The Effects of the European Integration on

Economic Growth

Aidan J O’Galligan

1602986

Supervisor: Dr Miguel Angel Gavilan Rubio

EC831 Project

Department of Economics

University of Essex

23/04/2019

Word Count: 8,638

Abstract
This paper tests the effect of European Integration on the growth of the European Countries,
by defining European Integration in terms of Single Market and Fiscal Convergence, whilst
accounting for the different membership positions within the European Union. The main
finding is that accounting for Single Market Integration, EU membership boosts economic
growth by 0.6%. However further Integration in terms of Eurozone membership and
Schengen both negatively affect growth by 0.62% and 0.52% respectively. The sample
covers twenty-four of the current twenty-eight EU members, and the model uses a Weighted
Least Squares Panel regression, with economic growth defined by the change in GDP per
capita.

1
Acknowledgements
I would like to thank my supervisor Miguel for the continuous support for this paper,
especially in terms of the econometrics and presentation of the final paper, helping me
elevate my analysis to next level.

2
 TABLE OF CONTENTS

INTRODUCTION ................................................................ 5

1. HISTORY OF EUROPEAN INTEGRATION .............. 6

1.1 VERTICAL INTEGRATION ................................................... 6

1.2 HORIZONTAL INTEGRATION ............................................ 7
2. LITERATURE REVIEW ................................................ 8

3. EMPIRICAL ANALYSIS .............................................. 12

3.1 EUROPEAN INTEGRATION MEASURE ............................. 12

Single Market .................................................................. 13
Fiscal Convergence ......................................................... 14
Membership ..................................................................... 15
3.2 GROWTH METHODOLOGY ............................................. 16
4. DATA DESCRIPTION .................................................. 18

5. RESULTS ........................................................................ 19

6. CONCLUSION ............................................................... 23

REFERENCES ................................................................... 25

APPENDIX A ..................................................................... 26

3
4
Introduction
The motivation for this paper arises from a quote from the current President of the European
Commission, Mr Jean Claude Juncker; “Europe must be about more than market, goods and
money”. Now why is this relevant? The European Plan is of further integration, and from this
quote it wouldn’t be unreasonable to envision a singular European State. These are
significant changes to the European culture and have and will continue to effect over five-
hundred-million citizens. Another reason why the topic of European Integration is
particularly relevant is the 2016 vote in the UK. This paper is not a paper discussing the
potential effects of Brexit to the British and European economies; instead we can estimate the
effect of European Integration to the economic growth.
The formal theory surrounding regional integration is from Bela Balassa (1963), indicating
the step-by-step process to achieve complete regional integration. The steps begin by initially
removing some tariffs in Preferential Trade Agreements (PTAs), this converts into a Free
Trade Agreement (FTA) whereby no trade is subject to tariffs. A Customs Union includes
free trade with the addition of a common external tariff. The other factors have free
movement within a Common Market, and Monetary Union requires a common currency. The
next step of integration, partners the Monetary Union in terms of policy effects, and this is a
Fiscal Union whereby central control of government expenditure and taxation. Political
Union completes the process of integration.
The specific European Integration steps can be found in Table 1, and what is clear from this
is that firstly the theory does somewhat simplify these steps. Another caveat of Balassa’s
theory of regional integration is the lack of distinction between vertical and horizontal
integration, when looking at Table 1 we can see that the two are in fact not mutually
exclusive.
Given the relevance of this topic, this paper will propose the following statement: European
Integration has no effect on the economic growth of the European economies. To test this
hypothesis, we first have to define European Integration, and doing so this paper aims to
define each country heterogeneously in terms of their European Integration. This will take the
form of accounting for convergence in the European Single Market and Fiscal Behaviour
relating to membership of the EU, and by using a Weighted Least Squares panel regression;
we measure the growth given by the GDP per capita to appropriately see the direct effect to
the half-a-billion EU citizens, over the years 1996-2016. The baseline results indicate that
European Integration, when accounting for the Single Market convergence boosts the per
capita GDP of the European countries. However, these results also show that membership of

5
the Eurozone and Schengen offset these effects. The Fiscal convergence yields no significant
results but should still be reported.
The project will first discuss the European Integration and then survey the literature
surrounding the question of European Integration and economic growth. Following that, the
empirical analysis and data will be described, yielding the results section of the paper. Then
the caveats of this discussion will commence.

1. History of European Integration

In this section, I will briefly discuss the history of cooperation and policy coordination in the
European sense, and then link back to the theory of Regional Integration from Balassa
(1963).

1.1 Vertical Integration

As stated above, Regional Integration is not one dimensional in the European context;
I like to define vertical integration as the traditional sense of integration whereby the
members of the bloc increase their coordination and therefore their interdependence
on each other. In terms of the European Continent, this was the central point of the
Industrial Revolution driving the ‘Smaller World Hypothesis’ through trading with
each other and driving global colonisation. The end of the 19th Century brought strong
growth in the region, however the impact of both of the World War’s lead to the
destruction of many European Nations. At the end of World War Two, the United
States (US) came together with the leaders of Europe to create the Marshall Plan
(1948). This agreement was to ensure that war in Europe was not an option and a
greater sense of cooperation existed between the countries. After the funds from the
US began to rebuild Europe, the Continent looked to coordinate the region. The Coal
and Steel Community (1950) was the first step of trade integration, comprising of the
‘Original Six’. In 1957 the Treaty of Rome was initiated, creating a Common Market,
often referred to as the ‘European Economic Community’ (EEC). It became a
Customs Union in 1968 with a greater protection of the Agriculture within the region.
In 1992 the Maastricht Treaty was ratified which transformed the EEC into the

6
 Table 1 – Vertical and Horizontal Integration in Europe
Vertical Integration Horizontal Expansion
1957: Common Market Original Six: Belgium, Netherlands,
Luxembourg, Italy, Germany, France
1968: Customs Union
1973: First Enlargement; UK, Ireland,
Denmark
1981: Greece
1986: Spain and Portugal
1992: Treaty of Maastricht
1995: Schengen Introduced Austria, Sweden, Finland
1999: Eurozone Introduced
2004: CEE101
2007: Bulgaria and Romania
2009: Lisbon Treaty
2013: Croatia
European Union and paved the way for the introduction of Monetary Union in 1999,
by providing the criteria for convergence2 and also created the Single Market. Finally,
in 2009 more parliamentary power was handed to the European Union through the
Lisbon Treaty. Future integration can be adhered to by assessing the ‘Five President’s
Report’ (2015), which pushes for a closer Union, specifically addressing the plans to
first update the economic union, whilst also determining the mechanisms for fiscal
and financial union.

