) (2.986)

SAIRjt 0.0032 0.0019 0.0040 0.0030(0.396) (0.242) (0.498) (0.371)

SHWAY jt 0.224 0.253 0.247 0.256(0.758) (0.857) (0.840) (0.868)

STAX jt 0.0499 0.212 0.0625 0.236(0.164) (0.718) (0.206) (0.799)

SCRM jt −0.994∗∗ −1.131∗ −0.975∗∗ −1.076∗(−1.855) (−2.155) (−1.829) (−2.052)

R2 0.038 0.034 0.038 0.032

Sample 882 882 882 882

Notes: The dependent variable is the share of industry and state specific FDI inflowsout of total FDI inflows in the respective industry and for all the states in thesample. T -statistics are reported in parentheses. A (∗) or (∗∗) next to a reportedcoefficient indicates its significance at 0.01 and 0.05 levels.

Relative labor productivity appears to be statistically signif-icant in increasing the state’s share of FDI inflows in a givenindustry with respect to its competing states. A 10% improve-ment in the state’s relative labor productivity in a given indus-try boosts the state’s share of FDI inflows in the same industrywith respect to its states-competitors for similar FDI by approx-

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


imately 5%. On the other hand, neither relative real wages norrelative unit labor cost appear to be statistically important inattracting relative FDI inflows.

However, the state’s relative real per capita spending on edu-cation seems to have a strong, robust and statistically significanteffect in attracting FDI inflows with respect to competing states.A 1% increase in (SEDU jt) boosts the state’s share out of FDIinflows in competing states by approximately 2%. Finally, thestate’s relative crime rate seems to depress its share of inwardFDI.19 A 1% increase in (SCRM jt) diminishes the state’s shareof inward FDI by approximately 1%.

Overall, it appears that relative labor productivity, relativespending on education and the relative crime rate are the sta-tistically important factors that help a given state in attractingmore FDI inflows with respect to its competing states. Appar-ently, productivity issues and safety are very important concernsin foreign companies’ decision to choose the location of theirforeign subsidiaries. In fact, John Manley, Canada’s Minister ofIndustry, cited the low productivity in the country as one of thefactors that deters foreign companies in choosing Canada as thelocation of their subsidiaries.20

Some might claim correctly that the choice of the specificUS state in locating a foreign subsidiary usually comes downto couple of state-finalists. In evaluating the factors that mightdetermine finally which state receives an FDI inflow, we constructa data set based on the two states in our sample that receivedthe most FDI in each manufacturing industry during the timeperiod our sample spans (1974–1994). For instance, as Table IIIindicates, in Food (SIC 20) California and Illinois were the top

19This finding is consistent with Woodward (1992).20“The Americas: How Canada Scares Away Investors and Talent,”Wall Street

Journal, 1/2/1998.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


two states in the sample as recipients of FDI inflows. Basedon this notion, we identify the top two FDI state-recipients ineach 2-digit SIC industry in manufacturing and we evaluate therelative importance of the economic conditions in the two statesin explaining the relative FDI inflows in the same states.

Specifically, we estimate again reduced form Eqs. (3) and (4)with (SFDI ijt) being the ratio between the FDI that flows in thetop state recipient and the one that flows in the second highestrecipient. Also, the independent variables are the ratios betweenthe corresponding variable in the top FDI state-recipient andthe one in the second highest state. The results are reported inTable VI.

Relative real wages and productivity have the appropriatesign but are not statistically significant. However, a relativelyhigher industry agglomeration in the top FDI state-recipient hasa statistically significant, robust and expansionary impact on therelative flow of FDI in the state. Finally, a relatively tighter labormarket along with relatively higher per capita taxes in the topFDI state-recipient discourages FDI inflows in the state in favorof its state-competitor.

These results though must be interpreted with caution, sincethe top two FDI state-recipients in each industry are not alwaysamong the final contenders in attracting a foreign subsidiary.Also, the sample size is rather small. Nevertheless, they empha-size some factors, such as relative industry agglomeration andrelative tightness of local labor markets, as pivotal among thestates that usually compete for the same FDI inflows.

