Fostering Growth: The EU as an Economic Engine. A Case Study on Former Jugoslavian Republics
Advanced Econometrics
Abstract
In this brief paper, I propose an analysis of the impact of joining the European Union on economic growth, presenting the case of Slovenia and Croatia, whose trajectory has been different as the former entered the EU nine years before the latter. Thanks to the Fixed Effects Estimation approach, I will check whether being in the EU affects some indicators such as productivity, employment and hours worked. By browsing regional data provided on the ARDECO database, I will subsequently try to assess last decades economic performance of the two countries, focusing on trends and growth rates, aiming at understanding whether Croatia lagged behind as Slovenia was benefiting of its status of EU member State.
Keywords
Slovenia, Croatia, European Union, economic integration, regional convergence, productivity, economic development, Fixed Effects Estimation
Presentation
Aiming at assessing the impact that joining the European Union and the Eurozone may have on economic performance, in this brief analysis I will try to do that by taking into consideration two of the current member countries: Croatia and Slovenia. These former Yugoslav Republics, located in the so-called Western Balkans, can act as an exemplar case study due to the fact that, despite being two really similar countries, one joined the EU nine years before the other: Slovenia did in 2004, whilst Croatia in 2013. Both have by now adopted the Euro as well, thus being part of the Euro Area: Slovenia since 2007, Croatia since 2023.
This allows us to carry out a study of the effects that this membership-gap had, by focusing on economic data for the period 2004/2024. It is worth remembering that unfortunately the time frame I was obliged to choose is not the best to assess economic growth, as Europe was hit by at least four major economic shocks: the 2007/2009 Great Financial Crisis, the subsequent 2011/2013 European Sovereign Debt Crisis followed by a very slow recovery, the 2020 COVID pandemic and finally the Ukrainian War which caused a spike of energy prices, the comeback of inflation in the Old Continent.
That said, there are some economists who have already been investigating this subject. Coricelli, Moretti and Campos studied whether the countries that joined the EU between 1973 and 2004 benefited from membership. Their work1 is particularly interesting as their findings - based on the difference-in-differences technique - look pretty strong, even if they diverged amongst member States, and positive, except for Greece, the only country that did not benefit from joining. In the case of Slovenia, the results are positive both for GDP per capita and labour productivity, but statistically significant only for the latter. Nevertheless, their research dates from 2014, so that it was not possible to assess Croatia, and they did not adopt a comparative approach, taking into account the eventual similarities between some member States. Tajnikar, Došenović Bonča and Rubinić, published a paper2 in which they specifically assess Eurozone’s influence on the cyclical macroeconomic performance through a comparison between Slovenia and Croatia. Mikulić, Lovrinčević and Nagyszombaty Galić published in 2013 an article3 concerning regional convergence in the EU compared to Croatia, arguing that this could not be as strong as we could think, as other local-based factors may dominate its effects thus determining economic growth according to the evidences they found thanks to a \(β\)-convergence analysis. Also, there have been multiple attempts of evaluating, at the NUTS 2 or 3 level, both economic convergence between EU regions, as in Postiglione, Cartone and Panzera 2020 paper4, as well as studies to assess the impact of EU development policies, such as Becker, Egger and von Ehrlich’s one5. The former find “a positive impact of EU transfers to disadvantaged regions on GDP per capita growth, but no employment growth effects” whilst the latter highlight how “regional economic β-convergence holds at the European level across the different periods and regions investigated” stressing on the “need for major integration between Western and Eastern European regions in order to support cohesion and reduce disparities in the whole EU”.
Case Study
It is not at all easy to determine whether - and how - joining the European Union can impact economic growth. First, because it is not always easy to identify what determines economic growth and prosperity in general. Furthermore, as reality provides us with only a few cases which could serve as samples to deepen. The Western Balkans one could be one of them, but we could handle it carefully anyway, acknowledging that two countries, despite their similarities, are still two different worlds.