1.2 Horizontal Integration

Horizontal integration can be defined as the widening of the Union. This can be
measured by the Expansions of the Union. The actual expansions will be described in

the table above, alongside the vertical steps to make comparisons as to the state of the
Union when each expansion took place.
Of course, in 2016 the United Kingdom voted to leave the European Union, it is yet to
be seen what relationship the UK and the EU will have once the departure is
complete. It is also important to stress the fact that countries can be vertically
integrated into Europe without being full members for instance Switzerland and
Iceland who participate in the Single Market and Schengen without being full
members of the Union. This paper will only focus on the EU members, but there is
scope to consider these countries.
1
CEE10 countries are the following: Estonia, Cyprus, Malta, Slovakia, Slovenia, Czech Republic, Latvia,
Lithuania, Poland and Hungary.
2
Discussed further in section 3.1 (pp.14)

7
2. Literature Review
From humble beginnings, European Integration was intended to merely act as a form of
cooperation between the European Nations. Fast-forward seventy years, and there is shared
institutions between European countries, as well as policy coordination and in the not too
distant future, full Economic Integration3. This process of Regional Integration has happened
at a very fast pace and has vast economic implications. The literature assessing different ways
European Integration has affected members, this paper will consider only those that have
captured growth effects, as growth captures the income of the entire country.
To distinguish the differences in previous literature, we must consider how different authors
have approached European Integration and Growth Modelling. The authors of previous
literature have applied three generalised measures to European Integration: firstly in terms of
various fundamentals such as trade or monetary union; then expansion of members and
therefore markets; and finally applying more sophisticated models such as Indices of
European Integration. There is also different econometric methodologies used, as well as
different countries and periods applied across all the papers, so the scope for direct
comparison of results is restricted. Instead this Literature Review will compare the different
ways of measuring European Integration and then discuss the conclusions and limitations of
individual papers.

The first measure of European Integration we can determine is Trade. Henrekson et al (1997)
was one of the earlier papers discussing European Integration in terms of trade. The model of
European Integration is very simple, as the paper compares if a country is within the
European Community or European Free Trade Area and assesses the effects on GDP growth.
The paper included a regression on growth of GDP with a dummy variable for EC/EFTA
membership. The model is limited and falls into the trap, like many other papers, defining
European Integration as homogeneous amongst all members. Despite this limitation, the
paper finds that the coefficient of the dummy variable for European Integration is positive
and significant, whilst also finding that the cause of the increasing growth rate was due to
technological transfer as opposed to investment, which follows partially the conclusions of
the Solow Growth Model4.
To further develop Henrekson et al (1997), Badinger (2001) adapted the econometric
approach to separate growth into technology and investment, with both temporary and

3
See the Five Presidents Report (2015), from the European Commission.
4
See Solow (1956)

8
permanent effects. The measure of integration is also in terms of trade, controlling for the
effects of the General Agreement on Tariffs and Trade (GATT). In his analysis, Badinger
finds that if no economic integration had taken place, between 1950 and 2000, GDP would be
one-fifth lower. Interestingly, Badinger attributes two-thirds of this result towards GATT
unilateral liberalisation of tariffs. Although a valid contribution, this paper does not account
for positive integration features including non-tariff barriers, such as common institutions and
policies, which means this paper can only partially explain the effect of European Integration.
Mann (2015) instead evaluates how successful European Integration has been for growth of
East European countries, who joined the EU in 2004 and 2007. The main findings are that
there are medium-run growth effects, found from a regression subject to the Augmented
Solow Growth Model. Due to a lack of data, the approach of an index is deemed not possible.
Instead this paper uses trade as a proxy, whereby European Integration is defined as trade
with EU27 as a percentage of total trade. This measure has stronger implications than
Henrekson, as countries are weighted on how integrated they are within Europe in terms of
trade. The issues apparent are that a country not in the Eurozone could have a higher trade
ratio than a country inside of the Eurozone, whilst being defined as more integrated. This
expresses a clear example of the potential limitations of this measure, and therefore limits the
conclusions.
There are other forms of integration we can isolate as potentially capturing European
Integration, such as monetary union. Luque and Taamouti (2013) want to assess the effect of
fundamental economic measures and if adopting the Euro affects these. Accounting for the
first twelve members of the monetary union by using a panel regression from 1980-2011,
they find that adopting the Euro and losing control of independent monetary policy that many
macroeconomic fundamentals are affecting growth. De Grauwe (2006) somewhat explains
these differences through Optimum Currency Area Theory, and that without full Economic
and Monetary integration, using Balassa’s (1963) definition, there will be issues within the
Eurozone. Luque and Taamouti (2013) do clarify that much of the dispersion is due to the
Financial Crisis of 2008, but De Grauwe (2013) reiterates the point that “the Euro is a
currency without a country”, which implies that had the European Union had a common
fiscal institution, maybe the effects of the financial crisis would have been somewhat limited.

