FDI & Ports

This study investigates the effect of the presence of a port on the attractiveness of inward FDI in the region. Port regions will be defined due its geographical location

and categorized by size. Afterwards, using a negative binomial regression model,

several related hypotheses will be tested. The results show a positive significant

relationship between a port region and the inward FDI.

The Effect of a Port on the inward FDI in a region

2

Table of Content

1. INTRODUCTION.........................................................................................................................3

2. THEORETICAL BACKGROUND ...............................................................................................5 2.1 Port regions and FDI ......................................................................................................................... 5 2.2 Positive effects of port on the city ...................................................................................................... 7 2.3 Business sectors and proximity to port .............................................................................................. 8

3. THE HYPOTHESES .................................................................................................................. 10

4. DATA & METHODOLOGY ...................................................................................................... 14 4.1 Data .................................................................................................................................................. 14 4.2 Ports and urban attractiveness ........................................................................................................ 16 4.3 Sectoral Analysis .............................................................................................................................. 16 4.4 Effect of Port size ............................................................................................................................. 17 4.5 Control Variables ............................................................................................................................. 18 4.6 Estimation Strategy .......................................................................................................................... 20

5. RESULTS ................................................................................................................................... 21

6. DISCUSSION & CONCLUSION ................................................................................................ 26

REFERENCES ............................................................................................................................... 30

APPENDIX ..................................................................................................................................... 32

3

1. INTRODUCTION

In the ceaseless game of dominance, multinational companies (MNCs) constantly pursue

competitive gains and advantages in order to secure sustainability. In a fast changing and

demanding global community, firms aim to expand beyond the boarder of their own country,

investing in different geographical areas around the globe, a strategy commonly refer to as

foreign direct investment (FDI). As simple as this may sound, the process of identifying

opportunities abroad is complex and usually involve businesses to analyze different factors

before committing to a specific location (Dunning J. H., 2001) (Porter, 2000). For businesses

interested in investing in European regions, ports could be of pivotal importance. Ports serve as

vital economic gateways as 74% of goods imported in or exported from Europe go by sea.1

European ports are not considered a homogenous set of ports (Notteboom, 2010) not only

because of the varied types of commodity handled but also for reasons related to connectivity

with various hinterlands and different location qualities. With many different ports within a

relatively small continent (Europe), competition to draw FDI becomes fierce among regions.

Another interesting aspect of the European region has to do with the formation of the European

Union. One of main purposes for forming the Union was to facilitate investments and potentially

increase wealth across the Union. The free movement of capital, goods and persons within the

Union as well as no trade barriers and tariffs within the E.U in theory make FDI very appealing

to those firms looking for opportunities abroad. In their study, Bevan & Estrin analyze the

determinants of FDI in European economies and found that “FDI is positively related to both

1 https://ec.europa.eu/transport/modes/maritime/ports/ports_en

4

source and host country GDP and related inversely to the distance between countries and to unit

labor costs” (Bevan & Estrin, 2004). An example of this could be a German firm looking for

opportunities to reduce labor costs in the Eastern part of Europe2. From previous research, we

know that many parts of Europe draw substantial amount of FDI in both services and

manufacturing sectors, the latter sector being explored more often especially in Eastern regions

(Disdier & Mayer, 2004). Furthermore, firms interested in investing in the European regions in

general do not seem to discriminate on the determinants of FDI before committing to a specific

location. This means that market size and agglomeration effect are considered equally important

for all EU regions (Disdier & Mayer, 2004). Disdier and Mayer also mentions that the

competition for FDI is not across regions (West vs East for example) but within the regions itself

(East vs East countries and West vs West countries). An explanation for this might be the

economic characteristics of the region drawing similar FDI interest within these areas.

The economic attractiveness as well as the geographical composition of European regions along

with the presence of many ports within this specific area makes it very interesting for economist

to research. Our research question is therefore if the presence of a port plays an active role in

urban competiveness and the attractiveness of inward FDI? The method used to answer this

question is the so-called “negative binomial model”; a model which is very suiting in order to

control for overdispersion in our data.

This paper continues as follows: In Section 2 we provide the reader with the necessary theory to

understand the dynamics of ports and FDI within the European region as well as our motives for

formalizing the different hypotheses within this paper. Section 3 introduces the composition of

2Data suggests that income in Germany is higher than in for example Bulgaria creating possibly certain advantages.

See:http://databank.worldbank.org/data/Views/Reports/ReportWidgetCustom.aspx?Report_Name=CountryProfile&

Id=b450fd57&tbar=y&dd=y&inf=n&zm=n&country=DEU and

http://databank.worldbank.org/data/Views/Reports/ReportWidgetCustom.aspx?Report_Name=CountryProfile&Id=b

450fd57&tbar=y&dd=y&inf=n&zm=n&country=BGR

5

our dataset, reasons for our choice of regression models and estimation strategy. Section 4 is

reserved for the empirical analysis and results of the main and sub hypotheses. Finally, in

Section 5 we discuss some of the limitations of this paper and conclude accordingly.

2. THEORETICAL BACKGROUND

2.1 Port regions and FDI

There are many definitions of what a port is but for simplicity we will use the definition of

Stopford (2009:81) a port is “a geographical area where ships are brought alongside land to load

and discharge cargo – usually a sheltered deep-water area such as a bay or river mouth”

(Stopford, 2009) (Nijdam & van der Horst, 2018). From this we can already get a “feel” that

ports function as node in transport chains and are important for economic activities involving

cargo and ship handling (Nijdam & van der Horst, 2018). Ports generally are an economic

catalyst for surrounding cities in the region, facilitating the integration of markets and the

agglomeration of services that generate economic benefits and socioeconomic welfare (Song &

van Geenhuizen, 2014) (Zhao, Xu, Wall, & Stavropoulos, 2017). The following paths in spatial

distribution show changes in the economic relationship between ports and port cities. Port cities

usually benefit from the port’s economic activities, by needing to be near the port. An example

of this is the lower transaction costs provided by ports to port related business activities.