In order to show how membership could affect economic issues, I suggest to start from investigating a DAG model such as the one presented below, in which I tried to identify the determinants of economic growth from their foundation. As we can see, even in a simple model, things become quickly complicated. But it turns out that two of the most important elements impacting growth are the currency and the institutions. Local institutions are influenced by national and international ones, notably in the case a country joins a union as integrated as the EU is. If the same country decided to join the Eurozone as well, also the currency factor would become almost completely dependent on EU issues.
To present an analysis as comprehensive as possible, I will be using regional data from the ARDECO database, provided by the European Union. They deliver panel data at the different NUTS levels, and even for European largest Metro regions. Since Croatia and Slovenia are two little countries, their NUTS 1 correspond to the entire countries, and they just have a total of six NUTS 2 regions. That is the reason why I decided to take NUTS 3 territories into account, this was more consistent with my idea of going more in-depth into the given data.
In the map below, we can take a look at NUTS 3 regions in Slovenia and Croatia.
To give some context, in 2003 Slovenia had a population of 2.000.000, and Croatia of 4.300.000. Today, Slovenia increased to 2.100.000, whilst Croatia decreased to 3.900.000. This different trend in population growth could be of interest. They were - and are - the richest countries in Western Balkans: the GDP per capita at current prices was of 7.205,97\(\text{€}\) in Croatia and of 13.151,78\(\text{€}\) in Slovenia; in 2012 this figure stood at 10.578,48\(\text{€}\) and 17.622,95\(\text{€}\) respectively. Concerning productivity, defined by Eurostat as the real GDP per hour worked, in 2003 it stood at 19,94\(\text{€}\) in Slovenia and at 13,38\(\text{€}\) in Croatia, increasing to 24,23 and 15 in 2012.
Estimating the Effects
As said, back in 2004, Slovenia joined the EU along with other nine Southern and Eastern European countries, whilst Croatia had to wait until 2013. Thanks to the fixest package, I am able to easily perform some Fixed Effects Estimation to evaluate whether the fact of being in the EU impacts some economic variables values and growth rates, accounting for between-variation through the fixed effect technique, thanks to which we are able to isolate all local stuff which would disturb our broader analysis.
I will investigate the magnitude of the lag that Croatia could have accumulated by staying out when Slovenia was already granted membership. In order to do that, in my regressions I will consider, as said, the time frame 2004/2024, because Fixed Effects Estimations should work better in accounting the effect of the dummy variable EU when applied to a period of time in which the latter is always active for at least one variable.
Labour Market and Productivity
I first performed a regression on GDP per capita at constant prices; in Model 1 I analyse whether GDP per capita \(log\) value has a relationship to the NUTS region being in the EU, whilst in Model 2 I look at annual percentage GDP per capita growth, that I calculated myself. I will proceed in the same way for all the next regressions. The result I cared about the most is that of growth, as we are interested in showing that the EU is an engine to growth, even for those less developed regions that join. The results are statistically significant, as shown by their low \(p\)-value, and the effect on GDP per capita growth looks pretty strong at 3,26, despite a very low \(R^2\), which is not necessarily a proof of model weakness, and a pretty high standard error.
| Model 1 | Model 2 | |
|---|---|---|
| EU | 0.13 *** | 3.26 *** |
| (0.01) | (0.34) | |
| nobs | 660 | 660 |
| r.squared | 0.93 | 0.06 |
| adj.r.squared | 0.93 | 0.01 |
| within.r.squared | 0.19 | 0.05 |
| pseudo.r.squared | ||
| sigma | 0.11 | 5.58 |
| nobs.1 | 660.00 | 660.00 |
| AIC | -1063.99 | 4175.24 |
| BIC | -911.26 | 4327.97 |
| logLik | 566.00 | -2053.62 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||
To have an idea of how growth varied between NUTS regions, graphs in Figure 1 show GDP per capita growth according to EU membership between 2004 and 2012, so that the points on the left represent Croat regions and those on the right Slovenian ones. Remembering that we are still talking about two similar, but distinct countries, we can see that variance looks narrower in NUTS regions in the EU. We also observe that this widens in periods of crisis, and notably during the European Sovereign Debt Crisis.
