Czech Republic
7.1.8 Robustness check
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
7.1.7 Investment per capita
Lastly, SCM was also built for investment per capita. However, as can be seen from the graph below, the fit of the pre-treatment period is bad and therefore, the results cannot be properly interpreted. The synthetic investment per capita consists of 0.557 Korea, 0.192 New Zealand, 0.143 Malaysia, 0.061 Albania and 0.047 Singapore.
Figure 13: Comparison of investment per capita, Czech Republic
Source: Author’s calculations
7.1.8 Robustness check
To make sure the results of the estimates are robust, three robustness checks were carried for each of the outcome variables. The estimated effects on the country are robust if the estimated synthetic outcome variable by the robustness check follows similar pattern as the SCM estimates. Below graphs present the robustness results for the Czech Republic. The robustness checks show similar trend for the outcome variables. Overall, the results seem to be robust for all variables except for net export and investment.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 14: Robustness check, Czech Republic
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
Source: Author’s calculations
Note: Synthetic CZ (add) refers to robustness check when Iceland, Norway and Switzerland are added to the donor pool. Synthetic CZ (leave-one-out) refers to robustness check when the country with highest assigned value in the weighted average is dropped from the dataset. Synthetic CZ (different donor pool) refers to robustness check when larger donor pool consisting of 37 countries is used.
The SCM analysis results for the Czech Republic showed that the GDP per capita was positively influenced by the EU membership. The results were fairly robust for most of the outcome variables with the exception of investment and net export. The obtained p-values for the GDP per capita were statistically significant for the first six years of the EU membership, which shows strong short-term impact of the EU accession on the Czech income per capita level.
Furthermore, the analysis showed the main driving force of the growth was export, which experienced statistically significant increase over the studied period. Due to the economic integration, the post-accession GDP per capita was higher on average by 1 415 USD. The EU membership also positively influenced the import per capita that remained statistically significant for most of the post-treatment years. Nonetheless, Czech Republic achieved during the whole post-treatment period positive net export per capita39, thus the net export positively influenced the output growth.
The government spending and private consumption per capita would be slightly higher without the EU membership, nonetheless, the p-values were not statistically
39 The results for the net export per capita were not fully robust and the effect may have been overestimated.
‧
significant for either of the components. This shows that the EU participation did not significantly influence the Czech government spending or private consumption per capita. In year 2011, the Czech government introduced several government spending cuts that affected the both above-mentioned GDP components. The Czech government aimed to lower its budget deficit by cutting on social support and salaries of the constitutional officers (Government of the Czech Republic, 2010).
For investment per capita, the trend of the graph showed possible negative effect of the EU accession on the investment, nonetheless, the pre-intervention period for the variable was not matched well and therefore, the results were not interpreted.
To sum up, we can conclude that the EU membership did not significantly influence the rest of the GDP component variables. The obtained results from the analysis are similar to the ones reported by Campos et al. (2014), the authors found positive insignificant effect of the EU membership on real Czech GDP per capita.
Furthermore, the Czech Republic also maintained good public debt results and avoided high level of inflation (Molendowski, 2015). Czech Republic is one the top exporters in the EU and after joining the EU, the welfare of the Czech citizens and the quality of their life increased (Jedlička et al., 2014). Overall, the EU membership was positive for the Czech Republic and boosted up the GDP growth.
Hungary
Just as the Czech Republic, Hungary has data available from year 1991. The pre-treatment and post-treatment periods remain the same, 13 and 14 years, respectively. The preliminary analysis showed nonuniformed results, therefore, we expect the SCM results for Hungary to be mostly small or negative.