Urban spaces near ports also provide ports with advantages that cannot be easily accessed

outside of urban agglomerations, such as labor pools and infrastructure, in this case the city

provides the port the human capital to efficiently run its labor and provide for roads to reach the

hinterland and vice versa (Hall & Jacobs, 2012). Port-related industries are attracted by such

environments, which allows ports and port cities have a relationship with one another and

economically benefit off each other (Zhao, Xu, Wall, & Stavropoulos, 2017). Because the

6

maritime transport network covers about 90% of world trade (Ducruet, Rozenblat, & Zaidi,

2010), it also has an effect on the global economy. Port and city networks become related to each

other during this process. Jacobs et al. (2010) stated that port-related advanced producer services

activities predominantly follow the overall global city trends, where some port cities have an

advantage over others because they act as hubs in global (commodity) flows on top of acting as

centers of advanced services related to shipping and port activities. But why is one urban port

area more competitive and/or attractive than another? First we need to understand the concept of

urban competitiveness. Kostiainen (2002) states that ‘the ability to attract flows of information,

technology, capital, culture, people, and organization is the key concept of urban

competitiveness (Kostiainen, 2002)’. Urban competitiveness can be measured by the capacity of

cities to attract investment and to promote development (Sáez & Periáñez, 2015). It is important

to note that there are two different types of FDI’s related to ports; inward FDI and outward FDI,

this is in our interest to analyze because inward and outward investment should be separated

from each other due to the different requirements of the involved parties making the investments

(Dooms, Lugt, & Langen, 2013) (Kolstad & Wiig, 2012).

Inward investments give the opportunity to attract international private multinationals, where the

host country can grow investments and provide for local economic growth. Outward investments

aim for outward activities (abroad), and look for exploiting new business opportunities and

relationships (Dooms, Lugt, & Langen, 2013). The port’s (authority) outward

internationalization strategies are to sell the port worldwide (customer seeking principle), create

value for their domestic customers and/or maximize profits through involvement in exploiting

market opportunity abroad (Dooms, Lugt, & Langen, 2013). However, because this study

focuses on the capacity to attract foreign investments in Europe, we solely analyze inward FDI

7

made within the mentioned region. The ability to attract (inward) FDI is a good indication of a

city's competitive success (Lovering, 2003). Inward investments allow flow of goods, capital,

resources, information and or services from external world to domestic market (host

country/region/city/port) (Karlsen, Silseth, Benito, & Welch, 2003). The port inward FDI

operations aim to attract more foreign direct investment, attract international private companies,

increase investments and traffic volumes and this all to aid the local economic growth.

2.2 Positive effects of port on the city

With the presence of a port in/near a city, the port has an influence on the city to focus on export

related industries. The transport costs are affected by the connectivity between inland countries

(region) and available ports. For example, in comparison with other inland countries, the Czech

Republic, Switzerland and Austria that are surrounded by ports in the European port system, and

this advantage affords these countries more negotiation power to decrease transport costs in

business opportunities (Merk & Hesse, The Competitiveness of Global Port-Cities: The Case of

Hamburg, Germany , 2012). Secondly, another benefit of ports on urban regions is the creation

of an additional added-value. For example, Rotterdam generated 12.8 billion U.S dollars of

added-value in 2007, which accounted for ten percent of its regional GDP (Merk, 2014) This

value comes from four sources (Ferrari, Parola, & Gattorna, Measuring the quality of port

hinterland accessibility: The Ligurian case, 2011):

 The increase of employment opportunities and income by the construction and operation

of port infrastructures

 The increase of employment opportunities and income by port related industries

 A stimulated domestic demand by the increase of income

 Foreign investment that is attracted by a port’s welfare

8

Merk (2014) also states that as the scale of a port becomes larger, the added value also grows e.g

up to ten percent in employment in the ports of North West Europe. Thirdly, the growth of (port)

employment: a study of the European port region shows that 100 million units of cargo

throughput will create 0.0003% of regional employment opportunities (Ferrari, O., Bottaso, &

Tei, 2012). To add to this, a port city is becoming an innovation center for port related industries

(2014). Fourth, the spillovers (technology, salary) of economic benefits will spread to other cities

in the vicinity of the port (region). A port can do this by increasing its port competitiveness in

terms of connectivity and port efficiency; the purpose of this is to improve its locational status in

global port networks and thereby increase its urban competitiveness, thus building a bridge

between ports and municipalities (Zhao, Xu, Wall, & Stavropoulos, 2017). In the case study of

Rotterdam, the positive spillovers of the Rotterdam port have even spread to nearby countries

such as German industries (Merk, 2014). Based on this research of Merk (2014) there are three

aspects of port competitiveness. The first one is maritime connectivity, the aspect to measure the

level of accessibility of a certain port. The second aspect is port efficiency, throughput of

container and bulk goods, as well as ship calling and other indicators of port activeness. The

third aspect is hinterland connectivity, how well a port is connected and reachable for the region

around the port.

In order to satisfy these demands of multinational corporations, the destination of foreign

investment should possess as many as attracting factors related to locational advantages. To

emphasize, because we focus on the inward investment, we should focus on the determinants of

investing in the host country.