We now assess the Literature with regard to European Integration expressed by an increase of
EU members. With the exception of Badinger (2001), the issues with using trade and tariff
analysis is that the global tariff liberalisation is not distinguished from European Integration.
9
Similarly, by isolating a single variable as “integration” the papers fail to capture the total
effect of European Integration. Instead many author’s define European Integration in terms of
European Expansion, as it allows us to compare growth of accession countries before and
after Integration takes place. Deardoff and Stern (2002) apply a theoretical model based off
increasing returns to scale and conclude that the incumbent members gain from expansion,
whilst having mixed success from countries that join later. The authors also had doubts that
European Integration has any affect on long run growth. Any theoretical model without any
econometric basis draws limits on any conclusions, and it’s also worth considering that their
analysis fails to include the largest expansion of the EU in 2004, highlighting the limited
scope of their research.
Another approach to capture the effects of European Expansion is through synthetic
counterfactual analysis and is used by Campos et al (2014). By using a difference in
difference measure, the authors take each enlargement of the EU between 1973 and 2004,
and compare GDP performance and labour productivity had the accession countries not
joined the EU. The conclusions are that, on average, GDP per capita would be 12% lower.
Austria and the countries of the 1973, 1986 and half the 2004 expansions saw large gains in
GDP and labour productivity. It follows that there were smaller effects for the remaining
1995 expansion countries, as well as Poland, Czech Republic and Slovakia. This paper found
that only Greece would have had higher GDP and labour productivity if they had never
joined the European Project. Any counterfactual analysis is subject to many unknowns, and
having been held over many years reduces the likelihood of ceteris paribus. In particular to
this paper, the conclusions do not consider effects to the Original Six members, and perhaps
more importantly, fails to address that Europe became more integrated over the period and
does not distinguish the effects of each integration point. This is fundamental as the CEE
countries did not join the same Bloc as the 1973 expansion, and therefore the latter captures
mainly the single market effect, whilst the former capture the whole European Union, which
limits the defendability of Campos et al’s analysis.
Another approach related to expansion would be countries actually leaving. Crafts (2016)
tackles the issue of European Integration in terms of what would happen if the UK left the
European Union, or “disintegration”. By surveying the literature and data analysis regarding
growth, Crafts finds that European Union membership has increased GDP per capita by 10%
per year, which is greater than the 1.5% GDP cost of membership. This is a comparison of
UK GDP as members of the EU against EFTA membership. Like Deardoff and Stern, this
paper has no econometric model, so when accounting for the validity of the conclusions, we
10
must be cautious. The comparison of different countries also falls in to the traps of Campos,
by taking the year a group joined and comparing the growth per capita.

By not accounting for individual weighting in terms of European Integration, we assume first
that all countries are equally integrated within Europe, as well as each step towards
integration carries the same weight in terms of effect on growth. The first is certainly false, as
the UK is not in the eurozone and therefore less integrated than Belgium. It would also be a
fair assumption that the steps towards integration (monetary union/single market access etc)
do not have identical effects on growth. Therefore, to measure European Integration, data
must be collated on various integration variables, and then weighted based off how each stage
of integration affects growth.

Rapacki and Próchniak (2009) apply a regression, derived from beta and sigma convergence,
to define GDP growth by two sub-factors: eight growth variables from the convergence
models and four measures of European Integration. The measures in terms of European
Integration include; Foreign Direct Investment, Economic Freedom, Transition Indicator, and
Aid. The analysis is regarding the CEE-10, who joined in the EU in 2004, thus the paper
focuses more about convergence to the EU average levels, but this can only be achieved
through European Integration for these countries. They find, similarly to Mann (2015), that
European Integration leads to convergence of the CEE 10 bloc, which in turn caused high
growth rates for these countries. There is limitations in this paper, namely that the variables
used for European Integration, such as Freedom and Transition, could be inflated by the fact
that many countries came from Ex-Soviet backgrounds where these variables would naturally
be much lower. A similar comparison would be the ‘Golden Age of Economic Growth in
Europe’ whereby the European Community countries were destroyed, and consequently their
GDP retracted significantly, and therefore recovery inflated growth to return to post-war
GDP levels. Another critique would be the fact that Slovakia and Slovenia joined the
Eurozone whilst many other CEE10 countries did not, so again this paper falls into the trap of
homogenising the Bloc, also not mentioning the effects to the incumbent members.
A more sophisticated model of Integration would be from König and Ohr (2013). They
manage to capture European Integration in terms of: the Single Market; Homogeneity;
Symmetry; and Conformity, and through a weighting procedure manage to rank the European
Countries in terms of how integrated they are in the years of 1999 and 2010. Despite not
having clear mathematical calculations, there is substance to their approach, as their findings
11
seem fairly realistic on the face of it. For example, countries that are not in the Economic
Monetary Union cluster together in terms integration scores. This paper does not apply the
index to measure growth, and one outstanding issue of König and Ohr (2013), is that they use
the cyclical growth data which may skew the results of their final index as the European
countries economic cycles are more likely to align given the nature of European Integration.
Therefore it is essential to utilize the approach of scoring each country in terms of European
Integration and compare how the precise level of integration affects the economic growth of
the European Union Members.

3. Empirical Analysis
This section will explain firstly how to capture Vertical and Horizontal Integration of the
European Countries, and then secondly formulation of a growth model to test the impact of
the different measures of European Integration.

3.1 European Integration Measure

Stated many times throughout this paper, I want to stress the importance of the
heterogeneity in terms of European Union membership. To avoid homogenising the
bloc, we need some way of differentiating the European countries within Europe. One
obvious way is to look at the membership options, but to capture European
Integration this is necessary but not sufficient as we again fall into the homogenous
trap, by saying all countries who have the same membership status are equally
integrated. This is an issue in the sense that if we take say Italy and France, both
countries are members of: the EU, the Eurozone and the Schengen Area for the same
number of years, and to then say these countries have the precise same relationship in
Europe would seem unclarified. To counteract this assumption, consideration to
different European Union policies is necessary to see if there exists a convergence
within the Union.
Inspired from König and Ohr(2013) who created an index for European Integration, I
believe a better approach would be to consider how integrated each country is in the
European Union. In their paper they account for a single market index, based off the
flows of the Four Factors, which have free movement in the European Single Market.
They also include indices for other macroeconomic fundamentals, including the

