The Impact of Pandemics on Workforce Joblessness in Central Europe in pre-Covid Era and during Pandemic

people with the employment of up to one year is in the age group 25-54. In none of the examined countries was a gender unemployment gap proved before Covid-19.


Introduction
Before Covid-19, it was predicted that for Europe as a whole, unemployment rates will return to their 2008 levels by 2030 (Cedefop, 2016). But unemployment rates globally vary dramatically in this time of Covid-19, even among the world's largest economies. Like the Great Recession and the recessions of the 1980s, the 90s, the early 2000s, and the 2010s, the Covid-19 recession caused sustained but unequal high unemployment. In the many countries severely affected by the economic crisis, the long-term unemployment (LTU) rate constitutes a general risk for the working population (Duell et al., 2016). The EU economy will experience a deep recession due to the coronavirus pandemic (EC, 2020).
In the pre-Covid-19 era, the Czech Republic retained the lowest unemployment rate (CSO, 2019), but due to the spread of the pandemic, its unemployment rate has risen (CSO, 2021). Overall, EU employment in the euro area rose to 8.3% and in EU-27 to 7.5 % in November 2020 (Eurostat, 2020). Since the early 1980s, unemployment has been a serious problem in Europe, especially among the youth, affecting the southern European countries the most (Hernanz and Jimeno, 2017).
Unemployment is not only a European but an intercontinental issue. Some of the sectors have gone into overdrive, e.g. health, manufacturing of food, beverages, transportation, while other large sectors, ranging from services to hospitality and tourism, have been deliberately shut down, resulting in high unemployment.
This study aimed to examine the effects of unemployment in three-generation groups in V4 and Austria in the pre-Covid-19 era and during Covid-19. The research question of this study was: Which age group in the surveyed countries is most affected by Covid-19 job losses?
In this paper, we briefly review the evidence and offer some general perspectives on its interpretation. The next paragraph describes the existing theoretical debate on the causes of unemployment. The second paragraph describes the methodology used in this paper. The subsequent part provides the results. The fourth section contains the discussion, and the last part offers a brief conclusion.

Causes of unemployment
LTU is felt to have disastrous effects on the individuals who suffer it, both in terms of their labor market opportunities and their more general physical and mental well-being (Machin and Manning, 1998). The most significant cause of youth employment is poor macroeconomic performance. This results from a combination of slower rates of economic growth, demographic trends, and structural factors (OECD, 1978). Further, lack of growth affects each person in the economy, and especially some age groups are severely affected (e.g. youth unemployment has greater cyclical amplitude than adult unemployment (OECD, 1982)). According to Ryan (2001), youth are more severely affected than adults.
Recessions naturally drive up unemployment across the population (Knotek and Terry, 2009;Tasci and Zaman, 2010). The effects are more serious for those who have left their educational system and started their professional life (so-called frictional unemployment). For instance, the unemployment rate rose sharply in the European Union after March 2008 due to the global economic crisis (Eurostat, 2014).
Unemployment may also be influenced by where people live, e.g. in Australia. McDonald (1995) highlighted the higher rates of unemployment experienced by those living in older industrial areas. Also, Gregory and Hunter (1995) found that there had been little or no employment growth for people living in low socio-economic areas between 1976 and 1991, in contrast with the better conditions experienced by people living in higher socio-economic areas (McClelland and Macdonald, 1998). Those living in countries where there are social security policies and small business development occurs suffer less from the adverse effects of unemployment (Farber and Valletta, 2015) than those in underdeveloped countries, which tend to suffer more from the negative effects of unemployment (Duygan-Bump et al., 2015;Startiene and Remeikiene, 2015).
Children from less-privileged backgrounds experience more adult unemployment but are less affected by it in terms of well-being (Clark and Lepinteur, 2019). These authors further add that both educational achievement and good behavior at age 16 reduce adult unemployment. Ahmad and Khan (2016) conclude that joblessness is a mixture of economic, community, and other specific elements. Based on Spermann (2016), the risk factors for LTU include old age and lack of vocational training. Black et al. (2015) stress that unemployment insurance policies benefit unemployed workers by giving them the resources to become qualified and reintegrate into the labor market. Rodenburg (2004) stresses that the exact underlying cause or causes of unemployment can seldom be identified separately, and explanations are often surrounded by a set of auxiliary assumptions. The author further adds that technical progress and the immobility of labor or union power are hard to measure. Corry (1995) further notes that economists are not in control of policy and hence cannot be pilloried for the failure of the economic system to create jobs for all.
The gender unemployment gap was positive until the early 1980s. The gap disappeared after 1983, except during recessions when men's unemployment rates always exceeded those for women (Albanesi and Sahin, 2017).
Say's the law, namely that supply creates its demand, failed in 1930 due to the Great Depression, and has failed all over again today, this time due to Covid-19, once more proving a triumph for the Keynesian economist, the great prophet of the principle that demand creates its supply (Sirah and Atilaw, 2020). According to Rodríguez-Caballero and Vera-Valdés's data (2020) on unemployment, the periods associated with the Great Pandemic of 1870-1875 and the Russian flu show a more persistently higher level of unemployment. Additionally, after the Spanish flu pandemic and the First World War, the level and persistence of unemployment increased.