To investigate how the EU may account for this acceleration in economic growth, I first looked at some labour market figures, starting from the indicator “Wage and Salary Earners”, which is workplace based, so that it accounts for where jobs are actually located and being created. According to the model, regions within the EU would benefit from 8% more wage work (the low \(p\)-value and standard error, and the very high \(R^2\) leave little doubt about this), and the number of workplace would be growing 1,96 pp faster in the EU, even if, again, a high standard error could make us question it.
| Model 1 | Model 2 | |
|---|---|---|
| EU | 0.08 *** | 1.96 *** |
| (0.01) | (0.24) | |
| nobs | 660 | 660 |
| r.squared | 0.99 | 0.07 |
| adj.r.squared | 0.99 | 0.02 |
| within.r.squared | 0.17 | 0.04 |
| pseudo.r.squared | ||
| sigma | 0.07 | 3.71 |
| nobs.1 | 660.00 | 660.00 |
| AIC | -1616.58 | 3638.20 |
| BIC | -1463.85 | 3790.93 |
| logLik | 842.29 | -1785.10 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||
If firms hire more people, this could also have an impact on - or be due to a variation of - hours worked, at which we will take a look right now. Performing the same type of regression as before, we see that both coefficients are statistically significant and marked by incredibly low standard errors. According to the model, a person employed in a NUTS region in the EU works -4% hours compared to an employee in a territory outside of the EU. Furthermore, its hours worked grow at a pace 0,44 pp lower, actually even possibly decreasing. By looking at the database, we corroborate this thesis, as a decline occurred in 348 out of 660 observations, both in expansion and crisis moments of the business cycle.
| Model 1 | Model 2 | |
|---|---|---|
| EU | -0.04 *** | -0.44 *** |
| (0.00) | (0.02) | |
| nobs | 660 | 660 |
| r.squared | 0.92 | 0.02 |
| adj.r.squared | 0.91 | -0.03 |
| within.r.squared | 0.38 | 0.02 |
| pseudo.r.squared | ||
| sigma | 0.02 | 1.45 |
| nobs.1 | 660.00 | 660.00 |
| AIC | -3234.17 | 2393.80 |
| BIC | -3081.44 | 2546.54 |
| logLik | 1651.09 | -1162.90 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||
As we could think that worked hours could decrease due to an enhanced income effect (i.e. when their hourly wage increases, individuals are adept at choosing to renounce to an even higher salary to enjoy more free time), it is worth checking how the model predicts that its evolution vary depending on our EU dummy variable. As Croat data are still not available for 2023, in this case the analysis will be limited to the period 2004/2022.
Surprisingly, even if the figure is 13% higher for regions in the EU, its growth is apparently 0,44 pp lower than outside the EU. Again, the low \(p\)-value and standard error make these figures statistically significant. By looking closer at the database, only 101 out of the 660 observations considered recorded a decline: they concern almost exclusively Croatia, often in the period in which it had already joined the EU, but not only during clear crises. This matter would probably need to be investigated further.
| Model 1 | Model 2 | |
|---|---|---|
| EU | 0.13 *** | -0.73 *** |
| (0.01) | (0.13) | |
| nobs | 627 | 627 |
| r.squared | 0.87 | 0.03 |
| adj.r.squared | 0.86 | -0.02 |
| within.r.squared | 0.13 | 0.01 |
| pseudo.r.squared | ||
| sigma | 0.14 | 4.02 |
| nobs.1 | 627.00 | 627.00 |
| AIC | -696.77 | 3557.86 |
| BIC | -545.78 | 3708.86 |
| logLik | 382.38 | -1744.93 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||
Increased productivity is the most intuitive determinant of decreasing hours worked. As we can see, “Real Labour Productivity per Hour Worked” - which is equal to real GDP per hour worked - is higher in EU regions, and also grows at a faster pace. Nevertheless, the standard error is non-negligible in the case of the growth rate.
| Model 1 | Model 2 | |
|---|---|---|
| EU | 0.12 *** | 1.62 *** |
| (0.02) | (0.44) | |
| nobs | 660 | 660 |
| r.squared | 0.93 | 0.02 |
| adj.r.squared | 0.93 | -0.03 |
| within.r.squared | 0.25 | 0.01 |
| pseudo.r.squared | ||
| sigma | 0.09 | 5.51 |
| nobs.1 | 660.00 | 660.00 |
| AIC | -1336.97 | 4159.43 |
| BIC | -1184.23 | 4312.17 |
| logLik | 702.49 | -2045.72 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||
Another element whose impact on economic matters I would like to analyse is trade, which is facilitated by the EU both by being a customs union and by its numerous trade agreements with countries around the world.