12
 business cycle, in terms of their homogeneity and symmetry indices. Finally, they
apply a conformity measure based on European Court of Justice Violations.
This precise calculation is not perfect when testing the effect of the integration on
economic growth as there exists collinearity issues as the cyclical nature is part of
their index. Alas, the single market measure is an important consideration in which
European Countries can deviate in terms of integration.
Another important policy of the European Union, which was part of the Maastricht
Treaty (1992) is the Euro Convergence Criteria. This was the policies set out by the
European Commission to ensure that before the Euro currency was introduced, the
countries had sustainable performance in terms of government finance, interest rates
and inflation.
To capture the Convergences, we use the following formula:

(𝑉𝑎𝑙𝑢𝑒𝑖,𝑡 − 𝑀𝑒𝑑𝑖𝑎𝑛𝑖,𝑡) /𝑀𝑒𝑑𝑖𝑎𝑛𝑖,𝑡

𝐶𝑜𝑛𝑣𝑒𝑟𝑔𝑒𝑛𝑐𝑒 = 1 − | |
𝑀𝑎𝑥 𝑉𝑎𝑙𝑢𝑒 ((𝑉𝑎𝑙𝑢𝑒𝑖,𝑡 − 𝑀𝑒𝑑𝑖𝑎𝑛𝑖,𝑡) /𝑀𝑒𝑑𝑖𝑎𝑛𝑖,𝑡

By using the median instead of the mean the score will robust to extreme outliers and
for the late 1990’s there exists some very large outliers. After applying this formula
each country in each year will receive the median deviation score, a number between
1 and 0. The closer the score to 1, the closer the value to the median and so greater is
the European Convergence and therefore the Integration.

Single Market
In König and Ohr’s (2013) Single Market Index, they apply the approach of
measuring the EU single market flows5 by two measures. First are the EU flows as a
percentage of total flows, and then they take the EU flows as a percentage of Gross
Domestic Product.
Due to the lack of data available I cannot apply the same measure and also the lack of
accurate service data, due to World services having very little barriers to trade, this
model will only consider labour, goods and capital.
Again, as the data for labour flows are so inconsistent6 we instead approach with an
economic argument to justify the use of the unemployment rate. When there is free
movement of labour, we assume that if there are deficits in one country and a surplus

5
The flows being Goods, Services, Capital and Labour.
6
The data is inconsistent as there are no requirements for countries to record their inflows of labour and so we
do not have accurate data.

13
in another, that the rational labour who is unemployed will move to the country where
they can get a job. In this sense, we are considering how well the host country can
attract the labour. Of course, this is not the only factor that drives the employment rate
in a given country, and we also have to look at the literature surrounding Optimum
Currency Area to actually see how flexible European citizens are to move to another
country within the Union, as there may be constraints such as language barriers but
we can still apply the economic intuition as a proxy for measuring labour mobility in
the EU. Solely using the unemployment rate does not hold well for European
Integration as the same economic logic can be used for international labour market
integration, and so by applying the above Median deviation formula, can see how the
unemployment rate in a given year convergences to the European average.

An interesting paper from Svirydzenka (2016) attributes all of the fundamental

measures that account for capital market performance and generate a Financial
Development Index. An immediate question would be to ask why that is relevant to
European Integration. The link is that when a country is part of a single market, and
for some countries a monetary union, their external capital controls are set by the
European Union. This means that all twenty-eight members abide by the same rules,
meaning internal and external investors follow the same regulations. This should
facilitate intra and extra-EU flows and increase their financial development, and so
we can use these scores to measure capital in the single market. Again, applying the
median deviation formula, we can find the relation to the EU average for Financial
Development, making the measure relevant to European Integration.

Another fundamental factor is the trade in goods. For this measure there is sufficient
data to allow measuring the importance of EU trade. This paper will take the value of
Intra-EU trade of a country and divide that by the value of total trade. By using this
measure, we can say how integrated a given country is within the goods market in the
EU, as measures the trade with the twenty-seven other members.

Fiscal Convergence
The Convergence Criteria stated in the Maastricht Treaty was originally designed to
stabilise the European Countries before they aligned their currencies and the policies
have developed over time for example, now including the Stability and Growth Pact.

14
 Table 2 – Original Convergence Criteria
Aim Technical Measure Target
Price Stability Consumer Price Inflation Not more than 1.5% above
Rate the three best performing
Sound Public Finances Government Deficit as % of Not more than 3%
GDP
Sustainable Public Finances Government Debt as % of Not more than 60%
GDP
Durability of Convergence Long-term Interest Rates Not more than 2% above the
three best performing in
terms of Price Stability
Exchange Rate Sustainability Deviation from Central Rate Participation in ERM for two
years without severe tensions

The original criteria focused on five main areas, specified in Table 2 above whereby
to join the Euro the countries must follow the guidelines of these rules. As this is a
major policy it is essential to include when defining European Integration.
As we are looking to use these criteria over a time period once some countries have
already joined the Euro, we will ignore the final row of the table and focus our
analysis in terms of Public Finances, Interest Rate performance and Price Stability.
These measures will also follow the median deviation calculation due to issues with
the target rates, namely being that other policies such as the Stability and Growth Pact
came into effect, making the original criteria somewhat redundant. There is also great
evidence for inconsistent enforcement and punishment in regard to exceeding these
policies, but the sentiment exists that the measures used are important when
considering European Integration.
The Exchange Rate is a difficult measure as nineteen of the twenty-eight current
members are in the Eurozone and therefore is excluded from this analysis.
By taking the deviation we are not measuring how effective is price stability in terms
of economic growth, we instead consider does converging to the European Union
average price stability have an effect on economic growth.