Research Method
The main purpose of this study was to examine the development of unemployment in three generations in selected countries during the pre-Covid era and Covid-19. We looked for answers for this basic research question: Which age group in the surveyed countries is most hit by Covid-19 job losses? Relationships based on OECD data for V4 and Austria were investigated. We have divided the unemployment data into three age groups: 1) 15-24, 2) 25-54 and 3) 55-64.
Firstly, a comprehensive review of the available literature for the given research question was done to explain the main causes of unemployment around the world. We briefly explain the nature and causes of unemployment. Secondly, Excel calculations and descriptive statistics (test of normality (Shapiro-Wilk test, Mann-Whitney U test), the test of homogeneity (Levene test), robust tests of equality of means (Welch Test), multiple comparisons, cross-tabulation, pairwise comparisons) were used to analyze the surveyed quantitative data. As Freeman and Julious (2006) emphasize, it is good practice to produce a table or tables that describe the initial or baseline characteristics of the sample. In this study, three basic tables have been prepared: 1) Unemployment rates from Q3-2018 to Q3-2020, 2) Unemployment rates from 2000 to 2019, and 3) Job tenure of less than one year from 2000 to 2019 (OECD, 2021). Descriptive analysis is data simplification. Good description presents what we know about capacities, needs, methods, practices, policies, populations, and settings in a manner that is relevant to a specific research or policy question (Loeb et al., 2017). The following methodological background was used in this study: induction, descriptive statistics, synthesis, deduction in development of results, and concluding. Results are interpreted in graphic and narrative form and differences are discussed.

Results
The current pandemic recession, like those in the past, has already driven up the number of people who are not employed. It cut the number of available vacancies or offered a short-time work model. Generally, the lowest paid, the lowest skilled and the least experienced workers are those who are most severely affected.

Influence of age on unemployment (Q3 2018-Q3 2020)
The economic effect of the coronavirus has taken the surveyed countries into unknown territory. Based on Axelrad et al.'s (2018) empirical data, older workers' difficulties are related to their age, while for younger individuals the difficulties are more related to the business cycle. Aging is the most important demographic change for employment (Zipperer, 2015).
Firstly, verification of normality using the Shapiro-Wilk test was done, as shown in Table 1 Source: Author's own elaboration All p-values (Sig.) are higher than the significance level of 0.01, therefore the data can be considered as normally distributed at the significance level of 1%. Further, to verify the difference in unemployment in the three age groups, an analysis of variance was used, see Table 2. Source: Author's own elaboration Verification of the assumption of homogeneity of variances using the Levene Test are shown in Table 3:

Source: Author's own elaboration
Based on the received data, the assumption of homogeneity of variances is not met in any surveyed country (Sig. <0.05). Thus, Welch's analysis of variance was applied, which takes into account the failure to meet this assumption (See Table 4). In all countries, the p-value of the test is lower than the significance level of 0.05, and there is at least one pair between age groups in all countries that differs significantly. Employing post-hoc tests of multiple comparisons, it has been found which groups differ, as displayed in Table 5.  Table 5 shows that in Austria, Hungary, Poland, and Slovakia, all three age groups differ in terms of unemployment. It can be concluded that unemployment decreases with age. In the Czech Republic, there is no significant difference between the 25-54 and 55-64 age groups. People aged 15-24 show significantly higher unemployment than other age groups.

