Impact Of Tax Revenue Components On Economic Indicators Economics Essay

THE IMPACT AND CONSEQUENCES OF TAX REVENUE’S COMPONENTS ON ECONOMIC INDICATORS: EVIDENCE FROM PANEL GROUPS DATA

ABSTARCT

The change in components of tax revenue may result the change in the economic growth and other economic indicators. The previous studies found the significant effect of the change in tax revenue to the economic growth in a country. This study is the research to test if the components of tax revenue and inflation rate are related to the economic growth, gross saving, and inflow of FDI in the 120 countries based on the different level of income from 1960 to 2009. The impact of tax components such as total tax revenue, taxes on income, profit and capital gain, taxes on goods and services, taxes on international trade, total tax rate, and taxes on export are examined with the three economic indicators. These three indicators are highly significant to the tax structures in four groups of country. Change in the tax revenue or tax policy for every single of tax will also change the growth of GDP, gross saving and FDI. The results indicate the different impacts and consequences of type of taxes for the four groups of country. This study conclude that tax to GDP ratio has negative impact to growth in low income countries but positive impact in high income countries that relate with the different inflation rate and tax policy in the country.

Keywords: tax; economic indicators; economic growth; GDP; Gross saving; Foreign direct investment ; Tax components.

INTRODUCTION

Taxes are one of the major revenue for a country in which collect taxes from citizens, company, investment and so on to generate economy. There have several impacts of taxes due to economic growth weather it is positive or negative impact. According to the theory of tax competition, the government will reduce the taxes on mobile asset when the globalization increases due to raise the economic growth in a country. Change in tax rate also will give the different impact to an open economy. According to Bretschger (2010), he was found the negative impact of corporate taxes on openness and total tax revenue to the economic growth in 12 OECD countries. He also mention about the tax competition theory that was agreed when tax rate of capital reduce, it will cause the capital inflow to a country. It’s because of the tax rate is one of the cost for capital holder, see Bucovetsky (1991) and Wilson (1991). These two researches were found that private return on investment will influenced by the changes in capital taxes.

More than 20 studies their looking at for evidence on tax rates and economic growth in the United States and internationally. With all of the studies, they come with concluded that reduce all marginal rates by 5 percent and average tax rates by 2.5 percent leads to increase 0.2 percent to 0.3 percent of long-term growth (1996: 34). Christina and David (2007) conduct the impact of changes in the level of taxation on economic growth in which their investigated the effects of tax on GDP in United State in the post-World War II period. The study found that a tax increase by 1 percent leads to reduced 2 percent to 3 percent of GDP in United State. According to Engen and Skinner (1996), reduce 5 percent in tax rate will increase 0.25 percent on growth in U.S.

However, some of the studies give opposite results in term of the negative relationship between tax and economic growth. According to Uhliga and Yanagawa (1999), increase capital income taxes will generate the economy. It’s because, the capital income accrues for the old, in which increase on the capital income taxes will burden tax for the young and increase their saving, if the interest elasticity of saving is low. The other study by Glomm and Ravikumar (1998), find that when the government reduce the capital income taxes, it will reduce the spending on education and the long-run growth. In this case, the capital income taxes have positive correlation with the economic growth. Besides that, Gober and Burns (1997) have done a study about the relationship between tax structure and economic indicators for 18 industrial or OECD countries. From their finding, total tax revenue has negative relationship with two economic indicators that are saving and investment. However, according to them, personal income tax, corporate income tax, sales tax (consumption tax) and other taxes are highly significant, in which have positive relationship with economic growth (GDP or GNP).

This study also involves another variable that was effect the collection of tax revenue in a country. The variable is inflation rates. Usually, if the country faces the economic crisis, the government will try to recover the problem using monetary or fiscal policy. In fiscal policy, the government will use either taxes or government spending based on the problem. High inflation rate in a country will force the government to increase the tax of goods and services due to increase the price and stabilize the consumption also aggregate expenditure. With that, excise tax on some products may be affected with the change in inflation rate, see Tanzi (1989).