2.3 Business sectors and proximity to port

9

In the previous section we discussed the attracting factors of port regions and cities but the

question remains; which business sectors are attracted the most by the presence of a port in an

area? Some business sectors (manufacturing, logistics, transport for instance) are attracted by

being near a port. Being close to a port provide these firms with advantages concerning

transaction and transportation cost (Jacobs, Ducruet, & De Langen, Integrating world cities into

production networks: The case of port cities, 2010). In recent times, ports have slowly moved

away from city centers (Jacobs, Ducruet, & De Langen, Integrating world cities into production

networks: The case of port cities, 2010). There are many reasons explaining this phenomena;

environmental awareness and the right for cleaner air (public good), lack of land to expand port

related business activities near city centers (Jacobs, Ducruet, & De Langen, Integrating world

cities into production networks: The case of port cities, 2010) (Hoyle, 1989) and governments

are more involved in spatial planning and designing creating specific places for these industries

to conduct their businesses. Furthermore, Jacobs (2007) argues that because port related

activities went from a more public (governmental) domain of business to private

(corporate/investors) business negatively impacting the port-city relationship because “the

dependence of ports on the urban labor market as well as the reduced dependence of cities on

ports for local economic growth” (Jacobs, Political Economy of Port Competition , 2007).

On the other side of production related businesses, we have services related activities. These

activities are mostly ICT, insurance and consultancy related businesses (see Jacobs 2007 et al.

table 1 for complete overview). It is tough to pinpoint that these sectors exist near ports regions

specifically to enhance/provide port business with for example consultancy or legal support. In

other words, it is near impossible to categorize financial service related businesses in port

10

regions that are there solely there because of the port simply because there isn’t extensive data

available to proof this.

3. THE HYPOTHESES

FDI is positively related to economic growth, more often than not regardless of a host country’s

human capital level. Furthermore, these investments can create jobs for in a region and increases

the competitiveness in the host country (Wang & Wong, 2009). As firms looking to make a

Foreign Direct Investment are assumed to be profit maximizers, they select an investment based

on the chosen region’s characteristics impacting profits relative to other regions (Makabenta,

2002). Some of the determinants of FDI inflows, based on existing theories are the market size,

the advanced infrastructure of big cities and ports, natural resources, the proximity to European

markets and finally, legislative and political risks (Ledyaeva, 2009). Accessibility to ports and

connected cities by rail and road are positively related to FDI locations especially for firms

seeking to locate in areas near ports to reduce transportation costs (Makabenta, 2002).

Makabenta (2002) shows in her research in the Philippines that the port and highway variables

have strong pull effects on manufacturing FDI, as potential investment areas do not only need

pools of skilled labour, but also need good transport options to markets and/or sources of

resources via ports and roads. The availability of a port, the marginal effect of the dummy

variable PORT, in the study increased the new FDIs by x1.62. This hints to the conclusion that

further improvement of the port (region) increases the attractiveness of foreign direct

investments. Similar results have been concluded in Nyamai & Wall’s (2015) research which

looked at competitiveness between port and non-port cities. The results showed that for the port

cities higher education and liner shipping were positive and significant indicating that smart

11

people and global shipping networks are needed to attract more FDI. Whereas for non-port cities

the employment rate was positive and significant indicating that an increase in employment

would increase FDI in the city. However, port cities remain the most competitive compared to

non-port ones because of the ability to attract FDI due to a growth in the number of smart human

capital (Nyamai & Wall, 2015).

Hypothesis 1: If a region includes a port, it is expected to attract more FDI compared to a

non-port region.

The growth of the port of Shanghai is related to the development of its economy, manufacturing

and foreign trade, this because Shanghai is a major exported of manufactured products.

Therefore, it is needed for this market to develop a major international port in/near Shanghai.

The port serves as a city-serving hub port-city, the port basically serves as an international

shipping center for the region for the mass-produced goods that need to be exported, as well as

the resources that need to be imported (Huang, 2009). To deal with these trades, a trade- and

financial service center will be set up in the city. This is supported by Makabenta (2002) as

shown in the results where manufacturing FDI firms looks for regions that have, among other

aspects, access to ports and highways as to which firms have to ability to efficiently transport

their manufactured products. These two effects also have the largest marginal effects on the

attraction of FDI. Furthermore, also European ports such as Rotterdam and Antwerp have

developed their ports to big facilities of production and manufacturing because they are

dependent on the import of raw materials and want to be close to the port (Jacobs, Ducruet, & De

Langen, Integrating world cities into production networks: The case of port cities, 2010).

Therefore, we would expect businesses within the manufacturing and “wholesale, warehousing,

logistics and transport” businesses to be drawn by the attractiveness of short distances or direct

12

accessibility of nearby ports due to the fact that the main goal of ports is to load and unload the

goods as quick and efficient as possible (Jacobs, Ducruet, & De Langen, Integrating world cities

into production networks: The case of port cities, 2010). Business sectors that might be drawn to

the port but need not to be near the port are those that are concerned with the service aspect of

the business.

Hypothesis 2: The effect of the port in attracting FDI is stronger for investments in the

Manufacturing sectors relative to the Services sectors.

