Abstract 5 Introduction 6 1. Overview of the model 7 1.1. Literature review 7 1.1.1. Related published paper 7 1.1.2. Research gap 8 1.2. Theoretical framework 8 1.3. Empirical model 10 2. Model specification and data 13 2.1. Methodology 13 2.1.1. Method used to collect data 13 2.1.2. Method used to analyze data 13 2.2. Data description 13 2.2.1. Data description and interpretation 13 2.2.2. Correlation matrix between variables 15 3. Estimated model and statistical inferences 16 3.1. Testing the econometric model 16 3.2. Estimated model 17 3.3. Statistical inference 18 3.3.1. Testing for the overall significance of the model 18 3.3.2. Testing for the statistical significance of the individual coefficients 19 3.4. Testing for violations of the model 20 3.4.1. Detection of multicollinearity 20 3.4.2. Detection of heteroskedasticity 20 3.4.3. Detection of serial correlation – autocorrelation 21 3.4.4. Detection of crosssectional correlation – autocorrelation 21 3.4.5. Detection of omitted variables 22 3.5. Fixing the model 22 4. Recommendations 25 4.1. Policy implications 25 4.2. Limitations and future research 26 Conclusion 27 References 28 Dofile 29
https://tailieuluatkinhte.com/ FOREIGN TRADE UNIVERSITY FACULTY OF INTERNATIONAL ECONOMICS - RESEARCH PAPER Determinants affecting underground economy of european union from 2008 to 2021 Student group : Class : KTEE318 Module Instructor : Ph.D Dinh Thi Thanh Binh Ha Noi, June 2023 https://tailieuluatkinhte.com/ Table of Contents Abstract Introduction Overview of the model 1.1 Literature review .7 1.1.1 Related published paper .7 1.1.2 Research gap 1.2 Theoretical framework 1.3 Empirical model 10 Model specification and data 13 2.1 Methodology 13 2.1.1 Method used to collect data 13 2.1.2 Method used to analyze data 13 2.2 Data description 13 2.2.1 Data description and interpretation 13 2.2.2 Correlation matrix between variables 15 Estimated model and statistical inferences 16 3.1 Testing the econometric model 16 3.2 Estimated model 17 3.3 Statistical inference 18 3.3.1 Testing for the overall significance of the model 18 3.3.2 Testing for the statistical significance of the individual coefficients 19 3.4 Testing for violations of the model 20 3.4.1 Detection of multicollinearity 20 3.4.2 Detection of heteroskedasticity 20 3.4.3 Detection of serial correlation – autocorrelation 21 https://tailieuluatkinhte.com/ 3.4.4 Detection of cross-sectional correlation – autocorrelation .21 3.4.5 Detection of omitted variables 22 3.5 Fixing the model 22 Recommendations 25 4.1 Policy implications 25 4.2 Limitations and future research 26 Conclusion 27 References 28 Do-file 29 https://tailieuluatkinhte.com/ Abstract The underground economy, also known as the informal, unofficial, or shadow economy, includes illegal activities and unreported income, from monetary or barter transactions The relationship between the increase in the unofficial sector and economic growth has yet to be clearly demonstrated by theoretical research The main problem is that it would reduce the economy’s growth rate The seven explanatory variables including GDP Growth rate, Tax Rate, Index of Economic Freedom, Unemployment Rate, Consumer Price Index, and Internet Users will be examined to test their relationship with the underground economy In order to examine the correlation between the mentioned variables, panel data was gathered from 27 European countries between 2008 and 2021 Based on the results, quantitative analysis and explanation will be done to show the influence of each, the unaffected factors, and other removed ones from the model Furthermore, other recommendations and perspectives of writers are offered for future policy-making of governments Keywords: underground economy, European Union https://tailieuluatkinhte.com/ Introduction The mainstream economy and the underground economy have coexisted side by side for as long as human civilization has existed The underground economy has a big influence on a nation's overall economy even if it is not widely acknowledged The underground economy goes by many names: shadow, informal, unobserved, unrecorded or unofficial economy The underground economy, in contrast to its mainstream counterpart, involves not only illicit activity but also unreported earnings from the creation of legal products and services, whether through monetary or barter exchanges It is challenging to determine the size of underground economies because, by their very nature, they are not subject to governmental control; as a result, the economic activity neither generates tax returns nor appears in official statistical reports However, even though the