LITERATURE REVIEW

The previous empirical studies found that most of the tax structures were highly significant and related with the economic growth in a country. One of the early studies done by Marsden (1983) mention that change in tax policy will affect the economic planning. He was found the different effect of tax to GDP ratio on growth in low and high income countries. The low income countries show the negative impact of tax to GDP ratio on growth but positive impact in the high income countries. According to Gober and Burns (1997), economic in a country may affect differently due to any changes in each tax components. Based on their finding, excise taxes as percentage of Ireland’s total revenue was four times the level in U.S. Change in economic growth is depend on each of tax structures (Gold, 1991). Mahdavi (2008) suggest that the effect of rises in total tax revenue will reduce the growth in developing countries. Causes by the fiscal crisis in the past several decades, several developing countries have to recover its economic by change the level of taxes. Two of the early studies by Hinrisch (1966) and Musgrave (1969) were examined the relationship between the ratio of tax revenue to GDP (TAX/GDP) and found it was relatively low in the developing countries.

One of the studies that focus on African countries by Leuthold (1991) was examined the effect of (TAX/GDP) from 1973 to 1981 use OLS estimation method. From his study, the share of agriculture will affect the level of taxation and robust the relationship of total tax revenue into direct and indirect taxes. The level of taxes will give the different effect to growth and other indicators causes by the macroeconomic variables such as extent of corruption and adversely affected by the inflation rate in nine African countries over the period 1985-96, see Ghura (1998). Agbeyegbe (2004) has been used the same geographical sample, in which sample of 22 countries in sub-Saharan Africa from 1980 to 1996. The study was examined the effect of tax revenue on trade liberalization or “openness”. He is focused on three components in total tax revenue (taxes on income, international trade, also goods and services tax) that all as ratio of GDP and found the weakly related among these three tax types.

There have three of the previous studies that mention the relationship between taxation, especially personal income taxes and corporate income taxes on growth. According to Arisoy and Unlukaplan (2010) that study the tax composition and growth in Turkey found that corporate income taxes are the most harmful for growth in OECD (2008) followed by personal income taxes and consumption taxes. Corporate income taxes has positive relationship while personal income taxes has negative relationship on growth in 23 OECD countries, see Widmalm (2001). The other study that found the impact of corporate income taxes on growth was conducted by Lee and Gordon (2005). They were studied the relationship among taxation and growth using cross sectional and time series data for the period of 1970-1997 and found the negative effect of corporate income taxes on growth in a country.

The potential variable that will change the amount of tax revenue is change in inflation rate in a country. High inflation rate in a country will force the government to increase the taxes on goods and services due to increase the price and stabilize the consumption also aggregate expenditure. With that, excise tax on some products may be affected with the change in inflation rate, see Tanzi (1989). A study by Mahdavi (2008) was mention the effect of income, profit and capital gain tax due to change in inflation rate and investment plans. Based on his study, when the inflation rates increase, the household will protect their assets by substitute it with the assets that less domestically taxes such as jewellery items. According to Ashworth and Heyndels (2002), real economic growth and inflation rate in a country will affect the change in tax structure. It’s because each of the tax components will respond to the change in growth and inflation rate, in which will force the government to reform the different tax policy. The impact of inflation rate on tax was study by Messere (1993), in which found an increase in income taxes it cause by increase in inflation, while the consumption taxes was remain constant and not affected by change in inflation rate. The other study that focuses in OECD countries by Kemmerling (2003) also found the same result, in which income taxes and inflation rate has positive relationship.

Effect on Foreign Direct Investment (FDI)

The other variable that included in this research is to study the effects of taxes on FDI. Usually, the tax rate on capital was measure by stock of capital or capital flows that related to the FDI. One of the earliest studies by Hartman (1984), in which he was studied the relationship between FDI, after-tax rate of return by foreign investors and capital in United State. From his study, he suggests that the taxes have a strong relationship with FDI. The tax regimes in Mexico and United State have quite responds to the U.S’s FDI. Two of these earliest studies give the direction about the impact of taxes on FDI. Based on Scholes and Wolfson (1992), tax will affect the decision of foreign investors to make investment in a country cause by the changing in rates of return on assets. They argue that higher in tax will reduce the rates of return and discourage the FDI in-flow to a country. Hines (1999) found that FDI is sensitive with the tax, in which high tax rates can change the foreign investment rapidly. He concludes that reduce only 10 percent on tax rates will increase more than 10 percent in FDI. FDI determination and corporate tax was conducted by Ghinamo, Panteghini, and Revelli (2007) show that corporate tax rate was effect the inflow of FDI in a country. They also found the corporate income tax to personal income tax ratio are lower in developing countries than developed countries especially OECD countries.