V. CONCLUSIONS

The present study focuses on the relative importance ofindustry and state specific economic factors on inward FDI inseveral states in the United States that compete for these inward

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Table VIAttracting Factors for FDI Inflows Among Competing States


SW ijt −5.781 −5.745(−0.592) (−0.587)

SAPLijt 6.604 5.305(1.225) (1.190)

SCOST ijt −1.664 0.139(−1.238) (1.367)

SAGGLijt 19.951∗ 20.340∗ 20.201∗ 20.178∗(2.209) (2.381) (2.242) (2.278)

SLDijt −5.594∗ −5.726∗ −5.664∗ −5.583∗(−2.214) (−2.528) (−2.244) (−2.364)

SEDU jt −7.536 −7.386 −7.143 −3.199(−0.574) (−0.579) (−0.546) (−0.246)

SAIRjt 0.338 0.343 0.346 0.432∗(1.536) (1.636) (1.585) (2.352)

SHWAY jt −7.791 −7.682 −8.223 −9.264∗∗(−1.459) (−1.463) (−1.551) (−1.719)

STAX jt −10.184∗ −10.239∗ −10.029∗ −7.559∗(−2.452) (−2.565) (−2.432) (−2.014)

SCRM jt 2.634 3.625 2.0803 −6.567(0.145) (0.223) (0.115) (−0.436)

R2 0.103 0.121 0.101 0.112

Sample 77 77 77 77

Notes: The dependent variable is the ratio between the share of FDI inflows inthe top FDI state recipient and the share of FDI inflows in the second highestFDI state recipient for each industry. T -statistics are reported in parentheses.A (∗) or (∗∗) next to a reported coefficient indicates its significance at 0.01 and0.05 levels.

FDI flows. It appears that relative labor productivity, relativespending on education, and relative crime rate are importantwhen states compete with other states in attracting inward FDI.Also, when the contest in attracting inward FDI comes down totwo states, then relative taxes also become important in deterringFDI inflows. Thus, multinational companies seem to evaluate

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


highly a productive and educated labor force when they decidethe location of their subsidiary in the United States. However,these results might be driven by a set of technology-intensiveindustries as Dunning (1992) suggested. This will be addressedin future research.

VI. APPENDIX

US States in the Sample: California, Connecticut, Florida,Illinois, Massachusetts, New Jersey, New York, Ohio, Pennsylva-nia, and Texas.

Foreign Direct Investment (FDI): FDI inflows in each2-digit SIC manufacturing industry and in each state of thesample. (Source: ITA Data Set, US Department of Commerce).

Employment: Thousands of persons employed in each 2-digit SIC manufacturing industry and in each state of the sample.(Source: US Department of Labor, Bureau of Labor Statistics).

Wages: Average hourly earnings in each 2-digit SIC man-ufacturing industry and in each state of the sample (in currentdollars per hour). (Source: US Department of Labor, Bureau ofLabor Statistics).

Shipments: Total nominal value of shipments in each 2-digitSIC manufacturing industry. (Source: Eric J. Bartelsman, RandyA. Becker, and Wayne B. Gray: NBER-CES ManufacturingIndustry Database).

Industry Value-Added: Nominal value-added in each 2-digit SIC manufacturing industry. (Source: Eric J. Bartelsman,Randy A. Becker, and Wayne B. Gray: NBER-CES Manufactur-ing Industry Database).

Total Capital Expenditure: Nominal expenditure on fixedassets in each 2-digit SIC manufacturing industry. (Source:Eric J. Bartelsman, Randy A. Becker, and Wayne B. Gray:NBER-CES Manufacturing Industry Database).

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Investment and Shipments Deflators: Appropriate de-flators for fixed investment and shipments. (Source: Eric J. Bar-telsman, Randy A. Becker, and Wayne B. Gray: NBER-CESManufacturing Industry Database).

State Specific Data: The data include state disposableincome; state spending on highways, airports, and education;state tax revenue; state civilian resident population (source:US Department of Commerce); and the state total crime rate(Source: Federal Bureau of Investigation).

REFERENCES

Axarloglou, K. 2004. Local Labor Market Conditions and ForeignDirect Investment Flows in the U.S. Atlantic EconomicJournal 32(1):62–66.

