Social Protection after the Pandemic: Lessons Learned from the Eurozone Crisis

As vaccinations against COVID-19 move forward and the health crisis will eventually reside, focus will ultimately turn solely on the economic consequences of the pandemic. Having faced a serious economic turmoil during the past decade and with the perceptions about the role of the welfare state having changed drastically in the meantime, the European Union needs to learn from its mistakes in dealing with such crises and ensure sustainable recovery for its members. In this paper, the correlation between social protection expenditure, income inequality and poverty is examined, as well as the GDP and the employment rates, using the Eurostat database for the period 2007 – 2017 (subject to data availability). By attempting to identify the mechanisms that could lead to such results, the aim of the policy brief is to paint a realistic picture of the role of social protection in the European economies. Among other things, the results show that social protection expenditure correlates positively with employment rates and the GDP per capita, and negatively with income inequality and poverty. We conclude that the welfare state retrenchment during the Eurozone crisis deepened the recession and probably led to long-term adverse effects, broadening the gap between the North and the South. Building on the measures taken during the pandemic to protect employment and the economy, post-pandemic recovery policies should focus on rebuilding strong welfare states, directly create jobs and aim at the convergence of the economies through expansionary policies and growth.


Introduction
Shortly after the global financial crisis and its effects on the EU both at an economic and a political level, the world is faced once again with an economic crisis, this time triggered by the COVID-19 pandemic. The extent and the universality of the economic consequences of the pandemic have activated a series of policies that hardly resemble the response against the economic turbulence in the Eurozone back in 2010, at least while the health crisis is under way. The provisions of the Stability and Growth Pact have been put on hold and the ECB continues to provide liquidity through an extended Qualitative Easing program, as the EU member -states " […] took decisive action to protect employment, income and access to services through a variety of support measures" (Social Protection Committee, 2020). As Tzagkarakis et al. (2020) note, the welfare state has once again emerged as a vital institution in modern economies, especially during a crisis. However, it is widely acknowledged that once the health crisis is over, there will be a return to austerity in the EU in order to tackle the debt and the deficits accumulated during this period.
The return of the Economic Adjustment Programs has been foreshadowed by the amendment of the ESM Treaty in February 2021. During the previous decade, these programs targeted the welfare state (as well as the labor market institutions) and led to significant cuts, severely underestimating the macroeconomic consequences and the contractionary effects of such policies (Blanchard & Leigh, 2013;IMF, 2013). 4 This policy brief aims to examine the correlation between social protection expenditure, income inequality and poverty during the Eurozone crisis along with employment and the GDP, to identify the policy orientation required for post-pandemic sustainable recovery and growth. In the first section, the methodology and the datasets used are presented. In the second section, we present the correlations between the variables and attempt to interpret the mechanisms behind them. Finally, the implications of the findings are discussed for the design of the postpandemic policies at an EU level.

Data and Methodology
This policy brief studies the EU-28 countries with the addition of Norway and Switzerland, using the Eurostat database. The datasets used include the social protection expenditure, the employment rate, GDP per capita, the Gini coefficient and material deprivation. Social protection expenditure (SPE) refers to social benefits that aim to support households and individuals from certain risks. For this study, SPE is presented both as a percentage of GDP and as social protection expenditure per capita in Purchasing Power Standard (PPS). GDP per capita is also expressed in PPS, to control for the different costs of living in the region. Employment rate describes the employment to population ratio, while the Gini coefficient, one of the most popular inequality indices, displays income inequality in an economy, taking a value from 0 to 1, with 0 signaling total income equality and 1 total income inequality. The material deprivation (MatDep) rate depicts the percentage of the population that cannot afford some basic goods or services out of a specific list of goods. 5 The average value of each dataset is calculated for the years 2007 -2017 or 2008 -2017 (according to data availability), and the Pearson correlation between these values is presented. It is important to note that the correlations that 4 Olivier Blanchard, Chief Economist of the IMF at the time, recognized that fiscal multipliers were underestimated during the design of the European Economic Adjustment Programs (Blanchard & Leigh, 2013). IMF (2013), in its evaluation for the 1st Program implemented in Greece, admits to the underestimation of the fiscal consequences of austerity too. These admissions, however, seem to have been forced by the obvious failure of their original predictions and not by a sincere realization of the inadequacy of their analytical framework, as the continuation of these policies showed. 5 According to the definition given in europa.eu (n.a.), material deprivation variable "measures the percentage of the population that cannot afford at least three of the following nine items: to pay their rent, mortgage or utility bills; to keep their home adequately warm; to face unexpected expenses; to eat meat or proteins regularly; to go on holiday; a television set; a washing machine; a car; a telephone".
vol. 2 | no. 1 | June 2021 72 are examined hold for each year separately too. Since correlation does not necessarily mean causality, there is a theoretical analysis of how each variable can affect the other so as to lead to these results.