DATA AND METHODOLOGY

This study follows Gober and Burns (1997), Mahdavi (2008) and Gordon and Li (2009). The dependent variables are change in GDP, ratio of gross saving to GDP (SAVING/GDP) and foreign direct investment as ratio to GDP (FDI/GDP) to relate it with all the component of taxes (independent variables) since 1960 to 2009. This study is using panel data approach. Table 1 describes all the variables that include in this study.

Table 1: Definition and sources of variables.

Variable

Definition

GDP

FDI

SAV

IPCT

GST

ITT

ET

TTR

TTX

INF

Gross Domestic Product annual percentage change.

Foreign Direct Investment (inflow) divided by GDP.

Gross Saving divided by GDP.

Taxes on income, profit and capital gain divided by total tax revenue.

Taxes on goods and services divided by total tax revenue.

Taxes on international trade divided by total tax revenue.

Taxes on export divided by total tax revenue.

Total tax rate divided by total profit.

Total tax revenue divided by GDP.

Inflation rate (consumer price index).

Notes: All the data was collected from World Data Catalogue (World Bank, 2010).

Panel unit root test and Correlogram

This cointegration test is to test for the existence of unit root for these variables. This study use Levin, Lin, and Chu test (2002), The Im, Pesaran and Shin (IPS, 1997) test, and Breitung (2000) test. Correlogram is use to detect for the randomness of data set based on value of Autocorrelation function (ACF) and Partial Autocorrelation function (PACF).

Variance inflation factors (VIF)

This method can detect the multicollinearity problem in the regression model that examines the value of variance. Based on this method, the value of variance will increase due to collinearity. According to Wooldridge (2002), variance inflation factor (VIF) can be shown as follows:

VIF (i) = 1/ (1- R²) (1)

The higher VIF, the greater of finding i insignificant that will detect the problem of multicollinearity.

Breusch and Pagan LM test

This method is used to detect the model regression for heteroscedasticity problem. This test also has been used to know either the panel data can be pooled in random or fixed effects models, see Breusch and Pagan (1979), Zaman (2000), and ‚iorn (2009). Breusch and Pagan state the null hypothesis of homoscedasticity, H0: € €½€ €° given that:

€ €½€ t ) (2)

in which zt represent the t-th row of the matrix regressors Z being tested as independent variables for the variance (.The significant in Breusch and Pagan LM test (based on probability of chi squares) will reject the null hypothesis and indicates that the data can be pooled in random or fixed effects models.

Hausman test

This test is important test to select the fixed or random effects in panel data approach. The Hausman statistic (1978) was used to test the consistency of Generalized Least Squares (GLS) estimator, see Hausman and Taylor (1981), Cornwell and Rupert (1988) and Baltagi and Akom (1990). The Hausman test (1978) is based on the difference between within-regression (βW) and generelized-regression (βG) with the alternative hypothesis as shown below:

H1 = (W – G) [Cov (W)]-1 (W – G) (3)

Significant of Hausman test (based on probability chi-squared) means that fixed effects model is suitable for the model of panel data approach.

The model regression function of panel data approach is shown below:

Ejit = β0it + β1(TTX)it + β2(IPCT)it + β3(GST)it + β4(ET)it + β5(ITT)it + β6(TTR)it + β7(INF)it + ­it (4)

where the jth represent economic indicators (GDP, SAV, and FDI) in the ith country at time t (Ejit).

HYPOTHESIS

Hypothesis one

For the first hypothesis, the dependent variable is annual change of GDP and the independent variables are the component of taxes as ratio to total tax revenue, total tax rate, and total tax revenue as ratio to GDP. The first hypothesis stated in null form is:

H0: There is no relationship between tax components and inflation rate on GDP for a given countries based on the level of income.

H1: There is a relationship between tax components and inflation rate on GDP for a given countries based on the level of income. .

Hypothesis two

Hypothesis two estimates the ratio of gross saving to GDP as dependent variable due to the ratio of each components of tax revenue to total tax revenue and inflation rate. The null hypothesis is:

H0: There is no relationship between tax components and inflation rate on saving for a given countries based on the level of income.