Barrell, R., and Pain, N. 1999. Domestic Institutions, Agglom-erations and Foreign Direct Investment in Europe. EuropeanEconomic Review 43(4–6):925–934.

Bartik, T. J. 1985. Business Location Decisions in the UnitedStates: Estimates of the Effects of Unionization, Taxes, andOther Characteristics of States. Journal of Business andEconomics Statistics 3:14–22.

Bartik, T. J. 1989. Small Business Start-Ups in the United States:Estimates of the Effects of the Characteristics of States.Southern Economic Journal 55(4):1004–1018.

Bernard, A. B., and Jensen, J. B. 1999. Understanding Increasingand Decreasing Wage Inequality. In: R. Feenstra, ed., Tradeand Wages. Chicago: University of Chicago Press.

Billington, N. 1999. The Location of Foreign Direct Investment:An Empirical Analysis. Applied Economics 31:65–76.

Braunerhjelm, P., and Svensson, R. 1996. Host Country Char-acteristics and Agglomeration in Foreign Direct Investment.Applied Economics 28(7):833–821.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Coughlin, C. C., Terza, J. V., and Arromdee, V. 1990. StateGovernment Effects on the Location of Foreign Direct In-vestment. Regional Science Perspectives 20:194–207.

Coughlin, C. C., Terza, J. V., and Arromdee, V. 1991. StatesCharacteristics and the Location of Foreign Direct Invest-ment Within the United States. Review of Economics andStatistics 73:675–683.

Culem, C. G. 1988. The Locational Determinants of Direct In-vestments Among Industrialized Countries. European Eco-nomic Review 32:885–904.

Dunning, J. H. 1992. Multinational Enterprises and the GlobalEconomy, Addison-Wesley.

Friedman, J., Gerlowski, D., and Silberman, J. 1992. WhatAttracts Foreign Multinational Corporations? Evidence fromBranch Plant Location in the United States. Journal ofRegional Science 32(4):403–418.

Friedman, J., Fung, H.-G., Gerlowski, D. A., and Silberman, J.1996. A Note on “State Characteristics and the Location ofForeign Direct Investment Within the United States. Reviewof Economics and Statistics 78(2):367–368.

Goldbreg, L., and Tracy, J. 1999. Exchange Rates and LocalLabor Markets. In: R. Feenstra, ed., Trade and Wages.Chicago: University of Chicago Press.

Hatzius, J. 2000. Foreign Direct Investment and Factor DemandElasticities. European Economic Review 44:117–143.

Head, C. K., Ries, J. C., and Swenson, D. L. 1995. Agglomera-tion Benefits and Location Choice: Evidence from JapaneseManufacturing Investments in the United States. Journal ofInternational Economics 38:223–247.

Head, C. K, Ries, J. C., and Swenson, D. L. 1999. Attract-ing Foreign Manufacturing: Investment Promotion and Ag-glomeration. Regional Science and Urban Economics 29:197–218.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014


Hubert, F., and Pain, N. 2002. Fiscal Incentives, EuropeanIntegration and the Location of Foreign Direct Investment.NIESR, Manchester School, 70(3):336–363.

Luger, M. I., and Shetty, S. 1985. Determinants of Foreign PlantStart-Ups in the United States: Lessons for Policymakers inthe Southeast. Vanderbilt Journal of Transportational Law18:223–245.

Wheat, L. 1986. The Determinants of 1963–77 Regional Manu-facturing Growth: Why the South and West Grow. Journalof Regional Science 635–660.

Wheeler, D., and Mody, A. 1992. International Investment Lo-cation Decisions: The Case of US Firms. Journal of Interna-tional Economics 33:57–76.

Woodward, D. P. 1992. Locational Determinants of JapaneseManufacturing Start-Ups in the United States. SouthernEconomic Journal 58(3):690–708.

Dow

nloa

ded

by [

Nor

thea

ster

n U

nive

rsity

] at

05:

34 1

3 N

ovem

ber

2014

Documents

What Attracts Foreign Direct Investment Inflows in the United States