Ledyaeva (2009) suggests that in general competition for FDI between regions with ports has

increased after the crisis. The negative spatial relationship in FDI within the group of port-

endowed regions indicates that if one region with a port can offer additional advantages in other

FDI determinants than what is offered by neighboring port regions, foreign investors will tend to

choose that region. Components of port competitiveness such as; cost‐related vessel and cargo

entering, efficient inland transport network, frequency of large container ships' calling,

inland transportation cost, port accessibility/congestion/safety, professionals and skilled

labor in port operations and reliability of schedules in port are factors than can either

increase or decrease a port’s attractiveness (Yeo, Roe, & Dinwoodie, 2011)

A prime example of such port competitiveness are the ports of Shanghai and Ninbo. With the

ever-growing interest in FDI in China, the Yangtze River Delta is expanding exponentially. In

addition, two of the fastest growing container lines in the world, Cosco Container Lines and

China Shipping Container Lines, not only have their headquarters in Shanghai but also use

Shanghai as their main hub port in China. Competitive wise this is obviously better for the port

in Shanghai than Ningbo port. This is especially the case since Shanghai’s throughput is largely

domestic cargo, with international import and export still playing a major role within the port.

13

The potential for a further significant expansion in demand for the port of Shanghai is therefore

obvious (Cullinane, Teng, & Wang, 2006). However, it should be mentioned that in the future, as

the development of smaller ports such as Ningbo increases and with the ever-growing interest for

FDI in China, the port of Ningbo will benefit the greatest marginal benefit from the economies of

scale and efficiency improvement (Cullinane, Teng, & Wang, 2006). Especially because of the

ever-growing costs when dealing with larger ports, Hong Kong’s port charges for example are

the highest in the world and are at least 63 percent more expensive than other Asian ports (Yeo,

Roe, & Dinwoodie, 2011) In Asia, there appears to be a link between the size of the port and the

attraction of FDI to the port region. We will therefore test if this is the case in Europe as well.

Hypothesis 3: The size of the port is positively related with the number of total investments and

this relationship is stronger for large, main ports.

14

4. DATA & METHODOLOGY

4.1 Data

To examine the level of attractiveness of inward FDI between European regions explained by the

presence of a port, we chose the number of investment projects as the independent variable of

our analysis since it is a widely accepted operationalization of the competitiveness among the

multiple location alternatives in the economic research. Information concerning investment

projects were extracted from the FDI & Market dataset which includes the number of

investments per sector in 237 NUTS-2 regions for the period 2003-2011. To get information

about economic, demographic and structural characteristics on regional level, we merged the FDI

& Market dataset with the Regional Data Europe 2003-2010 dataset (Eurostat). Interestingly,

even though our main focus is on the impact of ports in regional economy, we could not find any

dataset to satisfy the needs of our research for those NUTS-2 regions and ports so we created one

of our own which perfectly includes all information needed for testing our hypotheses. More

specifically, our main motivation was to make a distinction between regions and non-port

regions based on the following concerns: regions with and without ports do not share apparently

same characteristics in terms of socio-economic infrastructure and institutions. They might

reveal differences between their main sources of financial income, the quality of human capital

or the levels of criminality and uncertainty. Not exploiting this heterogeneity on regional level

might lead to biased results and invalid conclusions. Segregating regions in two main

categories enables us to make comparisons between the levels of attractiveness in port and non-

port regions and isolate the effect of the presence of a port. Therefore, we created a dummy

variable which takes the value 1 if the region is recognized as a port region and 0 otherwise. This

variable characterizes 131 NUTS-2 regions and the procedure we followed to assign the proper

15

values was by looking at each region geographical position and considering it as port region if

this is directly connected to the sea and served by at least one port within or close to the borders

of that region.

Our unique FDI dataset

consists of totally 13,441

investment projects in 278

NUTS-2 regions across 33

European countries from 2003

to 2010. Investments,

however, are not normally

distributed as 75% of them

range from 0 to 9, indicating a strong

positive skewness. Almost 60% of the total

investments are in the Services sectors; Business, Financial, ICT, Transport and ConServ.

(Figure 1).

Regarding the business

activities, those can be

Figure 1: Distribution of FDI per Business sector

Figure 2: Distribution of FDI across Europe

16

viewed as upstream (R&D, headquarters) and downstream (Sales & Marketing, Business and

supporting services and Logistics). Most of the investments fall into the Sales & Marketing

category followed by production, indicating a market seeking behavior of the investing firms.

Finally, Figure 2 indicates that half of the total investments are found in Central and Western

Europe, while the UK and Ireland hold together 20% of them.

4.2 Ports and urban attractiveness

Capturing the impact of ports in attracting FDI, we set as variable of interest the dummy variable

‘portregion’. This measurement, on the one hand gives a relevant distinction between the

competitiveness of port and non-port regions as necessary to test the main hypothesis. However,

some regions might be located near the sea but are not be considered a “port region”. Whilst in

other cases, there may be large areas not near the sea that have a large port that are considered as

port regions.

Alternatively, we estimate this effect by using the distance to the nearest seaport. This

measurement overcomes some confusion derived from the different sizes of the regions. At the

same time, distance is a rather general indicator since we do not know exactly the points of

reference taken in the NUTS-2 regions. We expect, in consistence with our main hypothesis, a

positive coefficient for the port dummy and therefore a negative coefficient for the variable

distance to seaport.

4.3 Sectoral Analysis

The utterly different nature and

structure of MNCs in service

and manufacturing industries

Figure 3: Distribution of FDI in Manufacturing industries per Business Activity

17

hint at different required circumstances and surroundings of the market of entry when it comes to

investing. Our interest turns to investigating possible differences between the effect of a port in

service and manufacturing industries. Unlikely, even our dataset provides information on the

number of investments through nine different sectors, the variation of the number of investments

between the sectors was not separately enough and exploitable. Thus, we summarized them in

two groups; ‘Manufacturing’ and ‘Services’. ‘Manufacturing’ consists of the sectors HighTech,

MedTech, LowTech and Proceeding Industry and ‘Services’ respectively, consists for Transport,

ICT, Consumer, Business and Financial Services. Since we do not know exactly if the

investments are related with port activities or not, we can’t have a clear disposition as for the

expected sign. We expect though a stronger effect for Manufacturing for reasons mentioned in

previous sections. A detailed breakdown on business categories, sectors, functions and activities

can be found in the Appendix. As mentioned before, it is expected the effect of the port to be

stronger for investments in Manufacturing. It is crucial here to have a closer look at the business

activities of these MNCs investing in manufacturing. As seen from Figure 3 only one third of

these investments are related to production plants in contrast with upstream and downstream

activities, something useful to keep in mind when it comes to interpret our results.