H1: There is a relationship between tax components and inflation rate on saving for a given countries based on the level of income.

Change of the tax components such as goods and services tax, consumption tax, personal and income tax will also affect the gross saving in a country. It’s based on the previous study by Burns and Gober (1997) that study about the relationship between tax structure and economic indicators using the cross sectional data of OECD’s countries. However, the affect is not constant with the different level of income for the countries.

Hypothesis three

For this third hypothesis, the dependent variable is foreign direct investment (FDI) as ratio to GDP and the independent variables are the components of tax revenue as ratio to total tax revenue and inflation rate. The null hypothesis form is:

H0: There is no relationship between tax components and inflation rate on FDI for a given countries based on the level of income.

H1: There is a relationship between tax components and inflation rate on FDI for a given countries based on the level of income.

Based on the previous studies, inflow of FDI also was influence by the components of tax revenue in a country. Usually, the return that the investors will get is earning per share (EPS) or dividend per share (DPS) by the company. The investors was attracted by higher EPS or DPS, in which both of this returns were influenced by inflation rate and tax rates. The formula of EPS can be shown as:

EPS = (5)

With that equation, increase in rates of corporate income tax and taxes on profit will reduce the EPS and discourage investors to invest.

FINDINGS: LOW, LOWER MIDDLE, UPPER MIDDLE, AND HIGH INCOME COUNTRIES

The result based correlogram shows that all the variables are stationary at level while the panel unit root test shows the mixed results. The lower values of VIF indicate that all the models were not suffered from multicollinearity problem. All the models shows the significant of Breusch and Pagan LM test means the sample of countries can be pooled for all models regression.

Table 2: Low income countries

(1) (2) (3)

GDP SAV FDI

ET -0.037254 0.079222 0.036361

(0.049322) (0.125174) (0.069157)

GST -0.076487* -0.009997 0.029167

(0.042568) (0.093690) (0.055408)

INF -0.000200 -0.000011 -0.000050

(0.000199) (0.000213) (0.000161)

IPCT -0.013428 0.132366 -0.023944

(0.060971) (0.115998) (0.073368)

ITT -0.107811 -0.140060 -0.034585

(0.066408) (0.129199) (0.081130)

TTR -0.020604*** 0.002045 -0.003727

(0.006447) (0.020935) (0.010086)

TTX -0.183144 0.066993 -0.007817

(0.122531) (0.263306) (0.155436)

Common C 0.120946*** 0.135949** 0.027807

(0.025352) (0.059446) (0.033755)

F-statistic – – –

p-values

Wald Chi Square 18.84*** 4.51 1.18

p-values (0.0087) (0.7197) (0.9913)

LR Chi Square 17.49** 4.27 1.41

p-values (0.0145) (0.7487) (0.9851)

R-squared 0.0532 0.0005 0.0071

Breusch Pagan LM Test 98.26*** 5539.56*** 1657.18***

p-values (0.0000) (0.0000) (0.0000)

Hausman Test 4.41 4.27 1.44

p-values (0.6210) (0.6403) (0.9631)

No. of observations 1500 1500 1500

No. of countries 30 30 30

Notes: Values in parentheses are standard errors. ***, **, and * indicates significant at 0.01, 0.05, and 0.1 level.

Table 2 indicates the coefficient of parameter for low income countries. Column (1), (2), and (3) on the top of table represents the results for the three hypotheses. Based on Table, only GST and TTR are significant and negative correlation with growth. Increase 1 point in TTX will reduce growth 0.18 point. GST will reduce 0.08 point while TTR reduce 0.02 point on growth. All the components of tax revenue show negative correlation with growth in low income countries.