Trade
As joining the EU could foster trade, both with other member States and with Extra-EU countries, I will use Eurostat data to analyse the impact of membership on this matter for Croatia and Slovenia.
In the first set of regressions, I look at exports and imports value within the EU (Models 1 and 2), in Extra-EU countries (Models 3 and 4) and in total (Models 5 and 6). Both exports and imports are dramatically higher when \(EU=1\), except for Extra-EU imports, which are only 1% higher.
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | |
|---|---|---|---|---|---|---|
| EU | 0.65 *** | 0.56 *** | 0.43 *** | 0.01 *** | 0.57 *** | 0.39 *** |
| (0.00) | (0.00) | (0.00) | (0.00) | (0.00) | (0.00) | |
| nobs | 40 | 40 | 40 | 40 | 40 | 40 |
| r.squared | 0.76 | 0.50 | 0.41 | 0.10 | 0.68 | 0.31 |
| adj.r.squared | 0.75 | 0.48 | 0.37 | 0.05 | 0.66 | 0.27 |
| within.r.squared | 0.34 | 0.38 | 0.12 | 0.00 | 0.25 | 0.15 |
| pseudo.r.squared | ||||||
| sigma | 0.33 | 0.26 | 0.43 | 0.55 | 0.36 | 0.34 |
| nobs.1 | 40.00 | 40.00 | 40.00 | 40.00 | 40.00 | 40.00 |
| AIC | 28.07 | 9.80 | 49.17 | 68.17 | 33.92 | 30.83 |
| BIC | 33.13 | 14.86 | 54.23 | 73.24 | 38.99 | 35.90 |
| logLik | -11.03 | -1.90 | -21.58 | -31.09 | -13.96 | -12.42 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||||||
For this second set, I calculated the part of trade within the EU for each country every year. In Model 1, we see that the part of exports to the EU increases with membership of about 5%. This gap is even more important for imports, at 11,7%, which could explain the little increase in imports from Extra-EU countries previously seen. In Model 3, we observe that, in general, by joining the EU the part of trade within the EU is around 9% higher.
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| EU | 5.16 *** | 11.72 *** | 9.06 *** |
| (0.00) | (0.00) | (0.00) | |
| nobs | 40 | 40 | 40 |
| r.squared | 0.67 | 0.26 | 0.34 |
| adj.r.squared | 0.65 | 0.22 | 0.31 |
| within.r.squared | 0.21 | 0.26 | 0.29 |
| pseudo.r.squared | |||
| sigma | 3.71 | 7.22 | 5.24 |
| nobs.1 | 40.00 | 40.00 | 40.00 |
| AIC | 221.33 | 274.49 | 248.97 |
| BIC | 226.39 | 279.56 | 254.04 |
| logLik | -107.66 | -134.25 | -121.48 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | |||
In this last set of Fixed Effects regressions, let us take a look at annual growth rates as well. Total imports would apparently grow way faster with EU membership, whilst the benefit on exports looks reduced. Imports grow more rapidly than exports inside the EU as well, but both figures are way higher than those for total. Focusing on Extra-EU trade, its growth is apparently impacted negatively, especially for what concerns exports.