Unlike the previous case, the results for Hungary show mostly negative impacts on the Hungarian economy. Furthermore, the estimated effects on the Hungarian GDP per capita were statistically significant for years 2012 and 2013. Private consumption, government consumption and investment per capita also experienced a decline after the EU accession. The only exception is net export that significantly increased after joining the EU. The match of the synthetic Hungarian imports and exports in the pre-intervention period was insufficient, thus the results provided cannot be used for further analysis.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
7.2.1 GDP per capita
The below graph presents the GDP per capita for Hungary and synthetic Hungary that have never joined the EU. The pre-treatment period is matched well. The results are contrary to those obtained for the previous country. The graphs show that if Hungary decided not to join the EU, the Hungarian GDP per capita would be higher than it currently is. The synthetic Hungary consists of 0.439 Korea, 0.365 Russia, 0.11 Albania and 0.085 Australia.
Figure 15: Comparison of GDP per capita, Hungary
Source: Author’s calculations
The obtained p-values were statistically significant on 10% level for years 2012 and 2013. For other years, the results were statistically insignificant. The negative effect was experienced one year after the EU accession, in year 2005. The results show that if Hungary did not access the EU, in 2017, the GDP per capita would be higher by 2 898 USD. Due to the EU accession, the post-treatment income per capita was on average lower by 2 869 USD. The negative effect was expected from the preliminary results that showed a drop in the Hungarian GDP per capita after the EU accession.
Below analysis of the GDP components will provide more information to describe the possible causes of the negative impacts of the EU membership on the output.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
7.2.2 Export per capita
For export per capita, the pre-intervention period was matched badly. Therefore, the results were impacted, and significant values were not reached. In this case, synthetic Hungary consists of 0.696 Korea and 0.304 Malaysia. However, the below graph demonstrates possible positive effect of EU membership on the Hungarian exports.
Figure 16: Comparison of export per capita, Hungary
Source: Author’s calculations
7.2.3 Import per capita
Below graph shows estimates for the import per capita. The synthetic Hungary consists of 0.607 Korea, 0.341 Belarus and 0.052 Singapore. Just as for export, import per capita was not perfectly matched and therefore, the retrieved p-values were statistically insignificant, and the results cannot be interpreted.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
Figure 17: Comparison of import per capita, Hungary
Source: Author’s calculations
7.2.4 Net export per capita
Unlike export and import, net export per capita offers better fit in the pre-accession period. In this case, the synthetic Hungary consists of 0.467 Australia, 0.284 North Macedonia, 0.108 Canada, 0.107 Belarus, 0.03 Albania and 0.003 Singapore.
Hungary experienced a positive impact on its net export per capita due to the EU accession. Furthermore, the import per capita was lower than export per capita and therefore, net exports positively influenced the GDP per capita growth in post-accession Hungary. This shows that even though Hungary’s GDP per capita decreased, the drop was not caused by a change in trade, however, another GDP component declined after joining the EU.
The p-values for net export per capita were statistically significant on 10% level for years 2010 to 2017. If Hungary did not join the EU, in 2017, its net export would be lower by 2 444 USD. On average, Hungarian net exports per capita increased by 1 800 USD.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 18: Comparison of net export per capita, Hungary
Source: Author’s calculations
7.2.5 Private consumption per capita
Below graph illustrates the synthetic and real private consumption per capita in Hungary. The fit of the variables is quite good and shows negative effect of joining the EU on the private consumption per capita. The graph shows discrepancy between the synthetic and real values of the private consumption after the accession date, however, the firstly slightly positive effect turned in 2007 to negative. Moreover, the impact was experienced one year before the EU accession.
The synthetic Hungary consists of 0.417 Ukraine, 0.321 Singapore, 0.14 North Macedonia, 0.065 Belarus, 0.034 Jordan, 0.02 Russia and 0.003 Australia. The graph illustrates the same pattern as the GDP per capita, however, the results were statistically significant on 10% level only for year 2014.
According to the results, if Hungary did not join the EU, the private consumption per capita would be higher by 1 264 USD than its real value in 2017. Furthermore, on average, private consumption per capita declined by 965 USD.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
Figure 19: Comparison of private cons. per capita, Hungary
Source: Author’s calculations
7.2.6 Government spending per capita
Below graph represents the government spending per capita in Hungary and synthetic Hungary. The graph lacks fit for some of the pre-treatment period years and shows that EU membership had affected the government spending even before the accession date.