4.4 Effect of Port size

Testing our third hypothesis requires the classification by size of the ports included in our own

created dataset which ensued from the following procedure: first we matched each of the port

regions with one -the largest and busiest in the area- port. In order to classify them, we utilized

the list of the twenty largest ports in Europe by the volume of TEUs in 2010 by Eurostat. For the

rest of the ports, we searched the numbers of TEUs by looking at relevant articles and annual

reports or websites of the port itself. Twenty-Foot Equivalent Units (TEUs) are counted annually

18

and individually for all ports and functions as a measurement of the largeness or busyness of a

port. Note here that for 11 out of 101 ports chosen, data on TEU volumes were not available.

Thus, we got a general picture of their size by looking at the volume of cargo they handle in

tons-from the same sources and by comparing them with the numbers of the ports. We

subsequently assigned each port properly in our ranking list. We therefore categorized ports in

four groups: Small ports (0-100.000 TEU), Medium ports (100.000-300.000 TEU), Large ports

(300.000-1.000.000 TEU) and Mainports (1.000.000 - more TEU) where ‘Small Port’ is selected

to be the bench category. The distribution of these ports can be found in Table 6 and Table 7 of

the Appendix.

Based on this and our theoretical background about the contribution of ports to the urban

competitiveness, we expect the effect of the presence of a port to the attractiveness of inward

FDI to be positive and stronger for main, large ports that act as hubs in the global networks.

For reasons mentioned in the previous section, we also expect that port size will be also

significant through different sectors, with stronger to be that on manufacturing.

4.5 Control Variables

Following the taxonomy of FDI motives as reported by Serwicka, Jones and Wren (2014) which

was based on Dunning’s OLI framework, we introduce our control variables by linking them to

the main motivations of inward FDI, adjusted to the European environment. Here, diverse

motives are translated as market-, resource-, efficiency- and strategic-asset-seeking FDI.

Market-seeking FDI aims to serve overseas demand and is mostly driven by factors like the size

of the host country and market and its singularity compared to the neighboring ones. We use

gross domestic product (GDP) per capita, as a measurement of the regional economy, widely

used in the investment literature and population in thousands of residents as a proxy for the size

19

of the country. We also include population density to control for unequal distribution of

population among the regions. The percentage of population with higher education following

the ISCEC classification by Eurostat in 2011, is included to capture the demand for high-skilled

employees which is mostly desired in-service industries and business upstream activities.

Resource and efficiency-seeking FDI aim at the best quality of production with the lowest

possible costs. In other words, MNSs tend to maximize their benefits by exploiting the

differences between the costs of acquiring these resources of production in the mother and the

host country. To capture labor costs, we include the average of annual wages per region.

Despite numerous previous studies, we do not include long-term unemployment in our

estimation because its effect is rather ambiguous with both negative and positive impacts; high

unemployment makes recruitment easier and cheaper in theory but it may also induce serious

socio-economic problems like rising crime and skills extinction, creating this way uncertainty

and unattractive surroundings for the investors. Another cost taken under consideration is

corporate taxes as a proxy for capital costs which differs to a great extent between the European

countries. To illustrate with an example from our sample, corporate taxes range from 9%

(Cyprus) to 40% (Germany). While the previously mentioned types of motives exploit actually

existing assets of the regional economy, strategic-asset-seeking FDI aims deeper, to acquire

foreign assets of the economy of entry. R&D is of immense importance for firms when strive to

create competitive advantages over their competitors. It enables companies to predict and -

hopefully- meet future demand and trends and be ahead of fellow companies within their sector.

In this sense, expenses on R&D should be better seen as a kind of investment and not as an

20

expense per se. Based on the definition by OECD on gross domestic spending on R&D3, we

incorporate in our model such expenses to capture the difference in the levels of attractiveness

between the alternative locations for those MNEs which work toward establishing well-founded

and profitable operations in a foreign market.

Numerous empirical studies have verified that MNCs are attracted by the positive externalities of

economic concentrations within both the same and different sectors (Bronzini, 2004).Marshall in

his book ‘Principals of Economic Theory’ (1890) made a distinction between localization and

urbanization economies which are driven by lower input cost, larger labor markets and

knowledge spillovers. It would be absurd to assume that investments occur randomly through

space and time and independently to each other so not controlling for economic concentrations,

might lead to bias estimates and imprecise conclusions. In order to capture the effect of

localization and urbanization we include in our model the share of the population of own

employment in 5 sectors and the percentage of urban land use respectively. In the first case the

firm has to benefit from lower input costs and specialization in each ever field and in the second,

from the amenities of a big area like in transportation and infrastructure regardless the industry.

4.6 Estimation Strategy

MNCs location choices indicate the attractiveness of each region and subsequently its

competency to pull in FDI, able to elevate the regional economic status. The nature of our

dependent variable commands the application of count data models where counts are the number

of events occurred within a fixed time period. In our attempt to specify the probability of an

3 Gross domestic spending on R&D is the total expenditure (current and capital) on R&D carried out by resident

companies, institutes, universities etc., in a country. It includes R&D funded from abroad, but excludes domestic

funds for R&D performed outside the domestic economy. This indicator is measured as a percentage of GDP.