Gender influence on unemployment (Q3 2018-Q3 2020)
The unemployment gender gap (female and male unemployment rates) was positive until 1980. The gap virtually disappeared after 1980, except during recessions, when men's unemployment rates always exceed those of women (Albanesi and Sahin, 2017).
Firstly, a verification of normality was done using the Shapiro-Wilk test, as shown in Table 6.

Source: Author's own elaboration
Since the assumption of normality is not met, the non-parametric Mann-Whitney U Test was used to compare the figures for males and females, as shown in Table 7.

Male
The Wilcoxon paired test for males confirmed the difference in unemployment before and during pandemics, as displayed in Table 9.

Asymptotic significances are displayed
The significance level is 0.05.

Source: Author's own elaboration
The distribution of unemployment before and during Covid-19 in the surveyed countries is shown in Figure 3a and b. On the basis of these data, unemployment has risen during Covid-19. a) Before b) During

Female
The Wilcoxon paired test for females did not confirm the difference in unemployment before and during pandemics, as displayed in Table 10.

Source: Author's own elaboration
The difference in female unemployment before and during Covid-19 was therefore not confirmed (Sig. > 0.05). The dramatic female share and the notable decline of the male share of unemployment have received considerable attention (Albanesi and Olivetti, 2016;Albanesi and Sahin, 2017;Greenwood et al. 2005;Olivetti 2006), and the data obtained in this study confirm this tendency.

Age
The T-Test confirmed the difference in unemployment before and during Covid-19 in categories of 15-24 and 25-54 years, as displayed in Table 11

Comparison of job tenure of less than one year (2000-2019)
An analysis of the tenure distribution for the individual member states of the EU revealed strong cross-country differences in the pre-crisis period (Bachmann et al., 2015).
Firstly, a verification of normality using the Shapiro-Wilk Test was done, as shown in Table 12. In order to use the parametric test, all groups must meet the normal data distribution. Normality is met at the 1% level of significance in all countries except Austria (see Table 12). For this country, the non-parametric equivalent Kruskal-Wallis Test was used.

V4 countries
Further, to verify differences in job tenure in three individual age groups, analysis of variance was used, see Table  13. Again, Welch's analysis of variance was applied, which takes into account the failure to meet this assumption (see Table 14). In all countries, the p-value of the test is lower than the significance level of 0.05. There is at least one pair between the age groups in all countries that differs significantly. We can find out which groups differ by using post-hoc tests of multiple comparisons, as shown in Table 15. In Hungary, Poland and Slovakia, all age categories differ in the number of people with a duration of employment of up to one year. In the Czech Republic, there is a significant difference only between the youngest group and the other two. In all countries, the largest number of people with the employment of up to one year is in the age group 25-54 years.

Austria
The difference between the number of people with employment of up to one year in the age groups is verified for Austria by a non-parametric analog of the analysis of variance, the Kruskal-Wallis Test. The results in Table 16 and Figure 5 show that at least one pair of age groups was confirmed (Sig.< 0.05).

Asymptotic significances are displayed
The significance level is 0.05.

Source: Author's own elaboration
Further analysis through pairwise comparisons of age confirms a significant difference between all age groups. The group with the most people employed for up to one year is again the age group 25-54 years. The group with the fewest of them is the age group 55-64 years, as displayed in Figure 6.

Comparison of unemployment rate by gender (2000-2019)
In this section, primarily, verification of normality using the Shapiro-Wilk Test was done, as shown in Table 17. The assumption is not met for Austria and the Czech Republic as Sig. < 0.05. For these countries, a parametric T-Test was used.

Hungary, Poland and Slovakia
Firstly, group statistics were compiled, as shown in Table 18. Verification of the assumption of homogeneity of variances using the Levene Test are shown in Table 19. The P-values are all higher than the chosen level of significance, therefore the difference in unemployment between males and females in the period 2000 to 2019 was not confirmed.