Furthermore, the synthetic Hungary consist of 0.386 Canada, 0.288 Belarus, 0.286 Hong Kong and 0.039 Albania.
The p-values were not statistically significant for any of the post-treatment years, which was expected due to the lack of fit in the pre-accession period. Nonetheless, the analysis showed negative effect of joining the EU on government spending per capita.
If Hungary did not participate in the EU, its government spending per capita in 2017 would be higher by 424 USD. Furthermore, on average, the government spending per capita decreased by 358 USD.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 20: Comparison of gov. spending per capita, Hungary
Source: Author’s calculations
7.2.7 Investment per capita
Lastly, below graph illustrate investment per capita for Hungary and synthetic Hungary.
The fit of the pre-treatment period is relatively good, and the graph shows negative impact of accessing the EU on the investment per capita level. The synthetic Hungary consists of 0.402 Australia, 0.356 Albania and 0.242 North Macedonia. Even though the p-values were statistically significant on 10% level only for year 2012, we can assume that investment decreased after the accession date.
If Hungary never joined the EU, its investment per capita in 2017 would be higher by 258 USD. The highest difference in the real and synthetic values was in 2012, when the investment per capita values reached the lowest point. During the last treated years, the country recovered, and its values converged to the synthetic ones. On average, if Hungary did not join the EU, its investment per capita would be higher by 942 USD.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
Figure 21: Comparison of investment per capita, Hungary
Source: Author’s calculations
7.2.8 Robustness check
To evaluate the validity of the SCM results for the GDP and its components, three robustness checks were carried. The below graphs illustrate the synthetic, real and robustness check values for each of the outcome variables.
The below graphs show that the estimates for import and export per capita are not robust and therefore, the results provided may be inaccurate. This was expected due to the lack of match of the variables in the pre-intervention period. Moreover, the results for net export per capita are not fully robust either, therefore, the effect on the trade can be overestimated. The results for the rest of the variables are fairly robust and therefore, we can consider those estimates accurate.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 22: Robustness check, Hungary
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Results
Source: Author’s calculations
Summing up, the analysis of Hungary and synthetic Hungary reveals ambiguous results.
For the GDP per capita, the estimated impact of joining the EU was negative.
Furthermore, the results were statistically significant on 10% level for two out of fourteen post-treatment years. The results for net export showed large positive effect of joining the EU on the GDP per capita with statistically significant results for years 2010 to 2017. Private consumption, investment and government spending followed the same pattern of negative effect of joining the EU on their values, however, the results were not generally statistically significant on 10% significance level40. For import and export, the match of the pre-intervention period was insufficient and therefore, the results were not interpreted. Furthermore, the results were not fully robust as was confirmed by the latter robustness check.
The analysis revealed EU membership had inconclusive effects on the GDP components in Hungary, with positive effect received for trade, which was, however, outbalanced by the negative estimated effects on the rest of the GDP components.
Overall, the EU membership had a negative impact on the Hungarian GDP per capita.
The results were consistent with the analysis presented by Campos et al. (2014)41 that estimated negative effects of the EU enlargement on the Hungarian output per capita.
Even though Hungary was the most prominent candidate for participating in the EU, the SCM analysis estimated a negative effect on the GDP per capita. Before the EU
40 Private consumption per capita was statistically significant on 10% significance level only for year 2014, investment per capita was statistically significant only in year 2012.
41 The negative effect occurs when the authors used year 2004 as a treatment. Using year 1998 as a treatment showed the opposite results.
‧
accession, Hungary’s income per capita was second highest in the V4 group. Slovakia overcame Hungary early in 2007 and in 2008, the Hungarian GDP per capita fell sharply and its values reach similar levels as for Poland, which had from the beginning the lowest GDP per capita in the V4 group. According to Pintér (2018), after the accession, Hungary lost its position of a regional leader and alongside with bad economic policy decisions, the country fell way behind the old EU members. Therefore, the Hungarian economy did not reach positive impacts from the EU membership. Moreover, the European integration skepticism after joining the EU grew. Pintér (2018) also refers to 2010 when the Hungarian political leadership changed, and positive results were reflected in the economic growth. The graph for GDP per capita shows a relative convergence of the synthetic and real GDP per capita levels for the few last years, which confirms the author’s theory.