21

investment to occur in a particular region, OLS would yield biased estimates so we turn to a

Poisson process. Poisson however, is violated more often than not since it is based on the

equality of conditional mean and variance. From descriptive statistics, we see that the variance of

our dependent variable is more than six times its conditional mean, a clear sign of

overdispersion. We first calculate a Poisson regression-even though we believe that the Poisson

distribution is not the appropriate one- in order to test the goodness of fit of the model. The chi

square statistic is very significant in this case; thus, Poisson is not the best option. To continue, a

generalized version of Poisson is implemented, the negative binomial model, which has an extra

parameter that controls for extra variation. The likelihood ratio test for over dispersion confirms

once again our selection for the negative binomial model. As discussed, we incorporate along

with the variables of interest, typical controls aiming to inferring about the main determinants of

investing in Europe.

5. RESULTS

The results of our negative binomial regression model are represented in Table 1. The most

important findings are depicted in the two first rows, “portregion” and “distancetoseaport”. To

test the first hypothesis, the number of investments is taken as the independent variable. As seen

from the estimation of the baseline Model 1, the portregion dummy is highly significant and

positive, clearly demonstrating that port regions attract more FDI compared to non-port regions.

The results of the alternative estimation by using ‘distancetoseaport’ in column two, are

consistent with previous findings indicating that if the distance from a region to seaport

increases, the number of investments in this region decreases. Accordingly, we find strong

support for hypothesis 1.

22

Columns 3 to 6 show the estimations of our second hypothesis which concerns the effect of the

port in different business sectors. With the number of investments in both sectors taken as the

independent variable, colums 3 and 4 reveal that the port has a strong effect in attracting

investments of both Manufacturing and Services industries. Contradictory with our expectations,

is the stronger effect for investments in Services. This result is well-based on extant literature

and will be discussed in our conclusions.

The consistent results of the alternative estimation are shown in columns 5 and 6. As both

coefficients are significantly negative, a larger distance to the nearest seaport implies a lower

number of investments for both sectors. This effect is again stronger for Services indicating no

support for our second hypothesis.

23

Tabel 1: Port Regions and Sectoral Dissimilarity

24

Findings regarding the effect of the port size in attracting FDI are represented in Table 2. The

baseline estimation depicted in column 1, offers partial support to our third hypothesis, yielding

an insignificant coefficient for the third size category ‘Large ports’. The significant, positive and

large coefficient in the fourth category ‘Mainports’ indicates that the expected log number of

investments in regions with a major port is 1.819 higher compared with this number in regions

with very small ports. As for ‘Medium Ports’ the result is significant at a 1% level, has the

expected positive sign and signify that the expected log number of investments in regions with a

port of category 2 is 0.665 higher than regions with a port in category 1.

Since data are available, we proceed with a sectoral analysis to control for possible differences in

the port size effect between attracting FDI in Manufacturing and Services Industries. Once again,

Services sectors seem to be more responsive in the presence of a port in an area. What is

paradoxical here, is the insignificant coefficient of the ‘Mainport’ category in investments in

Manufacturing. Extremely high standard errors and relatively big effects could be signs of

inadequate variation within the business sectors or omitted variable bias.

The results of the rest independent variables are identical through all the estimation models and

yield some critical findings about the determinants of investing in European regions for the

period 2003-2010. Those are discussed in the next section.

25

Table 2: The Port Size Effect

.

26

6. DISCUSSION & CONCLUSION

Our findings on the attracting effect of ports are in line with the idea that some cities and ports

are favored by nature, being in proximity to the main lines that connect different parts of the

world within the global networks (Zhao, Xu, Wall, & Stavropoulos, 2017). Ports play a

significant, positive role in the ability of urban competitiveness to attract FDI. Examples of such

ports and cities in our sample are the Mainport of Rotterdam, the port of La Havre in France and

the port of Felixstowe in UK. The NUTS-2 classification is particularly useful in socio-economic

analysis of regional policies4. In our case this acts as a limitation since it provides information on

aggregated areas. It is expected that the port effect will vary across different areas in a NUTS-2

region basically due to the different sizes and characteristics of these regions. To eliminate that

bias we estimated in two ways our model and the results were consistent. An interesting aspect

would be to make a case based on city and not regional information in order to eliminate any bias

induced by geographical and economic clusters. That would yield useful information for policies

and institutional strategic planning that strive to escalate the urban economic performance

influenced by the existence of a port in this particular area.

The fact that the effect for investments in Services is stronger than for Manufacturing, might not

satisfy our expectations, nonetheless there are some arguable explanations on this. As mentioned

in ‘Introduction’, Manufacturing sectors tend to concentrate in Eastern Europe. Clearly, this is

not the case here as main European ports are in the Western part of the continent. This finding

confirms the stronger effect we had for the Service Sectors but is not enough to rest in our

laurels, so we also considered some additional reasons. Extant literature suggests that port cities

have been developed far and segregated from their ports because of the rise of advanced service

4 http://ec.europa.eu/eurostat/web/nuts

27

industries where the port does not meet the requirements for such operations (Zhao, Xu, Wall, &

Stavropoulos, 2017). At the same time, ports create excellent circumstances for companies

specialized in advanced producer services, especially for maritime and port-related services

(Jacobs, Political Economy of Port Competition , 2007). It is also important to take under

account that regions with most investments in our sample are big financial centers which

facilitate international firms in shipping and maritime activities like insurances and other

supportive services such as London and Amsterdam. A list if the top 10 NUTS-2 regions with

their main specialization can be found in Table 3 in the Appendix where only three of the 131

port regions defined in our study are included. If with past data the effect for FDI in services was

stronger, imagine how different this effect could be today after 8 years of incessant advance and

automation integration. It might be also the case that investing MNCs have entered a particular

region driven by the attractiveness and the status of the host country where the latter, in turn, can

enhance the general performance of the former (Karreman & Van der Knaap, 2010). One

direction for further research is to collect adequate data that capture all these port and urban

characteristics and infer what matters more for location choices; the port, the city or a

combination of those two? Highly influential port-specific factors like the port authority - private

or public - should be taken under account as well as urban-specific characteristics like the

potentiality of a market captured by growth rates. In our case it was impossible due to data

unavailability but we would be glad to incorporate such aspects in a future study.