transactions are undetected, keeping track of outgoing expenses can provide a sense of statistics In other words, spending that is not reflected in documented transactions presumably reflects the scope of black market activities.in documented transactions presumably reflects the scope of black market activities In order to transfer the bulk of the illegal enterprises into the legal sector and enable the government and policy framework, it is crucial to examine this issue in detail, identify the core factors that determine its size and growth, and develop policies that target those causes The issues with the shadow economy are too numerous and detrimental to overlook, and their continued presence will ultimately lower overall tax receipts and harm the macroeconomic policy framework For this reason, we made the decision to investigate the elements influencing the economy of 27 nations over a 13-year period This paper will contribute to a deeper understanding of those important elements, including their significance for the operation of the shadow economy as well as their involvement in the economies of the various nations In this report, the content is organized as follows: The first part will be the state of knowledge in this field, the second part will present the model specification and data analysis and validations in the determinants of the shadow economy The third part will be about the estimated model and statistical inferences And the final part will summarize all the results as well as present the limitations and directions for future research https://tailieuluatkinhte.com/ Overview of the model 1.1 Literature review 1.1.1 Related published paper The very first beginning of studying the underground economy defined it as the output of products and services (legal or illegal) not included in the official estimates of GDP (Smith, 1994) Schneider (1994) and Lubell (1991) believed that it is the activity that is not included in the computation of the gross national product As the confirmation of choosing the explained variable, Underground Economy Size was used by Friedrich Schneider’s studies (2004, 2005, 2010) The widespread of this estimation has been recalculated by Mai Hassan and Friedrich in 2016 to give further information about the extent of the underground economy First and foremost, the tax rate is mentioned in most of the articles about this topic The higher the overall tax rate or lower monitoring, the stronger motivation for tax evasion and underreporting wages (Schneider and Williams 2013, Hassan and Schneider 2016) The tax burden can affect labor and increase the labor supply for the informal economy in which they want to lower the tax wedge and work in the informal field to minimize the gap between official wages and after-tax earnings (Thomas 1992; Johnson, Kaufmann, and Zoido-Lobatón 1998a, b; Dell’Anno, Gomez-Antonio and Alanon Pardo 2007) The connection between unemployment and the size of the underground economy should be taken into consideration There may be a negative impact as the increase in the unemployment rate would lead to a decline in both the formal and informal sectors (Gulzar et al., 2010; M Hassan & Schneider, 2016) The inflation rate has been indicated as explanatory for the underground economy (Erdinỗ, 2016) The tax rate and tax distortions can be reduced since the government no longer needs to impose high taxes in order to generate more revenue through seigniorage Lower taxes encourage formal sector activity as opposed to production for the illegal market (Gulzar, Junaid, & Haider, 2010) Another regulation from the government can be known as limiting people’s freedom, it raises the labor costs and incentive people to work in the shadow sector The burden placed on businesses and people leads them to participate in the underground economy (Schneider 2011; Hassan and Schneider 2016) Internet use is proven to have a negative effect on the growth of the underground economy Better insight into the drastic consequences of corruption or tax evasion can https://tailieuluatkinhte.com/ be acknowledged through using the internet Awareness of risks would mitigate the possibility of joining illegal activities (Elbahnasawy, 2014; Elgin, 2012) Corruption contributes a lot to the underground sector, consequently, reducing it results in the decrease of the shadow economy size Another key factor is GDP growth as the higher the GDP growth, the lower the motivation to work in the shadow economy, ceteris paribus (Elbahnasawy, 2014; Elgin, 2012) However, due to the strict regulations and standards of government, it is challenging for businesses and individuals