Membership
Another measure we need to consider is that there are different membership options
of the European Union. An example that I considered when I discuss the motivation
for this topic, as well as my criticism of the existing literature, is that for instance,
Sweden is currently not a member of the Eurozone whilst France is. Rationally if we
ignore this effect for Eurozone participation, we will ignore the effects of the

15
 differences in national currency. This is problematic, as intuitively I could argue that
an investor in France will consider the exchange rate stability when investing in
overseas assets. If the investor had the opportunity to invest in Sweden (not in
Eurozone) or into Germany (in the Eurozone), the currency removes some of the
uncertainty which will affect the investors final investment decision. If we aggregate
this effect to the whole European economy, we may see the effect of being in the
Eurozone. The same argument can be made for the Schengen Area whereby boarders
within Europe are essentially removed, and that without the restrictions across
mainland Europe, the movement of peoples may have a more efficient labour market
equilibrium. We also account for European Union membership to control for the
expansions discussed in section. These dummy variables proxy the horizontal
memberships over time, as the countries of Europe do not have homogenous
memberships in the Union and so there is a need to distinguish them.
The way in which this is determined is given simply by dummy variables for
membership of the European Union, the Eurozone, and the Schengen Area, for a
given country in a given year.

3.2 Growth Methodology

Now we have controls for European Integration, we can look at what happens to
growth as countries integrate. When looking for a model specification, we need to test
the variables described above to see the effect of European Integration on growth.
Economic growth can be interpreted in many ways; this model will focus on GDP per
capita growth. This is because it accounts for the standard of living within a given
country which better serves the motivation described in the Introduction of this paper.
The base model will take the following form:

∆𝑦𝑖,𝑡 = 𝑦0 + ∆𝑘𝑖,𝑡 + 𝜆𝐸𝑈 + 𝜆𝐸𝑢𝑟𝑜𝑧𝑜𝑛𝑒 + 𝜆𝑆𝑐ℎ𝑒𝑛𝑔𝑒𝑛 + ∆𝑃𝑟𝑖𝑐𝑒𝑠𝑖,𝑡 + 𝜀𝑖,𝑡 (1)

Where ∆𝑦𝑖,𝑡 is the growth rate of GDP per capita and 𝑦0 is the value of GDP per
capita in the base year 1996. ∆𝑘𝑖,𝑡 is the growth rate of the Gross Fixed Capital
Formation (GFCF) and λ represents the dummy variables for different membership
positions in the European Union. ∆𝑃𝑟𝑖𝑐𝑒𝑠𝑖,𝑡 represents the growth in the median
deviation of the Consumer Price Inflation.

16
 The growth of GFCF is used to measure for capital stock growth without considering
the depreciation as this shows the growth of Investments which is a fundamental
driver of the economy. Traditional models account for labour quality with a usual
measure being the Years of Schooling; this has been excluded from this model due to
accurate data records every five years and this model is annual growth.

The model allows us to consider convergence within the Single Market and Fiscal
policy separately to isolate the effects.
The first variation of (1) will be for the Single Market:

∆𝑦𝑖,𝑡 = 𝑦0 + ∆𝑘𝑖,𝑡 + 𝜆𝐸𝑈 + 𝜆𝐸𝑢𝑟𝑜𝑧𝑜𝑛𝑒 + 𝜆𝑆𝑐ℎ𝑒𝑛𝑔𝑒𝑛 + ∆𝑃𝑟𝑖𝑐𝑒𝑠𝑖,𝑡 + ∆𝐿𝑎𝑏𝑜𝑢𝑟𝑖,𝑡 +

∆𝐹𝐷𝑖,𝑡 + ∆𝐺𝑜𝑜𝑑𝑠𝑖,𝑡 + 𝜀𝑖,𝑡 (2)

Where we simply apply the growth rates of the convergence for labour; the
unemployment rate, capital; the Financial Development Index and the growth of the
proportion of Intra-EU trade out of total EU trade to control for goods.

The second variation applies the Fiscal Policy:

∆𝑦𝑖,𝑡 = 𝑦0 + ∆𝑘𝑖,𝑡 + 𝜆𝐸𝑈 + 𝜆𝐸𝑢𝑟𝑜𝑧𝑜𝑛𝑒 + 𝜆𝑆𝑐ℎ𝑒𝑛𝑔𝑒𝑛 + ∆𝑃𝑟𝑖𝑐𝑒𝑠𝑖,𝑡 + ∆𝐷𝑒𝑏𝑡𝑖,𝑡 +

∆𝐷𝑒𝑓𝑖𝑐𝑖𝑡𝑖,𝑡 + ∆𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑅𝑎𝑡𝑒𝑠𝑖,𝑡 + 𝜀𝑖,𝑡 (3)

Again, the regression remains the same, instead we allow the growth in convergence
for the government debt to GDP, government deficit to GDP and the Long-Run
Interest Rates. Although price stability exists within the Convergence Criteria, we will
use across both regressions as this variable does not have much direct inference to the
other fiscal measures but indirectly impacts the Single Market variables and the Fiscal
policy measures so must be included in both.

By separating the regressions by policy type we will get a picture of how the
convergence of different European policy measures has affected, if at all, the growth
of these countries. To reiterate the point, European Integration in these models is
given by converging to the average rate, noting that these results are not to say that for
instance “Higher Debt to GDP implies a coefficient of X on growth”. Instead the
results will show that converging to the EU average will have an effect on GDP per
capita. This is an important distinction to make as to not mis-interpret the results.

17
 The econometric model used is the Weight Least Squares Panel Model. There is no
specific paper this model will follow, as the measure of European Integration is an
extension of König and Ohr (2013) who do not apply regression analysis and so
interpretation of European Integration is unique, at least with these variables. The
model also allows for time dummies.

4. Data Description
The data for this test is used in the time period 1996-2016 for all EU28 members apart from:
Croatia, Estonia, Luxembourg and Malta due to severe missing data. The data descriptions
will be comprised in Table 5 of Appendix A.
As this model uses data over several years, we need to ensure that the real values of variables
are used to exclude the offset effect caused by inflation. This issue is of particular concern for
the variables in Table 5, which is why the constant PPPs is used for GDP per Capita Growth,
and then using Chain linked volumes for the components. When considering the variables
used to measure integration, particularly the Single Market Index, is not the perfect measures.
Unemployment was used as a substitute to directly measure the labour market flows between
countries but due to the constraint, this measure from AMECO is the strongest alternative.
Similarly, the Financial Development Index does not capture the flows of Capital within the
EU perfectly, and ideally this data would capture the relative importance of interactions with
the rest of the EU. Although, both these measures show European Integration as the result of
being members of the Single Market. Following the economic rationale, it should follow that
Unemployment and Financial Development should increase with membership of the Single
Market, but not all countries will have the same gain and so specific labour and capital
measures all for the heterogenous relationship of a country and the Single Market. The Fiscal
Convergence measures are the specified measures stated in the Convergence Criteria so the
selection process for that index was straightforward. The dummy variables show the
difference in membership option and are essential to generate an accurate Index for European
Integration and can be found simply by reading through the EU website to determine the year
a country joined the bloc with dates also for Eurozone and Schengen.