Results
Before we delve into the correlation between social protection expenditure, income inequality and poverty, it is interesting to examine the relation between social protection expenditure and other fundamental macroeconomic variables, such as the employment rate and the GDP.
Source: Eurostat (2020a;2020b;2020c), data processed by the authors * In Figure 1, Greece, Italy, France, Belgium, Luxembourg and Estonia are outliers and therefore not included. In Figure   2, Switzerland, Luxembourg, Norway and Ireland are outliers and therefore not included. In both cases, the inclusion of the outliers would show a statistically significant, moderate, positive correlation.
The reasoning behind examining the correlation between the social protection expenditure and the employment rate is quite clear: the welfare state in general and social protection in particular have been long accused of promoting complacency among the beneficiaries and stripping the motives for participation in the labor market. A problematic view for several reasons, yet a popular argument against the welfare state in the public debate. We would expect that if this argument is valid, there should be a negative correlation between social protection expenditure and employment rates. Figure 1 shows a statistically significant, strong positive correlation between SPE and Employment Rate. Consequently, the argument that social protection hinders labor market participation is not empirically verified by the aforementioned data. On the contrary, there seems to be a tendency for countries that spend more on social protection to present higher employment rates. To explain this result, the relation among the two variables can be examined from both perspectives. Higher employment rates can lead to increased social protection expenditures by broadening the tax base and thus provide the state budget with more resources to finance social protection. 6 Moreover, generally the higher employment rates should mean fewer citizens in need of social protection. These two effects combined can translate into higher and more efficient social protection. At the same time, higher social protection expenditure can lead to increased employment rates, by increasing the aggregate demand. By providing disposable income to low-income households (and, therefore, households with high propensity to consume) consumption in the overall economy is boosted, creating the need and the motives for increased production (and, consequently, private investment), resulting in the creation of more jobs. 7 As far as the correlation between SPE and GDP per capita is concerned, Figure 2 shows a statistically significant, strong positive correlation as well. Although in this figure social protection expenditure is calculated as a percent to the GDP, the same conclusion would have been reached with the use of the data on the first figure. The tendency that the higher GDP per capita, the higher the social protection expenditure is (expressed either as a percent of GDP or as per capita) can once again be interpreted both ways, which are not mutually exclusive. Increased affluence in an economy can sensitize and promote the need for the protection of vulnerable groups and low-income households.
At the same time, as mentioned above, social protection expenditure allows low-income households to actively participate in the economy, triggering the multiplier effect and increasing economic activity. As Dercon (2011) notes, social protection promotes inclusive growth, while at the same time improves economic efficiency by correcting market failures.
Given the correlation between the two variables, extensive social protection seems togenerallybe a characteristic (and some would argue a prerequisite) of a strong economy. As pointed out by Sestito & Ca'Zorzi (2001), who found a similar correlation for the period 1993 -1996, the level of social protection expenditure should not be attributed solely to the level of GDP (and vice versa). A characteristic example is the Nordic countries, where the extensive welfare state is part of their culture and on the foundation of their economic institutions, and not just a mere economic outcome derived from high GDP. 6 As the Euro is practically a foreign currency for the members of the Eurozone (in the sense that they don't have the authority to issue new money to fund their policies), and due to the provisions of the Stability and Growth Pact, countries in the EU need to fund their activities through tax revenues. 7 Since it is not a temporary positive shock in demand but a permanent feature of the economy, simply increasing operation to full capacity to meet increased demand on the short-run is not a viable solution for the firms, motivating them to invest in capital and employees in order to maintain or even increase their market share.
Source: Eurostat (2020a;2020c;2020d), data processed by the authors * In Figure 3, Malta, Czech Republic, Hungary, Slovenia, Slovakia, Switzerland, Ireland and Luxembourg are outliers and therefore not included. In Figure 4, Malta, Czech Republic, Hungary, Slovenia, Slovakia and Luxembourg are outliers and therefore not included. In both cases, the inclusion of the outliers would show a statistically significant, moderate, negative correlation. Figure 3, there is a statistically significant, strong negative correlation between the Gini coefficient and the GDP per capita, indicating a tendency for lower income inequality co-existing with higher GDP per capita (and vice versa). Higher GDP per capita usually translates into higher tax revenue, which can be used to finance social protection, and therefore income distribution policies, reducing income inequality. Moreover, as observed in Figure 2, social protection expenditure is positively correlated with higher GDP per capita, and the increased disposable income of lowerincome households plays an important role in this. On that basis, the mechanism behind the negative correlation between the Gini coefficient and the GDP per capita, is the same one that governs the correlations examined in Figures 1 and 2, concerning the increase in the aggregate demand via the redistribution of income to low-income households with higher propensity to consume, thus improving the performance of the overall economy.