to comply Therefore, GDP performance may not only enhance development but also drive people into the unofficial sector A very important aspect when measuring the underground economy is population growth, while there is no direct link As population growth develops, it fuels corruption which leads to an increase in the informal sector 1.1.2 Research gap Studying the underground economy is troublesome in that it includes illegal activities and lacks evidence and data available to carry out research The underground economy is an emerging topic, compared to other economic issues, leading to the scarcity of research papers and proper methods to investigate Moreover, the unofficial sectors also depend on other explanatory variables that have neither been detected nor investigated to examine the correlation, leading to insufficient knowledge 1.2 Theoretical framework The causes and factors that drive the shadow economy are various and numerous The majority of the research has established a wide variety of elements that impact and aid in the operation of this type of economy Most studies on the underground economy (2004, 2005; 2010) employ Friedrich Schneider's estimates of it as a dependent variable These estimations are often used, which shows their accuracy The most current estimates of the size of the underground economy from Mai Hassan and Friedrich Schneider's study (2016) were used to put the results of this study into context as they are among the most recent estimates and have not been used in any prior studies Tax income and tax rates are among the most important aspects of the shadow economy that are included in most literature publications The bigger the incentive for tax evasion and underreporting of salaries, the higher the total tax burden and/or lower the monitoring and enforcement (Schneider and Williams 2013, Hassan and Schneider 2016) The distortion caused by the total tax burden affects leisure time spending https://tailieuluatkinhte.com/ decisions and might boost the availability of workers in the shadow economy This tax wedge affects both the overall tax burden and the social security burden/payments, making them crucial components of the sustainability of the shadow economy (Thomas 1992; Johnson, Kaufmann, and Zoido- Lobatón 1998a, b; Giles 1999a; Tanzi 1999; Schneider 2003, 2005; Dell’Anno 2007; Dell’Anno, Gomez-Antonio and Alanon Pardo 2007) There is evidence from several papers that the underground economy and inflation rate are related Earnings from seigniorage rise in tandem with inflation Because the government no longer has to levy high taxes to increase income through seigniorage, both the tax rate and tax distortions may be decreased Lower taxes promote activity in the legal economy as opposed to manufacturing for the black market The bulk of past research (Erdinỗ, 2016; Gulzar, Junaid, & Haider, 2010; Schneider & Bajada, 2003) employed the consumer price index to calculate inflation The relationship between the extent of the shadow economy and unemployment is up for debate Given that an economic downturn would increase unemployment in both the official and informal sectors, there may be a negative correlation The bulk of earlier studies have calculated unemployment using the unemployment rate as a percentage of the labor force According to other studies (Gulzar et al., 2010; M Hassan & Schneider, 2016; Kanniainen, Päkkönen, & Schneider, 2004; Saafi, Farhat, & Haj Mohamed, 2015; Sarac, 2012; Savasan, 2003), other factors, notably the level of education, have an effect on the extent of the underground economy The literature claims that restrictions on people's freedom (of choice) in the formal economy include rules controlling the labor market and trade barriers As a result, people are more motivated to work in the shadow economy, which is why countries with greater levels of regulation often have a higher share of the shadow economy in their overall GDP (Johnson, Kaufmann, and Shleifer 1997; Johnson, Kaufmann, and Zoido-Lobatón 1998b; Friedman, Johnson, Kaufmann, and Zoido- Lobatón 2000; Kucera and Roncolato 2008; Schneider 2011; Hassan and Schneider 2016) Internet usage has been shown to be negatively correlated with the expansion of the shadow economy As more individuals utilize the internet, they will become aware of the severe repercussions of actions like corruption or tax evasion Due to increased knowledge and individual action to mitigate and scale back the hazards associated with such activities, the underground economy will decline (Elbahnasawy, 2014; Elgin, 2012; Goel, Nelson, & Naretta, 2012; Shrivastava & Bhattacherjee, 2014) https://tailieuluatkinhte.com/ Although there is no direct correlation between population growth and the shadow economy per se, corruption is a fundamental element of this dynamic, making it a highly important consideration when attempting to quantify the shadow economy As a result, there is a tenuous link between population increase and the shadow economy The shadow economy grows as a result of population expansion, which also encourages corruption In this research, we will evaluate seven major factors that operate as driving forces in the underground economy Following extensive study on the issue of the underground economy, we would like to propose the following independent factors that may operate as driving forces: 1.3 Empirical model Definition of variables - Underground economy (SUE) Underground economy, also known as a shadow, informal, or parallel economy, refers to economic activities not reported to the authorities in order to avoid taxes, social security contributions, labor laws, and regulations, and costs related to regulations The underground economy includes not only illegal activities but also unreported income from the production of legal goods and services, either from monetary or https://tailieuluatkinhte.com/ barter transactions There is no precise definition of the underground economy because of its development and changes in regulations and taxation over time (IMF, 2002) - GDP growth (GDPGR) Gross Domestic Product (GDP) is the total value of goods and services produced within a nation’s territory during a specific period of time GDP growth is mainly used to evaluate the economic growth of a nation The formula to calculate the GDP growth is: GDP Growth = - GDP2−GDP GDP1 CPI Consumer Price Index (CPI), published by the Bureau of Labor Statistics (BLS) is a measure of the overall level of prices based on the typical consumer’s “basket of goods” A basket of goods refers to goods associated with the cost of living such as transportation, food, medicine, energy, etc CPI is used to measure inflation, track changes in the typical household’s cost of living, and make comparisons of the value of domestic currency over time CPI in any month equals 100x Cost of the basket ∈that month Cost of thebasket ∈base period - Unemployment rate (UER) The unemployment rate measures the shares of workers in the labor force who are not currently employed but are actively seeking a job The formula to calculate the unemployed rate is Unemployment rate = Unemployed people x 100 Total labor force - Internet user (IUR) In the context of the survey on Internet use within households, an Internet user is defined as someone who has used the Internet within the last three months, while a regular Internet user is defined as someone who has used the Internet at least once a week within the reference period of the survey (the first three months of the calendar year), regardless of locations to use - Tax revenue rate (TRR) Tax revenue is defined as the funds collected from taxes on income and profits; Social Security taxes or “contributions”; taxes levied on goods and services, generally 10 https://tailieuluatkinhte.com/ IUR -0.28056945*** -0.07550494*** -0.07685027*** TRR 0.36090658*** -0.01144128 -0.00346089 POPG 2.421944*** 0.19461488 0.21986582 IEF -0.14615139* 0.02685206 0.02180726 _cons 44.007552*** 22.327339*** 22.447841*** Number of obs 369 369 369 R^2 0.34363033 0.33101562 Legend: * p < 0.05; ** p < 0.01; *** p < 0.001 Breusch and Pagan Lagrangian multiplier test: xttest0 chibar2(01) = 1532.79 Prob > chibar2 = 0.000 Hausman test to choose between FEM and REM: Hausman FE RE, sigmaless chi2(7) = 18.81 Prob > chibar2 = 0.0088 From Breusch and Pagan test xttest0: Prob > chibar2 = 0.000 < 0.05 → exists Therefore, we not use the POLS model to estimate Using the Hausman test to choose between FEM and REM: Prob > chibar2 = 0.0088 < 0.05 We choose FEM 3.2 Estimated model Command: xtreg SUE GDPGR CPI UER IUR TRR POPG IEF, fe We got the following results: SUE Coefficient SE t p-value GDPGR -0.0553499 0.0179004 -3.09 0.002 16 95% confidence interval -0.0905612 -0.0201386 https://tailieuluatkinhte.com/ CPI 0.0340762 0.0407384 0.84 0.403 -0.0460592 0.1142116 UER 0.0786969 0.0290463 2.71 0.007 0.0215607 0.135833 IUR -0.0755049 0.0088773 -8.51 0.000 -0.0929672 -0.0580427 TRR -0.0114413 0.0334468 -0.34 0.733 -0.0772335 0.054351 POPG 0.1946149 0.169558 1.15 0.252 -0.1389178 0.5281475 IEF 0.0268521 0.0321443 0.84 0.404 -0.036378 0.0900822 _cons 22.32734 2.501232 8.93 0.000 17.40724 27.24744 R^2 0.1082 Prob > F 0.0000 corr(u_i, Xb) 0.1720 ^ SUE = 22.32734 - 0.0553499GDPGR + 0.0340762CPI + 0.0786969UER - 0.0755049IUR - 0.0114413TRR + 0.1946149POPG + 0.0268521IEF We have: ^β = 22.32734, means that when the value of independent variables equals 0, the size of underground economy equals 22.32734 ^β = -0.0553499, means that when GDP growth rate (GDPGR) increases by 1%, the size of underground economy will decrease approximately by 0.055% (other factors remain unchanged) ^β = 0.0340762, means that when CPI increases by 1%, the size of underground economy will increase approximately by 0.034% (other factors remain unchanged) ^β = 0.0786969, means that when the Unemployment Rate (UER) increase by 1%, the size of underground economy will increase approximately by 0.078% (other factors remain unchanged) ^β = -0.0755049, means that when the Internet Users (IUR) increases by 1%, the size of underground economy will decrease approximately by 0.075% (other factors remain unchanged) ^β = -0.0114413, means that when the Tax Revenue (TRR) increases by 1%, the size of underground economy will decrease approximately by 0.011% (other factors remain unchanged) 17 https://tailieuluatkinhte.com/ ^β = 0.1946149, means that when the Population Growth rate (POPG) increases by 1%, the size of underground economy will increase approximately by 0.194% (other factors remain unchanged) ^β = 0.0268521, means that when the Index of Economic Freedom (IEF) increases by 1%, the size of underground economy will increase approximately by 0.026% (other factors remain unchanged) 3.3 Statistical inference 3.3.1 Testing for the overall significance of the model We construct the following hypothesis and use the p-value method to evaluate the overall significance of the model, considering the significance level of 5% Hypothesis: H0: R^2 = H1: R^2 ≠ From the estimated model: p-value < α (0.000 < 0.05) Thus, Ho is rejected Conclusion: The model is significant at α = 5% 3.3.2 Testing for the statistical significance of the individual coefficients We construct the following hypothesis for each coefficient and use the p-value method to evaluate the significance of estimated coefficients of estimators, considering the significance level of 5% Hypothesis: H 0: β j = H 1: β j ≠ Testing results: - GDPGR: p-value < α (0.002 < 0.05), reject H 0, thus the coefficient is statistically significant, which means GDP growth has an impact on underground economy size - CPI: p-value > α (0.403 > 0.05), can not reject H 0, thus the coefficient is statistically insignificant, which means the inflation rate has little impact on underground economy size - UER: p-value < α (0.007< 0.05), reject H 0, thus the coefficient is statistically significant, which means the unemployment rate has an impact on underground economy size 18 https://tailieuluatkinhte.com/ - TRR: p-value > α (0.733 > 0.05), can not reject H 0, thus the coefficient is statistically insignificant, which means the tax revenue rate has little impact on underground economy size - IUR: p-value < α (0.000 < 0.05), reject H 0, thus the coefficient is statistically significant, which means the number of internet users has an impact on underground economy size - POPG: p-value > α (0.252 > 0.05), can not reject H 0, thus the coefficient is statistically insignificant, which means population growth has little impact on underground economy size - IEF: p-value > α (0.404 > 0.05), can not reject H 0, thus the coefficient is statistically insignificant, which means the economic freedom index has little impact on underground economy size Conclusion: At α = 5%, 3/7 independent variables are statistically significant 3.4 Testing for violations of the model 3.4.1 Detection of multicollinearity We used the Variance inflation Factor (VIF) method to detect whether the model has multicollinearity or not If existing at least one value of VIF greater than 10, the model contracts this defect With the vif command in STATA, we have the result as follow: Variable VIF 1/VIF UER 1.60 0.634348 IUR 1.50 0.667328 POPG 1.44 0.692776 IEF 1.33 0.751147 CPI 1.29 0.777595 TRR 1.19 0.838795 GDPGR 1.09 0.918301 Mean VIF 1.35 19 https://tailieuluatkinhte.com/ As all the VIF value of independent variables are less than 10, there is no multicollinearity in the model 3.4.2 Detection of heteroskedasticity Hypothesis: H 0: The variance of disturbance is constant (Homoskedasticity) H 1: The variance of disturbance is not constant (Heteroskedasticity) With the significance level of 5% Using the Modified Wald test in STATA with xttest3 command to detect whether the model incurs heteroskedasticity or not, we have the result as follow: H0: sigma(i)^2 = sigma^2 for all i Chi2(27) = 3850.33 Prob>chi2 = 0.0000 As p-value = 0.0000 F = 0.0000 20