The data is selected for these years firstly to consider how the Expansion countries have
interacted within the European Union in terms of the Single Market. Then secondly to see

18
how convergence in fundamental Fiscal variables affects growth, given the Financial Crisis
and Sovereign Debt Crisis both exist during this time frame, it will be interesting to see if
convergence has any effect on the growth. This is because the EU has an idea for a Fiscal
Union in the future, and assuming a legitimate Fiscal Union will converge these variables,
then we can infer how the Fiscal Union will affect growth from this model. This model won’t
capture how monetary and fiscal policies will interact which may caveat any implication as
no policy is used in isolation.

5. Results
This section of the paper will discuss the results of the regressions stated in CH4, and then
describe the economic implications. After there will be a brief comparison to the results in
the found in the existing Literature, and then any caveats in the analysis will be further
explored.
Table 3 displays the results for five regressions. Columns 1, 2 and 3 display the pooled and
fixed effects models, for the both the single market and fiscal variations. After seeing these
models fail, we instead apply the Weighted Least Squares approach, displayed with both sets
of controls in Columns 4 and 5, which will be the points of reference hereafter.
Observing Table 3 displays particularly interesting results. The immediate point of note is
that none of the Fiscal Policy convergences show any significance whilst the three Single
Market measures show varying levels of significance. This does not mean that Fiscal policy
does not have any effect on the GDP per capita of a country; instead it indicates that there is
not clear reasoning to say that converging to the European average for Debt and Deficit to
GDP and Interest Rates has a sufficient effect on the economic growth. Conversely, we find
that convergence to the European average in regard to Unemployment, Financial
Development and Share of EU trade are significant. Column 4 of Table 3 also shows
significance in relation to membership options within the EU. In both models, percentage
change in Price convergence yield insignificant results.
Controlling for the Single Market leads to a positive significant result for the European Union
membership, whilst negative coefficients for Eurozone and Schengen membership at the
ninety-nine and ninety-five per cent confidence level respectively. The immediate
implications of these results show that countries that have been a member of the European
Union over the twenty-one-year period from 1996-2016 experience an overall positive

19
 Table 3 – Panel Estimates of Economic Growth
OLS Pooled Fixed Effects Fixed Effects WLS WLS
Variables (1) (2) (3) (4) (5)

Constant 2786.66* 2944.12** 459.25 629.297 −511.78

(1554.5) (1466.33) (1083.95) (853.777) (643.7)
Initial GDP −0.1053 10-3*** 0. 9740 10-4*** −9.36791e-05
(1.73E-05) (1.12E-050) (8.86E-06)
∆K −0.7292 10-2*** −0.6739 10-2*** 0.1589*** −0.5708 10-2*** 0.1847***
(0.9187 10-3) (0.8949 10-3) (0.01255) (1.45E-03) (1.09E-02)
EU 0.6852 0.6075 −0.5491 0.6021* −0.5600
(0.501) (0.5261) (0.6366) (0.3637) (0.5281)
Eurozone -0.4935 −1.660*** −1.5234*** −0.6250*** −0.4745***
(0.3028) (0.5066) (0.4574) (0.1741) (0.1297)
Schengen −0.9541*** −1.161** −0.7906** −0.5189** −0.2679
(0.3247) (0.4931) (0.4009) (0.2326) (0.1744)
∆Labour 0.9172 10-2*** 0.7114 10-2*** 0.7762 10-2***
(0.2197 10-2) (0.2152 10-2) (0.2310 10-2)
∆Capital −1.83377e-05** −1.66251e-05** −1.38626e-05**
(7.20E-06) (7.07E-06) (6.67E-06)
∆Goods 0.0706 0.0642 0.07332*
(0.0429) (0.0412) (0.03811)
-2 -2 -2
∆Prices −0.3490 10 −0.3851 10 0.4273 10 −0.3007 10-2 −0.2985 10-3
-2 -2 -2 -2
(0.3554 10 ) (0.3412 10 ) (0.3082 10 ) (0.2523 10 ) (0.2158 10-2)
∆Debt −0.4838 10-3 0.7578 10-3
(0.2891 10-2) (0.1655 10-2)
-3
∆Deficit 0.8657 10 0.1980 10-3
(0.7290 10-3) (0.4878 10-3)
-2
∆Interest Rates 0.1295 10 −0.1406 10-2
(0.2786 10-2) (0.2264 10-2)

Observations 433 433 370 433 370

R-squared 0.5508 0.5376 0.7224 0.6620 0.8230
Number of
country 24 24 24 24 24
H 0.000 1.10E-283 0.000
D-W 0.00593347 1.000
Pesaran 0.222 0.416 0.324 0.00635 0.00585
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
The tests are also recorded for the p-values where; H is the groupwise Heteroscedasticity test, D-W is the Durbin-Watson test, and Pesaran measures the
Cross-Sectional Dependence.

growth effect, to typically contribute 0.602% of the per capita GDP growth. This paper
references European Integration and considering the theory, the EU is currently in the fifth
step of Balassa’s Regional Integration process, discussed in the introduction (pp.5), which
represents Monetary Union. Testing for Eurozone membership yields a negative coefficient

