







































26 

 

Finance, Accounting and Business Analysis 
Volume 4 Issue 1, 2022 

http://faba.bg 

The Consequences of Covid-19 on Youth Unemployment in The Bulgaria  

Stanimir Stamatev1, Vasil Bozev2  

Department of Management, University of National and World Economy, Bulgaria1 

Department of Statistics and Econometrics, University of National and World Economy, Bulgaria2 

Info Articles  
 

Abstract 

Keywords:  
Youth unemployment, COVID-19, 
Education level, Bulgaria 

 Objective: The purpose of this article is to examine how COVID-19 has an impact 
on unemployment among young people in Bulgaria. It shows the impact of the 
achieved educational level on the change in the unemployment levels among young 
people. In addition, this academic work attempts to establish what the future 
tendency in the development of youth unemployment might be in a subsiding 
pandemic. 

Methodology: The article uses empirical data from the NSI (National Statistical 
Institute) and Eurostat regarding the level of youth unemployment in Bulgaria on a 
general and educational scale. A forecast has been made, which is based on a trend 
model, whose parameters are estimated by the method of the least squares. 
Results: Our findings indicate that COVID-19 has had a detrimental effect on 
unemployment among young people and it has risen during the pandemic. In 
addition, our forecast illustrates an increase in youth unemployment in the 
upcoming two years. Despite this, according to the data for Bulgaria, in 2021 
compared to 2019, youth unemployment among people with primary and lower 
education has seen an increase by 7.4%; with 4.2% among those who have 
acquired secondary education and only with 1% among those with higher 
education. The analysis illustrates that during the COVID-19 pandemic, young 
people in Bulgaria with higher education are at a considerably lower risk in 
comparison to those with lower education. 
Implication: The study’s results provides beneficial starting point in improving the 
performance of the Labour market policy and the development of the educational 
system. 
 

 

  

   

*Address Correspondence:   
E-mail: sstamatev@unwe.bg1,  
                v_bozev@unwe.bg2  
 
 

 

 



Finance, Accounting and Business Analysis 4 (1) 2022 

27 

 

INTRODUCTION 

 

The global pandemic COVID-19 broke out in the beginning of 2019 and has set new challenges to 
the economies of the world, including Bulgaria. The soaring growth of the Gross Domestic Product (GDP) 
distinctive for the years before 2019, has rapidly declined in the following years. The declining economic 
activity has also had a detrimental impact on the labor market. This has given good reason to many 
economists and analysts to discuss the state of the labor market in Bulgaria. These negative consequences 
have also taken their toll on the youth as the unemployment rate has gone up by close to 4 percentage 
points in 2021 compared to 2019.  

Pursuing things further, the current conditions of the deteriorating economic and demographic 
tendencies have determined the growing interest in the state, dynamics, and prospects of the results of 
the realization of young people in the labor market. The importance of using young people as an economic 
resource is determined by two main factors. Firstly, to date, they are around 25% of the workforce (NSI, 
Demographic and Social Statistics, 2021). Secondly, they are the future of the nation and the initial 
conditions of their economic activity are an important prerequisite for long-term economic growth. 
 The scope of this paper is to examine the impact of COVID-19 on the Bulgarian youth 

unemployment rate. In order to achieve this goal, we must take into account the empirical data on the 

unemployment of the population aged between 15-29, which is defined in the article as “young people”1. 

Our assumption is that even in a subsiding COVID-19 pandemic, youth unemployment will continue to 

rise. It will have the strongest impact on those with the lowest form of education and to a lesser extent on 

those who have acquired higher education.  
The scope of this paper is to examine the impact of COVID-19 on the Bulgarian youth 

unemployment rate. In order to achieve this goal, we must take into account the empirical data on the 

unemployment of the population aged between 15-29, which is defined in the article as “young people”. 

Our assumption is that even in a subsiding COVID-19 pandemic, youth unemployment will continue to 

rise. It will have the strongest impact on those with the lowest form of education and to a lesser extent on 

those who have acquired higher education. 