Finally, we find a positive relationship between the log number of total investments and presence

of a main port in a region. The confounding result on the sectoral heterogeneity has three

possible explanations. First, the port size is not the most important criterion. Commercial and

geopolitical conditions as well as just the big number of ports in an area are factors able to set the

28

port in the center of the urban economic activity. (Roa, Peña, Amante, & Goretti, 2013).Second,

In the same study, it is found that large ports do not exhibit high potentiality in goods

management but instead features that downplay its importance as for the volumes handled and

the size of the hinterland. Third, the ports included in our study are mainly deep-water seaports

that basically handle freight (measured in TEUs), neglecting recreational, military, fishing and

tourism activities as well as river, harbor and terminal ports. In both cases, a thorough analysis

must be done distinguishing between the different types of port and activities in order to account

for this heterogeneity and draw specific conclusions, appropriate for inference to a greater

population.

As it seems, FDI in Europian regions is strongly driven by efficiency and strategic-asset seeking

motives. Investing MNCs strive to benefit from costs differences between the country of origin

and entrance and is becomes apparent from the significance levels (1%) of the corresponding

controls -wages, taxes, R&D expediture. The latter especially is in line with what Dachs et al.

stated that R&D has been overly extended in Europe and is inescapable for multinationals that

strive to develop goods and services outside their country and not only produce and sell them

(Dachs, Kampik, Scherngell, & Zahradnik, 2012). At first sight, own sector employment does

not display any important role in attracting FDI, something hard to accept as true. However, high

standard errors combined with high standard deviation obtain by summary and descriptive

statistics, may indicate the following: our sample is not really representative of the actual

population or sector specialization varies a lot in our sample and our model in not appropriate to

exploit all this explanatory power that these variables add in the model. In every case, we give a

suggestion to any future researcher whose focus will be on the contribution of ports in the

attractiveness and socio-economic performance of the cities: it is impossible to draw generalized

29

conclusions not taking under account different port types and activities or controlling for cities

heterogeneity, so it might be more informative to make comparisons between the different

regions and cities of the same country rather on a continent level.

30

REFERENCES Bevan, A. A., & Estrin, S. (2004). The determinants of foreign direct investment into European transition

economies. Journal of Competitive Economics, 775-787. Bronzini, R. (2004). Foreign Direct Investment and Agglomeration: Evidence from Italy. European

Regional Science Association.

Burger, M. J., Van der Knaap, B., & Wald, R. (n.d.). Revealed Competition for Greenfield Investments

between European Regions. Tinbergen Institute.

Cullinane, K., Teng, Y., & Wang, T. (2006). Port competition between Shanghai and Ningbo. Maritime

Policy & Management.

Dachs, B., Kampik, F., Scherngell, T., & Zahradnik, G. (2012). Internationalisation of business

investments in R&D. Austrian Institute of Technology.

Disdier, A.-C., & Mayer, T. (2004). How different is Eastern Europe? Structure and determinants of

location choices by French firms in Eastern and Western Europe. Journal of Comparative

Economics, 280-296.

Dooms, M., Lugt, L. v., & Langen, P. W. (2013). International strategies of port authorities: The case of

port of Rotterdam Authority. Research in Transportation Business & Management.

Ducruet, C., Rozenblat, C., & Zaidi, F. (2010). Ports in multi-level maritime networks:evidence from the

Atlantic. Journal of Transport Geography.

Dunning, J. (1981). International Production and the Multinational Enterprise. London: Allen & Unwin.

Dunning, J. H. (2001). The Eclectic (OLI) Paradigm of International Production: Past, Present and Future.

International Journal of the Economics of Business, 173-190.

Ferrari, C., O., M., Bottaso, A., & Tei, A. (2012). Ports and Regional Development: A European

Perspective. Paris: OECD Regional Development Working Paper.

Ferrari, C., Parola, F., & Gattorna, E. (2011). Measuring the quality of port hinterland accessibility: The

Ligurian case. Transport Policy.

Hall, P., & Jacobs, W. (2012). Why are maritime ports (still) urban, and why should policy-makers care?

Maritime Policy Management.

Hoyle, B. (1989). The port—City interface: Trends, problems and examples. Geoforum, 429-435.

Huang, Y. (2009). The growth of global hub port cities under globalisation: The case of Shanghai

international shipping centre.

Jacobs, W. (2007). Political Economy of Port Competition . Nijmegen: Academic Press Europe.

Jacobs, W., Ducruet, C., & De Langen, P. (2010). Integrating world cities into production networks: The

case of port cities. Global Networks, Wiley, pp.92-113. <halshs-00459509>.

Karlsen, T., Silseth, P. R., Benito, G. R., & Welch, L. S. (2003). Knowledge, internationalization of the

firm, and inward–outward connections. Industrial marketing management.