Austria and the Czech Republic
For Austria and the Czech Republic, a non-parametric test similar to the T-Test, the Mann-Whitney U test, was applied, see Table 20. In none of the surveyed countries was a difference between female and male unemployment confirmed.

Discussion
Even before the pandemic, youth unemployment in the European Union was three times higher than among the over-55s (Grzegorczyk and Wolff, 2020). Data obtained in this study show the same threat because the younger generation is more severely affected than older generations. But, as stated by Cubanski et al. (2020), older adults are severely affected by Covid-19 and are also losing their jobs. As shown by the data from this study, the T-Test confirmed the difference in unemployment before and during the crisis in the age categories 15-24 and 25-54. Compared to previous recessions, the current recession has increased the number of people who are still able to work remotely (Eurofound, 2021).
But, while crises will naturally affect all workers differently, will this disproportion be experienced more severely by the most vulnerable: the youngest, the lowest paid, the lowest skilled, and the least experienced? Does Covid-19 put the future employment of millions of workers and the viability of thousands of businesses at risk? Based on recent data, women are facing a greater risk of unemployment and/or being placed on furlough or equivalent employment protection schemes (Wenham, 2020). In the surveyed countries, however, Covid-19 has a greater threat of unemployment for males. Significantly, a gender unemployment difference was observed in the Czech Republic and Slovakia.
The question remains: Will countries recover from this crisis after Covid-19 ends with a possible jobs boom? This crisis comes on top of pre-existing challenges. Since 2000, there has been a shift in the US job tenure distribution toward longerduration jobs. A substantial number of these changes are caused by the aging of the workforce and the decline in the entry rate of new employer businesses (Hyatt and Spletzer, 2016). But, according to the data obtained, in Hungary, Poland, and Slovakia, all age categories differ in the number of people with a duration of employment of up to one year. In the Czech Republic, there is a significant difference only between the youngest group and the other two. In all countries, the largest number of people with employment up to one year is in the age group 25-54 years.

Conclusions
This analysis of unemployment among different age groups presents differences related to different variables, the sample of countries, the time horizon, and the statistical method used.
Based on received data, unemployment decreases with age. The 15-24 group shows significantly higher unemployment than the other two groups. A gender unemployment difference was confirmed only in the Czech Republic and Slovakia. Unemployment has risen during the Covid-19 pandemic. An unemployment gap before and during Covid-19 was not confirmed for females. The T-Test confirmed a difference in unemployment before and during the crisis in the age categories 15-24 and 25-54. In Hungary, Poland and Slovakia, all age categories differ in the number of people with a duration of employment of up to one year. In the Czech Republic, there is a significant difference only between the youngest group and the other two. In all countries, the largest number of people with the employment of up to one year are in the age group 25-54 years. In none of the examined countries was a gender unemployment gap before Covid-19 proved.
If this analysis is correct, the prospects of unemployment in the surveyed countries seem to be rather turbulent for the younger workforce. More comparative analyses such as this are to be recommended because unemployment rates are going to fall to historic lows before the Covid-19 pandemic ends. In addition to socio-economic and technological changes, more people have the possibility of working from home compared with past crises. This means that the rate of unemployed people was not so high. The lesson from the crisis of the 1930s is that if the current crisis leads to a similarly bad downturn, the policy reaction in terms of greater state intervention will not be conducive to improved growth prospects (Crafts, 2011). This necessitates a focus on apprecentships, onsite jobs, or remote jobs for all in the labor market.

Implications
This study investigated the impact of unemployment and job tenure in different generation groups about age and gender in V4 and Austria in the pre-Covid era and during the pandemic. Like the Great Recession and the recessions of the 1980s, the 90s, the early 2000s, and the 2010s, the Covid-19 recession caused sustained but unequal high unemployment. The future holds both significant obstacles and possibilities for the different workforce generations. Research conducted in this study indicates that the younger generations are more affected than the older generations. However, policy adjustments and investments in modern technology and e-education can help to improve the job market.