 

  LITERATURE REVIEW  

 
A study by the Resolution Foundation (Henehan, 2021) established that in May 2020, a third of 

the people aged between 18-24 have been fired or have lost their main job. Contrary to popular belief, in 
2020 unemployment increased for both graduates and non-graduates (by the same percentage) despite 
the numbers of years that have passed since leaving school. For instance, among those who chose to 
discontinue their education a year earlier, the unemployment rate has jumped from 14% to 18% between 
2019 and 2020. In addition, the unemployment rate for graduates rose from 10 to 14%. African-
Americans took the biggest hit; nearly one in three had recently discontinued their education and were 
left unemployed in 2020 compared to 2019. Asians are in second place where one in four young people 
were left unemployed in 2020 compared to 2019. 

Another study is connected to our neighbor Greece (Katris, 2021). The conducted study examines 
the unemployment in Greece compared to that of the European Union. Youth unemployment is in part 
also touched upon. The analysis studies what the possible consequences of COVID-19 could be on the 
unemployment rate among a variety of groups – young people, men, women, etc. The findings show that 
Greece will be less affected than the European Union. Unemployment among women in Greece and youth 
unemployment in EU27 are expected to take the biggest hit from COVID-19, which suggests the need for 
policy measures to be adopted in order to alleviate the effect it will have on the said groups. 

 

METHODS  

  
The unemployed and unemployment rate 

The levels of employment and unemployment characterize the state of a given labor market. 
Together they form the labor force (the presently economically active part of the population) – all people 
of working age who put in or offer their labor for the production of goods and services (Methodology of 
“Monitoring of the labor force”, NSI). 

Of key importance in the analysis of the labor market is the clear definition and identification of 
the unemployed individuals.  

 
1 According to the Bulgarian legislation, and in particular the "Law on Youth" in Bulgaria, it is accepted that the 
population in the age between 15 and 29 should be included in the group of "young people". 



Finance, Accounting and Business Analysis 4 (1) 2022 

28 

 

According to the methodology of the Periodic labor force surveys (PLFS) conducted by 
EUROSTAT unemployed individuals are those aged 15 to 74 who simultaneously meet the following 
criteria: firstly, they have no job in the week of the survey; secondly, they have actively looked for a job in 
the previous four weeks; thirdly, they are available to start working in the following three months 
(Eurostat). Therefore, the primary criteria, used to identify an unemployed individual, are to not have a 
job at the moment, to have actively looked for one and to have the necessary skills to work. These 
principles are written down in a separate resolution of the ILO (Pavlov, 2008) and have been adopted by 
almost all market economies in the world, Bulgaria included. 

There are two state institutions in Bulgaria, which gather, summarize, and analyze information 
regarding unemployment – the Employment agency (EA) at the Ministry of Labor and Social Policy 
(MLSP) and the National Statistical Institute (NSI). However, the methodologies of the two institutions 
differ drastically.  

The EA’s methodology of determining unemployment is based on a monthly registration through 
an “administrative system” (in labor offices) of jobless individuals and those on unemployment benefits 
(Employment agency). EA data have certain advantages related to the frequency of collecting, processing, 
and announcing – all done monthly. The information includes important demographic characteristics 
such as sex, age, level of education, location, etc. for each registered unemployed individual. In spite of the 
reliability of the data, collected by the EA, there are certain weaknesses in the approach the agency uses. 
Firstly, the data does not include unemployed individuals who are not registered with labor offices. 
Secondly, it is possible for some of the unemployed to register in order to receive unemployment benefits 
while simultaneously working on the grey market. Therefore, in practice, the unemployment level 
determined by the EA’s administrative system for registration may be far lower than its real value. 