Karreman, B., & Van der Knaap, B. (2010). The Geography of Equity Listing and Financial Centre

Competition in Mainland China and Hong Kong. Journal of Economic Geography.

Kolstad, I., & Wiig, A. (2012). What determines Chinese outward FDI? Journal of World Business.

Kostiainen, J. (2002). Urban economic development policy in the network society. Tampereen Yliopisto.

Ledyaeva, S. (2009). Spatial Econometric Analysis of Foreign Direct Investment Determinants in Russian

Regions. The World Economy.

Lovering, J. (2003). MNCs and wannabes: inward investment, discourses of regional development and

the regional service class. Cheltenham, UK: Edward Elgar.

Makabenta, P. M. (2002). FDI LOCATION AND SPECIAL ECONOMIC ZONES. RURDS Vol. 14.

Marshall, A. (1890). Principles of Economics ,8th edition.

Merk, O. (2014). The Competitiveness of GlobalPort-Cities:Synthesis Report.

Merk, O., & Hesse, M. (2012). The Competitiveness of Global Port-Cities: The Case of Hamburg,

Germany . OECD Regional Development Working Papers.

Meyer, K. (1995). Foreign direct investment in the early years of economic transition: a survey.

Economics of Transition, 301-320.

31

Nijdam, M., & van der Horst, M. (2018). Port definition, concepts and the role of ports in supply chains.

In Ports and Networks (p. 10). Abingdon: Routledge.

Notteboom, E. (2010). Concentration and the formation of multi-port gateway regions in the European

container port system: an update. Transport Geography, 567-583.

Nyamai, D., & Wall, R. (2015). WEALTH AND THE CITY: The Competitiveness of Port cities and Non-

port cities. Rotterdam.

Porter, M. E. (2000). Location, Competition, and Economic Development: Local Clusters in a Global

Economy. ECONOMIC DEVELOPMENT QUARTERLY, 15-34.

Roa, I., Peña, Y., Amante, B., & Goretti, M. (2013). Ports: definition and study of types, sizes and

business models. Journal of Industrial Engineering and Management, 1055-1064.

Sáez, L., & Periáñez, I. (2015). . Benchmarking urban competitiveness in Europe to attract investment.

Cities.

Song, L., & van Geenhuizen, M. (2014). . Port infrastructure investment and regional economic.

Transport Policy.

Stopford, M. (2009). Maritime Economics . In M. Stopford, Maritime Economics (3rd ed., p. 81).

Abingdon: Routledge.

Wang, M., & Wong, S. (2009). What Drives Economic Growth? The Case of Cross. Kyklos.

Yeo, G., Roe, M., & Dinwoodie, J. (2011). Measuring the competitiveness of container ports:

logisticians' perspectives. European Journal of Marketing.

Zhao, Q., Xu, H., Wall, R., & Stavropoulos, S. (2017). Building a bridge between port and city:

Improving the urban competitiveness of port cities. Journal of Transport Geography, 120-133.

32

APPENDIX

Table 3: Investment Portfolios of the Top 10 NUTS-2 Regions

33

Table 4: Business Sectors Taxonomy

34

Table 5: Business Functions Taxonomy

35

Table 6: Distribution of ‘Large’ and ‘Main’ ports in TEU

0 2000000 4000000 6000000 8000000 10000000 12000000 14000000

Port of Rotterdam

Port of Antwerp

Port of Hamburg

Port of Bremen

Port of Algeciras

Port of Valencia

Port of Felixstowe

Port of Gioia Tauro

Port of Piraeus

Port of Marsaxlokk

Port of le Havre

Port of Genoa

Port of Brugge-Zeebrugge

Port of Southampton

Port of Barcelona

Port of London

Port of Las Palmas

Port of Marseille

Port of Gdansk

Port of Gothenburg

Port of Livorno

Port of Venice

Port of Liverpool

Port of Bilbao

Port of Dublin

Port of Lisbon

Port of Napoli

Port of Leixoes

Port of Sines

Port of Helsinki

Port of Aarhus

Port of Trieste

Port of Klaipeda

Port of Teesport

Port of Thessaloniki

Port of Grimsby

Port of Milford

Port of Dunkirk

Port of Riga

Port of Limassol

Zeeland Seaports

Distribution of 'Large' and 'Main' ports in TEU

36

Table 7: Distribution of ‘Small’ and ‘Medium’ ports in TEU

0 50000 100000 150000 200000 250000

Port of Belfast Port of Grangemouth

Port of Vigo Port of Ravenna Port of Messina Port of Rostock

Port of Tallinn Port of Oslo

Port of Nantes Port of Ancona

Port of Heraklion Port of Rochelle

Port of Copenhagen/Malmo Port of Bristol

Port of Cartagena Port of Szczecin

Cromarty Firth Port Port of Ghent

Port of Plymouth Port of Sandefjord Port of Stavanger

Clydeport Port of Civitavecchia

Port of Cardiff Port of Gijón

Groningen Seaports Port of Bordeaux

Port of Odense Port of Palma de Mallorca

Port of Stockholm Port of Tyne

Port of Amsterdam Port of Bergen

Port de Saint Valery sur Somme Port of Barrow

Port of Corfu Port of Kavala

Port of Koge Port of Monaco

Port of Oskarshamn Port of Patras

Port of Trondheim Port of Melilla

Port of Brest Port of Aalborg

Port of Aberdeen Port of Faro

Port of Galway Port of Pescara Port of Tromsø

Port of Santander Port of Brindisi

Distribution of 'Small' and 'Medium' ports in TEU

Table 7: Distribution of ‘Small’ and ‘Medium’ ports per TEU