20
of 0.62%, whilst also being in the Schengen will contribute another negative 0.52% to
economic growth. When we consider a country that is a member of all three memberships,
whilst controlling for Single Market convergence, the net effect of membership is the sum of
the coefficients, which in this case approximates to a negative effect of 0.54% on economic
growth. Do these results indicate that the European Central Bank should close, and the
current Schengen countries should start “building walls”? Not necessarily. All that we can
infer from these results is that European Union Countries who are part of the Eurozone and
Schengen Zone experienced an overall negative effect to per capita GDP growth when
accounting for the Single Market heterogeneities. Another way of analysing is to group the
countries based on their positions within the bloc; see Table 4 in Appendix A.
If we distinguish these countries in the sample who joined the EU in prior to 2004 and
thereafter; it is clear that ten of the EU14 members have “complete membership”, whilst only
four of the Expansion countries are included as complete. Noting also that half of the
Expansion members are not yet Eurozone members, whilst considering that these countries
were poorer before they joined, they are likely to converge to the incumbent member’s
growth rates. This convergence relationship exists for the Single Market measure: intra-EU
trade in goods, as the coefficient in Column 4 of Table 3 attributes a 0.07% boost to per
capita GDP growth for a 1% increase in European Trade reliance.
Following Solow’s (1956) predictions regarding convergence to the steady state of growth,
this result is not too surprising. We cannot say definitively that country’s that are in the
Eurozone and Schengen Area are richer, but we can say that this model accounts for the
following: in 2000, pre-accession of the Expansion countries, none of these countries were in
the Schengen or Euro Area and so any growth effects from membership is likely to coincide
with the fact that the richer countries were measured. This is an important consideration as
the more developed a country is, the more difficult it is to sustain higher levels of economic
growth. In analogy, France would find it more challenging to grow at the same pace as say
Slovakia due to the starting point of the analysis, whereby in 1996 France’s resources are
used more efficiently than Slovakia. Development of the Slovakian economy leads to capital
and labour being used more efficiently and the hence a higher return to growth than that of
France. This model does appear to capture this effect.
All of that being said, we cannot disregard the Eurozone and Schengen Area as having no
effect on GDP per capita growth. De Grauwe (2013) discusses the limitations of having a
monetary union without a fiscal union, and this paper covers both the Financial Crisis and the

21
Sovereign Debt Crisis, where there was evidence of a prolonged negative growth for some
Eurozone countries.
Returning to the Single Market coefficients, it is clear to see that converging to the average
European Unemployment rate boosts growth, whilst converging to the median Financial
Development score reduces the growth, albeit by a very small proportion. The inference we
can take with these results is that converging to the European average unemployment change
helps economic growth whilst diverging from the average Financial Development change
indicates growth. These results are very small in the context of growth but still significant.
The Economic implications of Column 5 of Table 5 certainly are limited, the only significant
result relating to the European Integration displays a negative coefficient for Eurozone
countries. As stated above, this is not surprising given the issues Eurozone countries have
experienced, namely the PIIGS who are all complete EU members. The convergence in fiscal
measures are all insignificant for the growth and so using these as controls for heterogeneity
limit the analysis and have no statistical inference.
To compare these results to the Literature we must allow for the fact there is no direct
comparison to this paper. Alas, the Literature does typically find that the European
Integration, however it be defined, has a positive effect on the economic growth of the
European Nations. For the papers who use trade as a proxy, we must compare the total effect
but also consider the Single Market Model. Henrekson et al found that being in the EC/EFTA
increased growth, Badinger (2001) also found that regional integration boosted the growth,
mainly through tariff liberalisation and Mann (2015) achieved evidence for medium-run
growth effects. Direct comparisons to these papers are not conclusive but we do find that
when accounting for trade, European Integration has a positive effect for economic growth.
Badinger(2001) justifies this through tariff analysis, but given the difference in sample years
selected between his and this paper, and relatively how there has been little change in tariffs
in the years of this paper, a more appropriate control could be some globalisation measure. In
a sense this would aim to capture “Global Integration” opposed to “Regional Integration” and
has scope for further research. Other groupings in the literature overlap in the sense of the
more sophisticated model that accounts for regression analysis in terms of economic growth,
apply their models to countries to a sample of countries that have joined the European Union.
Similarly, to Campos et al (2013), this paper finds a positive growth effect for the Expansion
countries, if we use the Eurozone and Schengen acting as proxy for incumbent and expansion
counties, although at a lower rate compared to their Difference in Difference model. Rapacki
and Próchniak (2009) focus more on the convergence of the CEE10 countries to the EU
22
average growth and make predictions when the expansion countries will converge to the EU
level of growth. This model does not make those same predictions but supports the growth
effects from convergence within the Single Market.
The frustrating limitation of this analysis is the lack of data. I believe that the intra-factor
flows within the European Union could certainly better capture single market integration, and
the limitations of data particularly for the Expansion countries does caveat the analysis.
It has also been difficult to isolate the effects of vertical and horizontal integration, namely
due to the two not being mutually exclusive. Other Econometric issues include using the
median deviation, it is robust to outliers but arguments supporting mean deviation are
abundant. However, this analysis does contribute to the literature in terms of reassessing the
measures of Regional Integration, and despite the fiscal measures having no significance in
this case; it certainly acts as a base for further research.

6. Conclusion
This paper aims to test the hypothesis that European Integration has no effect in terms of
economic growth. The purpose of answering this question is to see if the economic and
political coordination has actually increased the European citizen’s standard of living. To
measure European Integration this paper defines the European Integration in terms of
Membership status, Fiscal Convergence and Single Market Integration. The results show that
European Integration, when accounting for Single Market Integration, has boosted the
standard of living for EU members. Interestingly the integration does have negative
coefficients for deeper membership for Eurozone and Schengen area countries. Why these
results are important when considering European Integration? Looking at Tables 1 and 2 we
see that Europe is not yet fully integrated, and so a smaller or negative return to growth could
mean that further integration could have minimal effects for economic growth, or in other
words, growth maximisation could be achieved at a lower stage of Balassa’s Theory.
Following on from this potential conclusion, we could see that a Fiscal Union may have no
effect on the economic growth of European Countries. This follows from Column 5 of Table
3; Fiscal Convergence has no significance in the Growth Rates. This anticipation is not
conclusive as this model does not consider tax integration, which of course is part of fiscal
policy. There is evidence to in this paper to support the idea that Eurozone membership
decreases economic growth by 0.62% and a justification for this could be found by De
Grauwe (2013), who argues “the Euro is a currency without a country”, and a Fiscal Union

23
could have reduced the negative effects of the Eurozone Crisis. If this analysis is accurate,
then the coefficients could have been different had there been an efficient Political Union.