Unlike the methodology of the EA, which is generally based on the voluntary registration of 
unemployed individuals in labor offices, the NSI methodology consists of observations of the labor force 
via a sample statistical survey including 2438 nests (census sections) and 19504 regular households. 
According to the methodology of monitoring of national statistics, in line with ILO recommendations and 
EUROSTAT requirements, unemployed are those individuals aged 15 to 74 who do not have a job in the 
calendar week of the survey but are actively looking for one over a four-week period and are available to 
start work within two weeks of the observed period (“Labor force survey” methodology, NSI). 

When it comes to calculating the unemployment rate, the two institutions of Bulgaria, the EA 
and the NSI, also use different methodologies. The EA calculates the unemployment rate as the ratio 
between the number of unemployed individuals and the number of economically active individuals (the 
labor force) as of the last census of the country, i.e., it uses the same basis for one period. On the other 
hand, the NSI does not use the numbers from the previous census, rather a dynamically changing basis, 
obtained as a result of quarterly surveys on a representative sample of the population. Therefore, the 
unchanging basis for calculating the unemployment rate used by the EA and the above remarks on the 
approach to data collection in practice make the NSI’s methodology more preferred and realistic. Aside 
from the general unemployment rate of the working age population in a given country, the 
unemployment rate is also calculated by age groups (aged 18-24; aged 25-34; aged 35-44; aged 45-54; 55 
and over, etc.), as well as by sex, place of residence and level of education.  

Since the first quarter of 2012, a new approach has been introduced in the weighing of the units 
from the labor force survey sample (“Labor force observation” sample, NSI), the influence of which needs 
to be taken into consideration when utilizing data from the labor force survey. This is why the scope of 
the survey begins from the following year – 2013, and continues until the latest available data (2021). 
 Since the first quarter of 2021 some changes has been implemented in compliance with the 

Regulation (EU) 2019/1700 of European Parliament and of the Council establishing a common 

framework for European statistics relating to persons and households, based on data at individual level 

collected from samples, and consequent implementing acts in the field of labour force statistics. These 

changes concern mainly employment and unemployment definitions, in particular:  

− persons on leave for looking after a child between 1 and 2 years of age who are receiving fixed 

compensation for the duration of the leave, are considered employed (they used to be considered 

economically inactive up to the end of 2020);  

− persons on unpaid parental leave for looking after a child between 2 and 8 years of age are 

considered employed only if the expected one-time duration of using that leave is at most three 

months. All persons on this kind of leave used to be considered employed up to 2020; 

− persons who are absent from work due to reasons other than holidays, illness, accident or 

maternity and parental leave are considered employed only if the duration of this absence is 3 

months or less (even they are being partially compensated);  



Finance, Accounting and Business Analysis 4 (1) 2022 

29 

 

− persons who produce agricultural goods for self-consumption are excluded from the employed 

person’s category even if they satisfy their household’s main consumption needs by that 

production. Employed are considered only persons growing agricultural produce, which main part 

is intended for sale or barter.  

 Due to changes in the LFS methodology data for the first quarter of 2021 are not fully comparable 

with those for previous periods. 

Methodoly for predicting the unemployment rate 

A forecast for the unemployment coefficient will be made in order to establish how it will 
develop over the following two years. To do so a trend model with the following function will be used: 

 

Y = ƒ(t) 

Where: 

Y – the unemployment coefficient; 

t – time as an independent variable. 

 

Trend models eliminate the effect of random and periodic causes and consider only the 
development trend. The functions of these models can take many forms: 

 

Linear model:      Cubic model: 

Ŷ𝒕 = 𝛃𝟎 + 𝛃𝟏𝒕      Ŷ𝒕 = 𝛃𝟎 + 𝛃𝟏𝒕 +𝛃𝟐𝒕²+𝛃𝟑𝒕³                     

Quadratic model:    Exponential model: 

Ŷ𝒕 = 𝛃𝟎 + 𝛃𝟏𝒕 +𝛃𝟐𝒕²    Y= 𝛃𝟎+(ɛ𝛃𝟏𝒕)                       

Logaritmic mocel:    Others…. 