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | |
|---|---|---|---|---|---|---|
| EU | 1.53 *** | 5.67 *** | 5.23 *** | 9.24 *** | -4.58 *** | -0.41 *** |
| (0.00) | (0.00) | (0.00) | (0.00) | (0.00) | (0.00) | |
| nobs | 40 | 40 | 40 | 40 | 40 | 40 |
| r.squared | 0.01 | 0.03 | 0.03 | 0.07 | 0.05 | 0.04 |
| adj.r.squared | -0.04 | -0.02 | -0.03 | 0.02 | -0.01 | -0.01 |
| within.r.squared | 0.00 | 0.02 | 0.03 | 0.07 | 0.02 | 0.00 |
| pseudo.r.squared | ||||||
| sigma | 11.65 | 13.59 | 11.72 | 12.31 | 13.54 | 21.07 |
| nobs.1 | 40.00 | 40.00 | 40.00 | 40.00 | 40.00 | 40.00 |
| AIC | 312.79 | 325.14 | 313.31 | 317.23 | 324.87 | 360.21 |
| BIC | 317.86 | 330.21 | 318.38 | 322.30 | 329.94 | 365.27 |
| logLik | -153.40 | -159.57 | -153.66 | -155.62 | -159.43 | -177.10 |
| *** p < 0.001; ** p < 0.01; * p < 0.05. | ||||||
As I also wanted to get an idea of how the trend for the part of trade within the EU evolves over time, I plotted the data in Figure 2. We see that the two countries have been having pretty different trajectories, as Slovenia is more of an export-oriented economy than Croatia, and developed important commercial flows with Switzerland in particular, in the sector of “Packaged Medicaments”, as measured by OEC. So, outside of the EU, but not that far from it. What is common to both countries, is the spike in the part of trade within the UE during the first year as members (it is also worth remembering that Croatia joined only in July, which makes its figure even more impressive). In the case of Croatia, we can observe a persisting upward trend.
In Figure 3, we clearly see how exports to the EU are boosted by membership, with Slovenia drastically accelerating starting in 2004, whilst Croatia lagged behind.
Just to play around with data, but also to try to account for the impact that joining later had on Croatia, I made the hypothesis that, by joining the EU in 2004, Croatia would have experienced exports growth rates similar to those of Slovenia. In Figure 4, we can take a look at this hypothesis. According to this simplistic model, in 2023 Croatia would have exported goods for around 5.000.000.000\(\text{€}\) more, which is for sure non-negligible.
As we saw, joining the EU can have a large impact on trade, which could itself foster economic growth.
Investigating the Effects: A Regional Perspective
In order to deepen this analysis, I will go over some of the figures provided by ARDECO at the regional level. As Croatia and Slovenia are two neighbouring countries, there are several regions which are located on the border, and it would be interesting to identify different trends in territories with a similar economic and industrial structure, as they could be due to exogenous factors, such as being part of the EU.
To do that, I will present several maps comparing both values and growth rates in the two countries’ regions in different moments and over multiple periods, focusing on the dates we care about the most as they are related to EU enlargements.
Labour Market
Starting with the labour market, I first decided to analyse the variation of hours worked per employed person between the different regions of the two countries.
In 2003, Slovenians worked way fewer hours per year compared to Croats. Whilst in Slovenia almost everyone worked the same amount of time, in Croatia, the hours worked varied more between regions. Those who had the longest working years were the people living in the NUTS HR050, corresponding to Zagreb, the capital city.
The situation evolves in 2012. Whereas in Slovenia, the hours worked continued decreasing, with no region’s average citizen working more than 1.656 hours a year, in Croatia they increased in all the inner regions, decreasing only in the capital city and on the coasts.
In 2023, we see that the two countries still have a significantly different situation, as Croats work 200 hours more per year than Slovenians.
Focusing on growth rates, between 2003 and 2023 the hours worked decreased faster in Slovenia than in Croatia, and notably than its inner regions.
As we can see, in the period 2003/2012, the hours worked notably decreased in Slovenia, whilst they increased in many Croat regions (the inner ones in particular) or were stable (on the coasts), except in the capital city.
By looking at 2012/2023, we see that in Croatia there was an increase much stronger than in Slovenia. This could be due to the effects of having joined the EU.
To provide some more context, I will now present some maps showing which sectors are the most significant (the first, the second and the third employing the most people as a percentage of all the employed) in each region, and how this evolved over time.