This paper adds to the Literature by first adding another dimension to European Integration
by considering the Fiscal Convergence as a policy measure of the European Union. There are
of course improvements that could be explored. Perhaps the economic growth approach could
look at the fundamental growth variables7 with other controls variables to see what in
particular is driving the GDP per capita growth. Mentioned previously the data constrains
provide difficulty in isolating European Single Market Integration and so this could be
explored with more reliable data in the future.

The take-home message of this paper, what we can say definitively, is that the European
Union at this in-between stage of Economic Integration is not the most efficient and is
probably justification for the plans of the Five Presidents to further integrate. This paper
simply describes the convergence of European policies, and the effect on the economic
growth. To fully discuss European Integration and the European Union, there needs to be
discussion on the Political Integration of the Union to supplement the economic analysis,
with the aim of answering the key questions that are driving the concerns within Europe, that
drove the Brexit Vote in 2016.

7
Fundamentals being Consumption, Investment, Government Expenditure and Trade.

24
References
i. Badinger, H. (2001). Growth Effects of Economic Integration - The Case of the
EU Member States (1950-2000). IEF Working Paper Nr. 40
ii. Balassa, B. (1963).European Integration: Problems and Issues. The American
Economic Review, Vol. 53, No. 2, Papers and Proceedings of the Seventy-Fifth
Annual Meeting of the American Economic Association (May, 1963), pp. 175-
184
iii. Campos, N; Coricelli, F; Moretti, L.(2014). Economic Growth and Political
Integration: Estimating the Benefits from Membership in the European Union
Using the Synthetic Counterfactuals Method. IZA Discussion Papers, No. 8162,
Institute for the Study of Labor (IZA), Bonn
iv. Crafts, N. (2016). The Growth Effects of EU Membership for the UK: Review of
the Evidence. Social Market Foundation, Global Perspectives Series: Paper 7.
v. Deardoff, A; Stern, R. (2002). EU Expansion and EU Growth. School of Public
Policy. The University of Michigan Ann Arbor, Michigan 48109-1220,
Discussion Paper No. 487
vi. De Grauwe, P. (2006). What Have we Learnt about Monetary Integration since
the Maastricht Treaty. JCMS 2006 Volume 44. Number 4. pp. 711–30
vii. De Grauwe, P. (2013). Design Failures in the Eurozone: Can they be fixed?. LSE
‘Europe in Question’ Discussion Paper Series, LEQS Paper No. 57/2013
viii. Henrekson, M; Torstensson, J; Torstensson, R. (1997). Growth Effects of
European Integration. JEL F15, F43.
ix. König, J; Ohr, R. (2013). Different Efforts in European Economic Integration:
Implications of the EU Index. CMS 2013 Volume 51. Number 6. pp. 1074–1090
x. Luque, J; Taamouti, A. (2013). Did the Euro Change the Effect of Fundamentals
on Growth and Uncertainty. UC3M Working papers, Economics, 12-21.
xi. Mann, K. (2015). The EU, a Growth Engine? The Impact of European Integration
on Economic Growth in Central Eastern Europe. FIW Working Paper, No. 136,
FIW - Research Centre International Economics, Vienna.
xii. Rapacki, R; Próchniak, M. (2009). The EU Enlargement and Economic Growth In
the CEE New Member Countries. European Commission, Economic Papers 367.
xiii. Solow, R. (1956). A Contribution to the Theory of Economic Growth. The
Quarterly Journal of Economics, Vol. 70, No. 1. (Feb., 1956), pp. 65-94.
xiv. Svirydzenka (2016). Introducing a New Broad-Based Index of Financial
Development. IMF Working Papers.

25
 Appendix A

Table 4 – Country Memberships

Country EU / EC Euro Area Schengen Area
Austria 1995 1999 1997
Belgium 1957 1999 1995
Bulgaria 2007 X X
Cyprus 2004 2008 X
Czech Republic 2004 X 2007
Denmark 1973 X 2001
Finland 1995 1999 2001
France 1957 1999 1995
Germany 1957 1999 1995
Greece 1981 2001 2000
Hungary 2004 X 2007
Ireland 1973 1999 X
Italy 1957 1999 1997
Latvia 2004 2014 2007
Lithuania 2004 2015 2007
Netherlands 1957 1999 1995
Poland 2004 X 2007
Portugal 1986 1999 1995
Romania 2007 X X
Slovakia 2004 2009 2007
Slovenia 2004 2007 2007
Spain 1986 1999 1995
Sweden 1995 X 2001
United Kingdom 1973 X X

Table 4 shows the years a Country joined various membership options within the current EU
framework. An “X” indicating the country is not

26
 Table 5 - Data Descriptions
Variable Technical Measure Source Date Extracted
Labour Unemployment rate: total :- Member AMECO 22.03.19
States: definition EUROSTAT
(ZUTN), % of total population

FD Financial Development Index IMF 29.03.2019

Goods Imports / exports of goods, Millions IMF 25.03.2019
US$. final calculation = EU / total
trade

Prices All items HICP, annual average rate of Eurostat 24.03.2019

change, %

Long Term EMU Convergence Criteria Bond Eurostat 23.03.2019

Interest Rates Yields (10 Year)
Deficit General Government, net lending and Eurostat 23.03.2019
borrowing, % of GDP

Debt General Government Debt Total, % of Eurostat 23.03.2019

GDP

GDP per Capita Gross Domestic Product (Expenditure OECD 20.03.2019

Approach), Per head, US $ constant
prices, constant PPPs, OECD base
year (2010)

K Gross Fixed Capital Formation, Chain Eurostat 20.03.2019

linked volumes (2010), million euro

27