Y = 𝛃𝟎 + 𝛃𝟏𝑳𝒏 t     

          

Where: 

𝛃𝒊 – parameters in the equation; 

Ŷ𝒕 – adjusted values of the unemployment coefficient. 

               

The model’s parameters can be found via the ordinadry method of least squares. This means that 
the sum of the squares of the residuals „ ɛk“ will be minimized.  

1
2

1

N

k

t


−

=

 = 
1

N

t=

 ( Yk - Ŷ𝑘 )² = min 

 

In order for this condition to be met, a system of equations needs to be solved.   
 

The forecasted value is reached by following these steps (Mishev and Goev, 2010): 

1. Performing an adequacy check on 10 trend models using an F-test and leaving only the 

adequate ones; 

2. Choosing the most adequate model for the trend – the one with the highest coefficient of 

determination ( R²); 

3. Constructing the point estimate of the forecast, which represents  the value of the 

dynamic array. The point forecast is obtained with a longer trend line in the future. 

The forecast is marked with с Ŷ𝑁+𝐿 , where N is the final period and L is the forecast horizon. The 
forecast will be realized using the program IBM SPSS.  
 
FINDINGS 

 
Analysis of unemployment rates among young people 

 There are different tendencies in the levels of youth unemployment for the period 2013-2021 

(Figure 1): 

 



Finance, Accounting and Business Analysis 4 (1) 2022 

30 

 

 
Source: NSI 

Figure 1. Unemployment rate among young people aged 15-29 

 

Due to the economic crises from 2008, unemployment rates from the discussed period amounted 
to 21.8. In 2014, the labor market in the country showcased signs of stabilization and youth 
unemployment, up to 2019, was going down with the lowest coefficient of 6.9%. The COVID-19 pandemic 
has had a direct effect in the increase of youth unemployment. The deteriorating economic situation in 
the nation increases the indicator by 3.5 percentage points up to 10.4%. 

Young people are most often the first ones to be terminated from employment in times of an 
economic recession. This can be seen from the following graph which illustrates the percentage change in 
the number of unemployed people aged 15-29 as a part of all the unemployed people in the country. The 
Eurostat data indicates that in 2013, youth unemployment had reached close to 30% from the total 
number of people aged 15-74. These high numbers are due to the economic crisis from 2008 which has 
left its mark. The percentage gradually decreased to 22% in 2020 and increased to 24% in 2021. 
 

 
Source: NSI 

Figure 2. Percentage change in the number of unemployed people aged 15-29 as a part of all the 

unemployed people in the country 

 

Youth unemployment began to rise further after 2019 and in 2020 their number increased by 
15% compared to 2019. In 2021, compared to 2020, the percentage increased by another 11%. In total, 
assuming that the rise in youth unemployment is due to the effect of COVID-19, then we can say that the 
pandemic has caused an increase in the number of the unemployed by more than ¼. 

21.800

17.700

14.400

12.200

9.900
8.300

6.900

8.800
10.400

.00

5.00

10.00

15.00

20.00

25.00

2013 2014 2015 2016 2017 2018 2019 2020 2021

4
3

6

3
8

5

3
0

5

2
4

7

2
0

7

1
7

3

1
4

3

1
6

9

1
7

1

1
2

9

1
0

0

7
8

6
1

5
1

3
9

3
2 3
7 4
1

2
9

.5
4

4
%

2
6

.0
3

4
%

2
5

.6
9

6
%

2
4

.8
3

8
%

2
4

.6
5

0
%

2
2

.5
6

2
%

2
2

.3
3

9
%

2
1

.7
6

7
%

2
3

.7
2

9
%

0

50

100

150

200

250

300

350

400

450

500

0%

5%

10%

15%

20%

25%

30%

2013 2014 2015 2016 2017 2018 2019 2020 2021

Total uneployment 15 - 29 г. Share of 15 - 29 г. from total unemployment

Number of 
people
in thousnads

Share                      



Finance, Accounting and Business Analysis 4 (1) 2022 

31 

 

Despite the observed trends of declining unemployment over the last few years, a serious issue for young 
people in the labor market is “long-term unemployment” (Eurostat, Statistics Explained). The main 
problem with this age group is the negative influence on future employment opportunities (Gregg and 
Tominey, 2007). It is considered that if a young person has been unemployed for a longer period of time, 
that leaves a bad impression on future potential employers, mainly due to an insufficient amount of 
knowledge, skills, and work experience (Ryan, 2001). 