These are the ten NACE sectors used by ARDECO: A “Agriculture, Forestry and Fishing”, B-E “Industry”, F “Construction”, G-I “Wholesale Retail, Transport, Accommodation & Food Services”, J “Information and Communication”, K “Financial and Insurance Activities”, L “Real Estate Activities”, M-N “Professional, Scientific and Technical Activities; Administrative and Support Services”, O-Q “Public Administration, Defence, Education, Human Health and Social Work Activities”, R-U “Arts, Entertainment and Recreation; Other Services; Activities of Household and Extra-territorial Organisations and Bodies”.
In the first map, we can look at the sectors that dominated each region’s economy in 2003. Many Croat regions still principally relied on the primary sector (i.e. agriculture) with three territories in which it accounted for more than 40% of the employment. Almost all Slovenian regions were predominantly industrial. Both capital cities NUTS are more tertiary-oriented, as well as most of the coastal regions of the two countries, probably also thanks to activities such as tourism.
For what concerns the second principal economic sector in 2003, it is that of the Public Administration in both capital cities, as well as in some other regions across the two countries. We find the primary and secondary sectors in some Slovenian regions, the services in all the others. In Croatia, there are much more regions in which O-Q stands at the second place.
For the third places, we find agriculture in three Croat regions, industry in both capital cities and a lot of O-Q in Slovenia, and of G-I in Croatia.
In 2012, many inner and one coastal Croat regions still rely on agriculture as their main economic activity, even if the percentages decreased a bit. In Slovenia, industry dominates in all the regions (with percentages of over 30% in eight of them), but that of the capital city, where now O-Q is ahead of the services, that held in Zagreb and its surroundings. All the other coastal regions are tertiary-oriented.
Concerning the second places, services dominate in Slovenia, including in the capital city, but in one agricultural and three Public Administration regions. O-Q plays an important role in many Croat regions as well, leading in Zagreb as well. The other Croat regions are mostly secondary-based, except for five G-I.
In third places, mostly O-Q in Slovenia and a new entry in Ljubljana (M-N); a more mixed situation in Croatia, where we find both industrial, O-Q and agricultural regions.
In 2022 (last available data), industry is still in the lead in Slovenia, and the situation changes in Croatia, where there are almost no more predominantly agricultural regions: G-I dominates the coastal ones, whilst the industry conquered some of the inner areas. Both capital cities are again mostly services-oriented.
For the second places, we have by now mostly O-Q in Croatia, and a mix of O-Q and tertiary in Slovenia. There are still three agricultural regions in Croatia.
Moving to the third places, mostly O-Q in Slovenia (M-N in the capital city and in the Istrian region close to Trieste), and a mix of services and industry (including in Zagreb) in Croatia.
In general, Croatia moved from being predominantly agricultural to industrial, and this process has accelerated since it joined the EU in 2013. We had the confirmation that Croat regions lagging behind in terms of richness and productivity were the inner ones, that still had to be industrialized.
On the other hand, Slovenia has a strong industrial base, to which probably is due the growth experienced by the country in the last decades. This also relates to its economy being an export-oriented one.
Productivity and Economic Growth
Concerning productivity, I will consider real GDP per hour worked, as before. I will proceed by looking at regional values in 2003, 2012 and 2023, to then analyse productivity growth over the same three time periods as above.
Getting back to 2003, we can see that Slovenia - and notably its Western regions - was much more productive than Croatia, whose most productive regions are as well located on the coasts, in the North-West.
The gap widens in 2012, as all Slovenian regions experienced a pretty strong growth, whilst some of the Croat ones actually decreased their productivity.
As Slovenia continued growing, there is still an important difference between the two. But we clearly see that Croatia’s productivity has increased since the country joined the EU.
In the period 2003/2023, most Slovenian regions increased their productivity much faster than Croat ones, where only the NUTS HR065 - which surrounds Zagreb - experienced a terrific growth.
Between 2003 and 2012, Slovenia’s real GDP per hour worked increased double-digit everywhere, whilst some Croat regions stagnated or actually even decreased their productivity. Probably the country outside of the EU suffered more from the crisis.