According to
For this mechanism to apply in all these correlations, it is implied that social protection expenditure does, actually, lower income inequality. Figure 4 shows that indeed, higher social protection expenditure correlates with lower income inequality, as there is a statistically significant, strong negative correlation between the two variables. Although quite intuitive, this result is very useful to be born in mind when assessing the efficiency of social protection policies in an economy. The volume of spending alone does not guarantee either efficiency or successful social protection. Cosmin It is important to address that in mainstream economic theory, the role of income inequality in economic performance is highly debated, and until recently the dominant (neoliberal) perspective had been in favor of letting inequality rise, predicting that this would be beneficial for the overall economy. This perspective leans on the notion of "trickle-down economics", claiming that savings from the higher-income households (and firms) would lead to higher productive investments, thus creating jobs and benefiting the society as a whole. Although it works in theory, real-world economics show that large portions of the higher-income savings leave the economy, being invested in international financial products or ending up in tax havens. This leads to reduced effective demand, hurting the economy and growth. The stagnation that advanced capitalist economies have been facing has led to OECD and IMF-led studies which recognize that increasing inequality has been hindering growth and should be contained (Cingano, 2014;Dabla-Norris et al., 2015). Of course, the debate is far from over. Source: Eurostat (2020a;2020d;2020e), data processed by the authors * In Figure 5, Bulgaria, Malta, Estonia, Czech Republic, Slovenia and Luxembourg are outliers and therefore not included. The inclusion of the outliers would show a statistically significant, moderate, negative correlation. In Figure 6, Czech Republic, Hungary, Slovenia and Slovakia are outliers and therefore not included. The inclusion of the outliers would still show a statistically significant, strong, positive correlation. Figure 5 shows a statistically significant, strong negative correlation between the material deprivation rate and social protection expenditure. This result is not really surprising, given that the goal of the welfare state is to mitigate material deprivation by promoting access to basic goods for lower income households. It is important to be noted, though, as arguments against higher social protection expenditure may question whether social protection actually reduces poverty.
Contrary to the debate around income inequality that was addressed earlier, there seems to be a consensus on the need for the elimination of poverty. However, the debate on how poverty should be eliminated remains, while opinions vary, ranging from the role of a big, interventionist state to the completely unregulated free market that promotes an efficient allocation of resources. Regardless of the various theories, European states do tend to be more interventionist in order to ensure social protectionat least to a certain extentin comparison to other countries (e.g. the USA). Given the results portrayed in Figure 5, the argument that less state intervention leads to more economic efficiency and socially-desired results does not seem to apply in Europe. Higher social protection expenditure is strongly correlated with lower rates of population living in conditions of material deprivation.
According to Figure 6, there is a statistically significant, strong positive correlation between material deprivation rates and income inequality in the European economies of this study. As mentioned before, the debate on the role of income inequality remains strong and topical. In line with the previous findings of this study, however, a strong correlation between poverty rates and income inequality appears reasonable: states with higher income inequality tend to show higher material deprivation rates.

Conclusions and Discussion
This study presents some basic trends that appear in the European economy. Far from an in-depth analysis, that certainly is useful but also in danger of biased models that lose touch with the reality they are supposed to examine, this policy brief aims to take a step back and try to comprehend the bigger picture. Given the results that the welfare state retrenchment produced both at a social, as well as at an economic level during the past decade (Kotroyannos et al., 2013;Tzagkarakis et al., 2021), the findings presented at the study seem to depict a realistic picture of the European economy.
Decreased social protection seems to weaken employment and the GDP instead of strengthening them, while increasing poverty. Rising income inequality doesn't seem to translate in better economic results via increased productive investments, but rather in increased poverty and lower GDP.
When the health crisis subsides and the focus will turn solely on the economic effects of the pandemic, the orientation of the EU policies should prioritize the rebuilding of a strong welfare state, inclusive and sustainable growth and the creation of quality jobs. The pandemic-induced economic crisis is characterized by both a demand and a supply shock, which differentiates it from the oil crisis of the '70s and the global financial crisis of 2007. As the restrictions are being lifted and trade and travel will be returning to normal, the supply chains will be "rebuilt". However, as shown by the secular stagnation that followed the global financial crisis, the effects of a demand shock are long-lasting, with fiscal policy being a very effective tool to tackle them (Stockhammer, 2021).
The post-pandemic EU has the chance for a new social contract and a restart of its failed economic framework. The orientation of a European New Deal should be towards the convergence of the economies through growth, investments and solidarity, and not through contractionary policies that dangerously broadened the gap in the past decade (Sbarouni et al., 2019). From the convergence of the trade balances through institutional reform (Emmanouil-Kalos, 2020) to the implementation of ECB-financed ELR programs, a series of radical yet necessary reforms must take place, in order for the EU to adequately rise to the challenges that lie ahead.
The modest (yet important in the European reality) attempts to cope with the recession and keep unemployment from skyrocketing during the pandemic should be maintained and expanded. ELR programs must finally be implemented in member-states throughout the EU, supported by the European Commission and financed by the ECB. Admittedly a radical reform for the conservatism of the EU, but a much needed one in order to fight unemployment and the stagnation that threatens European economies for the next decade. As Anastasakis (2020) points out, ELR programs can also be used as a tool towards countering climate change, and be combined with green public investments.
Although tackling unemployment is a priority in order to ensure social protection, it should not be ignored that in-work poverty is actually an important part of overall poverty, a result possibly linked to the deregulation of the labor market (Dafermos & Papatheodorou, 2012). Decent wages must be ensured throughout the EU, for a sustainable recovery and long-term growth.