Table 3: Lower middle income countries

(1) (2) (3)

GDP SAV FDI

ET -0.167365*** -0.398656** 0.023932

(0.060384) (0.171424) (0.067659)

GST -0.045822 -0.203413** 0.034781

(0.040883) (0.085168) (0.037303)

INF -0.001041*** 0.000509 -0.000422**

(0.000378) (0.000443) (0.000191)

IPCT 0.012664 -0.079806 -0.008630

(0.027308) (0.059308) (0.025533)

ITT 0.042575 -0.25354*** -0.025733

(0.032879) (0.084689) (0.036087)

TTR 0.019828 0.036881 -0.000293

(0.020670) (0.058276) (0.024907)

TTX -0.001933 0.141735 0.385848***

(0.072650) (0.129275) (0.057190)

Common C 0.038715 0.291930*** -0.032822

(0.025227) (0.057945) (0.025007)

F-statistic – – –

p-values

Wald Chi Square 15.95** 27.09*** 56.26***

p-values (0.0256) (0.0003) (0.0000)

LR Chi Square 16.87** 22.56*** 55.92***

p-values (0.0182) (0.0020) (0.0000)

R-squared 0.0413 0.2307 0.0470

Breusch Pagan LM Test 288.01*** 5995.57*** 5215.52***

p-values (0.0000) (0.0000) (0.0000)

Hausman Test 9.37 0.00 11.31

p-values (0.2269) (1.0000) (0.1257)

No. of observations 1500 1500 1500

No. of countries 30 30 30

Notes: Values in parentheses are standard errors. ***, **, and * indicates significant at 0.01, 0.05, and 0.1 level.

For lower middle income countries in Table 3, only ET and INF are negative and significant on growth with ET will reduce 0.16 point for each increase in 1 point of ET. Three of tax revenue (ET, GST and ITT) are highly significant and negative correlation with gross saving, in which ITT shows the most harmful to SAV. For FDI, only TTX and INF are significant with inflation will generate the tax revenue to increase and reduce inflow of FDI.

Table 4 examined the coefficient for each component of tax revenue on economic indicators in upper middle income that involve developing countries. Total tax revenue to GDP ratio was significant and positive correlation with all the three economic indicators. Increase 1 point in TTX will generate 0.28 to 0.40 point on economic. Increase 1 point in ITT will discourage 0.46 point in gross saving. Its means that taxes on international trade will reduce the people to save their money. Total tax revenue to GDP ratio (TTX), ET, GST and INF shows the highly significant effects on inflow of FDI. Inflation rate as usual will discourage FDI while increase 1 point in ET and GST will generate 0.2 point in FDI.

Table 4: Upper middle income countries

(1) (2) (3)

GDP SAV FDI

ET 0.155207 0.521871*** 0.290943***

(0.159057) (0.108057) (0.081076)

GST 0.146085* 0.005607 0.236552***

(0.077449) (0.067522) (0.044235)

INF -0.003172*** 0.000125 -0.000737***

(0.000877) (0.000457) (0.000205)

IPCT 0.001737 -0.018526 0.014753

(0.065399) (0.052281) (0.026349)

ITT 0.259307 -0.461960*** 0.004031

(0.173209) (0.127476) (0.069712)

TTR 0.218429* 0.065636 0.060201

(0.109140) (0.070611) (0.049866)

TTX 0.275306* 0.350892** 0.401550***

(0.162046) (0.176655) (0.123840)

Common C -0.198967*** 0.126263** -0.171769***

(0.076626) (0.054521) (0.041810)

F-statistic 4.65*** 5.33*** 7.65***

p-values (0.0000) (0.0000) (0.0000)

Wald Chi Square – – –

p-values

LR Chi Square – – –

p-values

R-squared 0.0337 0.0220 0.0745

Breusch Pagan LM Test 748.17*** 8164.86*** 6525.48***

p-values (0.0000) (0.0000) (0.0000)

Hausman Test 16.28** 15.86** 40.29***

p-values (0.0227) (0.0264) (0.0000)

No. of observations 1500 1500 1500

No. of countries 30 30 30

Notes: Values in parentheses are robust standard errors. ***, **, and * indicates significant at 0.01, 0.05, and 0.1 level.

Based on the World Bank Data (2010), high income countries have the lowest inflation rate compared with the other groups. Table 5 represents the coefficients for each components of tax revenue on economic indicators in high income countries. TTX has positive and significant effect to GDP and gross saving. The opposite result compared with low income countries. This result is strongly suggests that the different tax policy due to the different inflation rate in the different levels of income in a country will give the different effect of taxes on growth. The most harmful on SAV is ITT in which increase 1 point in ITT will reduce 0.68 point on SAV. However, none of tax revenue’s components are significant with inflow of FDI in high income countries.