 
Forecast 

 Based on the information on the unemployment rate from 2013 to 2021, a forecast of the 
coefficient for 2022 and 2023 will be made. As previously mentioned in the methodology section, this will 
be done on the basis of trend models. 10 trend models have been tested and their results are illustrated in 
Table 1 

 

Таble 1. Model Summary and Parameter Estimates 

Equation 

Model Summary Parameter Estimates 

R Square F df1 df2 Sig. Constant b1 b2 b3 

Linear .727 18.670 1 7 .003 19.867 -1.520   

Logarithmic .900 63.345 1 7 .000 21.427 -6.440   

Inverse .873 48.007 1 7 .000 7.240 15.991   

Quadratic .987 222.485 2 6 .000 27.212 -5.526 .401  

Cubic .993 247.056 3 5 .000 25.251 -3.649 -.045 .030 

Compound .711 17.216 1 7 .004 20.460 .891   

Power .836 35.632 1 7 .001 22.623 -.476   

S .757 21.856 1 7 .002 2.083 1.142   

Growth .711 17.216 1 7 .004 3.018 -.115   

Exponential .711 17.216 1 7 .004 20.460 -.115   

Source: Calculations by autors’ 

 
 The cubic model is the most appropriate one to make a forecast on as it has the highest coefficient 

of determination (R Square = 0,993). This is what it looks like:  

Ŷ𝑡 = β0 + β1𝑡 +β2𝑡²+β3𝑡³ 
 

А оценената му форма е: 

Ŷ𝑡 = 25,25 − 3,65𝑡 −0,05𝑡²+0,03𝑡³ 
 

 
Source: Calculations by autors’ 

Figure 3. Forecast on youth unemployment for 2022 and 2023 

 

 According to our forecast, the unemployment rate is expected to increase up to 14% in 2022. 
With a 95% probability, it can be stated that the coefficient will be between 11% and 17%. These 

.00

5.00

10.00

15.00

20.00

25.00

30.00

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Unemployment rate Forecast

Lower Confidence Limit Upper Confidence Limit



Finance, Accounting and Business Analysis 4 (1) 2022 

32 

 

numbers will continue to rise in 2023 and will reach 19%.  Once again, with a 95% probability, it can be 
stated that the coefficient will be between 14% and 25% The width of the interval is not very informative 
due to the short length of the dynamic array. Despite this, the indications we have observed since the 
beginning of the year showcase that forecast will likely come true. A long-term forecast is not advisable as 
the economic situation is currently shifting at a rapid rate and other long-term forecasts would be 
untenable 
 

Analysis of unemployment rates among young people at certain educational levels 

The data on the unemployment rate regarding young people who are at certain educational levels 
are also of great interest. Due to the aforementioned restrictions, the NSI does not maintain publicly 
available statistics regarding the work performance of people aged between 15-29. Therefore, we will 
once more use Eurostat statistics to analyze the unemployment at certain educational levels.  

In the case of youth unemployment, the levels of education undoubtedly have an impact. The 
importance of education among youth unemployment can be seen in figure 4. 