Focusing on the period 2012/2023, we can see an important increase in Croat productivity, as the country - albeit with some exceptions - accelerated compared to Slovenia. The impact of joining the EU looks positive.
Now, it is worth analysing whether this increase in productivity positively impacted economic growth.
In 2003, Croatia clearly lagged behind Slovenia in terms of GDP per capita at constant prices. Zagreb - and the two coastal NUTS HR031 and HR036 - was the only region that could compete with Slovenian ones.
Nine years later, Slovenia still held its advantage. It grew, but was probably also limited by the economic turmoil of those years. On the other hand, most Croat regions’ GDP per capita stagnated or even decreased.
Compared to 2012, in 2023 Slovenia is still in the lead, but Croatia is in better shape, as its capital city became the region with the second-highest GDP per capita after the Slovenian SI041 one (the most populous of the country as well as that where the capital city Ljubljana is located).
Focusing on growth rates, we can see that over the period 2003/2023, many Croat regions recorded a very strong increase in GDP per capita, and Slovenian ones grew at important rates as well. It is worth mentioning one more time that some of the highest increases concern territories that lagged behind, for which the hypothesis of a catch-up effect holds.
Between 2003 and 2012, whilst all Slovenian regions grew, some of the Croat ones recorded strong increases, but others became poorer. Again, this could also be due to the negative effects of the business cycle (i.e. subsequent recessions in those years). As shown in Figure 1, apparently inside the EU growth tends to be much stable and less subject to large territorial variations.
Looking at the period 2012/2023, we can see how Croatia clearly accelerated, as many regions recorded very high growth rates. This concerned especially the poorer inner and Southern ones, but also the more developed ones grew at important rates, including the capital city gaining 33,59% in GDP per capita during those years.
By looking at these figures, we can make the hypothesis that joining the EU may have a positive impact on economic growth.
Conclusions
Thanks to this brief analysis, we saw that joining the EU can provide some benefits to countries. Nevertheless, we should keep in mind some elements: first, for countries to join the EU, a certain level of development is implicitly demanded, so that this could be the reason for which the values for certain of the variables we studied were slightly higher for regions in the EU.
Also, we should not forget that some of the EU laws and customs could actually negatively impact economic growth. Think about Maastricht fiscal rules, or the Stability and Growth Pact. By joining the Eurozone, countries lose as well control of their monetary policy, which could affect their ability to act and thus eventually be even detrimental to their economic performance in some cases, notably in the case of States used to devalue their currency to support and foster exports. On the other hand, this can also provide them with much more monetary stability, which can be helpful in containing both bond rates and inflation.
The catch-up effect may have a role as well in determining economic growth in the specific case I analysed, as, despite being the two most developed countries of the Western Balkans, Slovenia was in advance compared to Croatia, even before joining the EU. This could explain Croatia’s high growth at the start of the 2000s, and it could be a bias in the context of this assessment.
Footnotes
TAJNIKAR, Maks; DOŠENOVIĆ BONČA, Petra; RUBINIĆ, Ivan. The Eurozone’s Influence on the Cyclical Macroeconomic Performance: Comparative Analysis of Slovenia and Croatia (2023)↩︎
TAJNIKAR, Maks; DOŠENOVIĆ BONČA, Petra; RUBINIĆ, Ivan. The Eurozone’s Influence on the Cyclical Macroeconomic Performance: Comparative Analysis of Slovenia and Croatia (2023)↩︎
MIKULIĆ, Davor; LOVRINČEVIĆ, Željko; NAGYSZOMBATY GALIĆ, Andrea. Regional Convergence in the European Union, New Member States and Croatia (2013)↩︎
POSTIGLIONE, Paolo; CARTONE, Alfredo; PANZERA, Domenico. Economic Convergence in EU NUTS 3 Regions: A Spatial Econometric Perspective (2020)↩︎
BECKER, Sascha O.; EGGER, Peter H.; VON EHRLICH, Maximilian. Going NUTS: The Effect of EU Structural Funds on Regional Performance (2010)↩︎