Table 5: High income countries

(1) (2) (3)

GDP SAV FDI

ET N/A N/A N/A

GST 0.000985 -0.264359** 0.029092

(0.072409) (0.123414) (0.052473)

INF -0.006773** 0.000619 -0.002089

(0.002819) (0.001092) (0.001820)

IPCT -0.167629** -0.164691* -0.017356

(0.072801) (0.097129) (0.027655)

ITT 0.106079 -0.687393*** -0.139963

(0.117691) (0.248920) (0.106403)

TTR -0.022672 -0.038467 -0.020008

(0.016539) (0.025076) (0.026488)

TTX 0.307797* 0.969083*** 0.103271

(0.158782) (0.301466) (0.086166)

Common C 0.071033*** 0.258489*** 0.027746**

(0.010115) (0.024032) (0.013785)

F-statistic 3.54*** 3.92*** –

p-values (0.0017) (0.0007)

Wald Chi Square – – 8.94

p-values (0.1773)

LR Chi Square – – 8.86

p-values (0.1814)

R-squared 0.0154 0.0131 0.0087

Breusch Pagan LM Test 189.46*** 5366.12*** 8900.88***

p-values (0.0000) (0.0000) (0.0000)

Hausman Test 17.76*** 15.75** 3.74

p-values (0.0069) (0.0152) (0.7122)

No. of observations 1500 1500 1500

No. of countries 30 30 30

Notes: Values in parentheses are robust standard errors (exclude model for hypothesis 3). ***, **, and * indicates significant at 0.01, 0.05, and 0.1 level. N/A indicates the data is not available (omitted) in World Bank Data (2010).

CONCLUSIONS

From the regression analyses, the main findings may be summarized as follows:

Taxes on income, profit and capital gain (IPCT) was negatively affecting both low and high income countries, in which were the most harmful for growth in high income countries. This result supports Widmalm (2001), Lee and Gordon (2005) also Arisoy and Unlukaplan (2010). Total tax rate (TTR) also have negative correlation on growth in low and high income countries. This result support Engen and Skinner (1996)

Increase in total tax revenue (TTX) will encourage gross saving to increase in a country. However, taxes on international trade (ITT) have negative impact to gross saving in a country.

The most important result is total tax revenue to GDP ratio (TTX) has negative effect on growth in low income while positive effect in high income countries. This result support Marsden (1983). Besides, the potential variable such as inflation rate will affect the tax reformed and tax policy that give different impacts to the economic growth and other economic indicators. This result support Ashworth and Heyndels (2002).

Appendix: Countries in the sample.

Low income

Lower middle income

Upper middle income

High income

Afghanistan

Bangladesh

Benin

Cambodia

Comoros

Congo Dem Rep

Eritrea

Ethiopia

Ghana

Guinea

Haiti

Kenya

Liberia

Malawi

Mali

Mauritania

Mozambique

Myanmar

Nepal

Niger

Rwanda

Senegal

Somalia

Tajikistan

Tanzania

Togo

Uganda

Uzbekistan

Vietnam

Zimbabwe

Angola

Belize

Bolivia

Cameroon

China

Cote d’Ivoire

Ecuador

Georgia

Guyana

Honduras

India

Indonesia

Iraq

Jordan

Kosovo

Maldives

Moldova

Mongolia

Paraguay

Philippines

Samoa

Sri Lanka

Sudan

Swaziland

Thailand

Timor-Leste

Tonga

Tunisia

Ukraine

Vanuatu

Algeria

Argentina

Belarus

Bosnia and Herzegovina

Brazil

Bulgaria

Chile

Colombia

Costa Rica

Cuba

Dominica

Fiji

Gabon

Jamaica

Kazakhstan

Latvia

Malaysia

Mexico

Namibia

Palau

Panama

Peru

Poland

Romania

Serbia

South Africa

Suriname

Turkey

Uruguay

Venezuela RB

Andorra

Australia

Bahrain

Belgium

Brunei

Canada

Croatia

Czech Republic

Denmark

Finland

France

Germany

Greece

Hong Kong China

Italy

Japan

Korea Rep

Kuwait

Netherlands

Norway

Oman

Portugal

Singapore

Spain

Sweden

Switzerland

Trinidad and Tobago

United Arab Emirates

United Kingdom

United State