 

 
Source: Eurostat 

Figure 4. Youth unemployment rate at different educational levels (15-29) 
 

The highest percentage of unemployment can be found among young people who have the lowest 
possible degree of education. In 2013, it began at 44% and reached 16.8% before the pandemic 
(Eurostat). In 2020, there is a slight increase of the coefficient reaching 17.1% due to COVID-19, however, 
in 2021 it is much more tangible, reaching 24.2%. The lowest percentage of unemployment can be found 
among young people who have higher education. In the beginning of the period, the unemployment rate is 
3.3 times lower (13,5%) in comparison to the young people who only have primary education. At the end 
of the period, the unemployment rate becomes 4.5 times lower (5,2%) among young people who have a 
Bachelor's degree, a Master’s degree, and a PhD. What is more, this is the only educational level whose 
unemployment rate has started to rise in 2020 as a result of COVID-19 (from 4.1% to 5.5%), however, in 
2021 it started to decrease from 5.5% of 5.2%). All other levels of education mark an increase after the 
pandemic. 

  

DISCUSSION AND CONCLUSIONS 

 
In times of economic crises, one of the most logical reasons for an increase in unemployment 

among young people is the lack of job opportunities. A large part of the produce is reduced, which equates 
to job losses. In this case, young people are considered at risk and suffer the most in the labor market. 
According to our forecasts, the consequences of COVID-19 on the unemployment rate among young 
people are likely to increase over the next two years. What is more, the low level of education also 
contributes to this. According to the data for Bulgaria, in 2021 compared to 2019, unemployment among 
young people with primary and lower education has increased by 7.4%, with 4.2% among those who 
have acquired secondary education, and only with 1% among those with higher education. The analysis 
shows that young people in Bulgaria with higher education are at a considerably lower risk during the 

00

05

10

15

20

25

30

35

40

45

50

2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 2 0 1 8 2 0 1 9 2 0 2 0 2 0 2 1

Primary education

Secondary education

Higher education



Finance, Accounting and Business Analysis 4 (1) 2022 

33 

 

COVID-19 pandemic in comparison to those with lower education. In addition, for the individuals with a 
low level of education, the salary for their work is lowered, close to minimum wage, which naturally leads 
to a lower standard of living. In general, these young people often become directly dependent on the 
social benefits of the Social Welfare System and thus fall into another group connected to the 
phenomenon known as the “welfare trap”.  
 
REFERENCES 
 

1. Eurostat, official website - www.ec.europa.eu/eurostat 
2. Eurostat, Statistics Explained, Long-term unemployment, direct link:  

https://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:Long-
term_unemployment 

3. Gregg, P. and Tominey, E. (2007). The wage scar from youth unemployment, Elsevier, 
Netherlands, 2005, p. 487-509, The Prince’s Trust, The Cost of Exclusion. Counting the cost of 
youth disadvantage in the UK, London 

4. Henehan, K. (2021). Resolution Fondation, Changes in youth unemployment and study since the 
onset of Covid-19: https://www.resolutionfoundation.org/app/uploads/2021/04/Uneven-
steps.pdf  

5. Henehan, К. (2021). Uneven steps: Changes in youth unemployment and study since the onset of 
Covid-19, 

6. Katris, К. (2021). Unemployment and COVID-19 Impact in Greece: A Vector Autoregression (VAR) 
Data Analysis, Engineering Proceeding, vol. 5 (41) https://doi.org/10.3390/ 
engproc2021005041  

7. Mishev, G., Goev, V. (2010) Statistical analysis of time series, Avangard Prima, Sofia 
8. NSI, Demographic and social statistics, population and demographic data, official website - 

www.nsi.bg 
9. NSI, Labor Force Survey Methodology, official website - www.nsi.bg 
10. Pavlov, N. (2008). Labor standards and unemployment, Faber, Sofia, p. 59 
11. Ryan, P. (2001). The School-to-Work Transition: A Cross-National Perspective. Journal of 

Economic Literature, 39, pp. 34–92 
12. The Employment Agency, official website - www.az.government.bg 

 

http://www.ec.europa.eu/eurostat
https://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:Long-term_unemployment
https://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:Long-term_unemployment
https://www.resolutionfoundation.org/app/uploads/2021/04/Uneven-steps.pdf
https://www.resolutionfoundation.org/app/uploads/2021/04/Uneven-steps.pdf
http://www.nsi.bg/
http://www.nsi.bg/
http://www.az.government.bg/