Jedlička et al. (2014) also blame the relatively low income convergence to the old EU15 states on the domestic policy mistakes. Hungary had from year 2003 to 2006 twin deficit (Jedlička et al., 2014). In the mentioned years, the budget decision-making in Hungary was rather politically oriented. The government expected to receive subsidies from the structural funds that would cover their spending in the future. Nonetheless, the EU funding did not cover the debts made by the Hungarian government and the inflow was not immediate. The political campaign led to many bad policy decisions to buy the voters, which caused large fiscal deficit (Győrffy, 2007). It can be assumed that the budget deficit and unstable Hungarian economy caused foreign investors avoiding investing in Hungary and domestic investment discouragement, thus may led to the negative effect on the investment. Furthermore, to cut the large budget deficit, the government had to implement cuts of the social benefits, governmental jobs and government expenditure that all together led to decrease of the private consumption and government spending.
The case of Hungary presents that the EU membership as itself is not a riskless way to an economic success. Even though the single market leads to opening of the economies, high investment potential and increase in net exports, the policy decision-making must be conducted with consideration of the economic situation of the country. Even though Hungary experienced certain level of capital inflow from the EU funds and EU investors, it did not offset the budget deficit and alongside with the incoming recession, the Hungarian economy experienced a large slump from which it
‧
did not fully recover. Hungary is an example of badly carried EU accession and should be considered by the new possible EU applicants before deciding to join the EU.
Poland
The data for Poland are available since year 1995. Thus, the pre-treatment period is shorter by four years and in total reaches 9 years. The post-intervention period remains the same as for the previous countries, 14 years. Before the EU accession, the value of the Polish GDP per capita reached the lowest values in V4 group, and after the EU accession remained the lowest. According to the preliminary analysis, the results for Poland are expected to be positive.
The results revealed that EU accession slightly positively affected the GDP per capita in Poland, however the effect occurred after year 2014. Before year 2014, there was none or slightly negative effect on the GDP per capita level. Furthermore, very much alike results were retrieved for import and private consumption. The export and government spending were positively influenced by the EU participation. On the contrary, investment per capita experienced negative effect of joining the EU.
7.3.1 GDP per capita
The below graph represents the SCM analysis for the GDP per capita in Poland. The below graph shows that after year 2014, Poland reached positive effect from the EU accession. Before that, the effect was somewhere between none and negative. The average effect on the GDP per capita was negative. The synthetic Poland consists in total of 0.418 Korea, 0.373 Belarus, 0.197 Uruguay and 0.011 Australia.
The retrieved p-values for Poland were not statistically significant for any of the post-treatment years. If Poland decided not to join EU, the value of GDP per capita in 2017 would be lower only by 1 381 USD. On average, the GDP per capita was lower by 314 USD after the joining the EU.
Even though there is not a strong evidence in the GDP per capita, there might have had a statistically significant positive effect on some of the GDP components.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 23: Comparison of GDP per capita, Poland
Source: Author’s calculations
7.3.2 Export per capita
The graph below shows strong positive effect of the EU on Polish export per capita.
The synthetic export per capita consists of 0.343 Belarus, 0.304 Chile, 0.139 Ukraine, 0.135 Japan, 0.073 Brazil, 0.006 Korea. The pre-treatment period provides good fit.
The p-values for export per capita were statistically significant on 10%
significance level for year 2013 to 2017 and thus, the EU had a positive and statistically significant effect on the Polish exports. If Poland decided not to join EU, the export per capita would be in year 2017 lower by 7 324 USD. In the post-treatment period, on average, the Polish export per capita experienced an increase of 2 611 USD.