European Journal of Economic Studies, 2017, 6(2)
UROPEAN of Economic
Has been issued since 2012.
ISSN 2304-9669. E-ISSN 2305-6282
Issued 2 times a year
Dr. Vidishcheva Evgeniya – Sochi State University, Sochi, Russian Federation (Editor-in-Chief)
Dr. Simonyan Garnik – Scientific Research Centre of the Russian Academy of Sciences, Sochi, Russian Federation
Dr. Levchenko Tatyana – Sochi State University, Sochi, Russian Federation
Dr. Tarakanov Vasilii – Volgograd State University, Volgograd, Russian Federation
Dr. Balatsky Evgeny – Central Economics and Mathematics Institute (RAS), Moscow, Russian Federation
Dr. Dinh Tran Ngoc Huy – Banking University HCMC Viet Nam – GSIM, International University of Japan, Japan
Dr. Gerasimenko Viktor – Odessa State Economic University, Odessa, Ukraine
Dr. Gvarliani Tatjana - – Sochi State University, Sochi, Russian Federation
Dr. Gunare Marina – Baltic International Academy, Riga, Latvia
Dr. Kryshtanovskaya Olga – Institute of Sociology of the Russian Academy of Sciences, Moscow, Russian Federation
Dr. Minakir Pavel – Economic Research Institute of the Far Eastern Branch Russian Academy of Sciences, Khabarovsk, Russian Federation
Dr. Papava Vladimir – Ivane Javakhishvili Tbilisi State University, Tbilisi, Georgia
Dr. Prokopenko Olga – Sumy State University, Sumy, Ukraine
Dr. Vishnevsky Valentine – Institute of Industrial Economics of the National Academy of Sciences of Ukraine, Donetsk, Ukraine
Journal is indexed by: CrossRef (UK), EBSCOhost Electronic Journals Service
(USA), Electronic scientific library (Russia), 1 (Australia), Open Academic Journals Index (Russia), ResearchBib (Japan), ULRICH’s WEB (USA).
All manuscripts are peer reviewed by experts in the respective field. Authors of the manuscripts bear responsibility for their content, credibility and reliability.
Editorial board doesn’t expect the manuscripts’ authors to always agree with its opinion.
Postal Address: 1367/4, Stara Vajnorska str., Bratislava, Slovakia, Nove Mesto, 831 04
Website: http://ejournal2.com/ E-mail: firstname.lastname@example.org
Founder and Editor: Academic Publishing House Researcher s.r.o.
Release date 01.09.17. Format 21 29,7/4.
Order № 112
l of Econom
European Journal of Economic Studies, 2017, 6(2)
UROPEAN of Economic
Издается с 2012 г.
ISSN 2304-9669. E-ISSN 2305-6282 2017, 6(2). Выходит 2 раза в год
Видищева Евгения – Сочинский государственный университет, Сочи, Российская Федерация (Гл. редактор)
Левченко Татьяна – Сочинский государственный университет, Сочи, Российская Федерация
Симонян Гарник – Сочинский научно-исследовательский центр Российской академии наук, Сочи, Российская Федерация
Тараканов Василий – Волгоградский государственный университет, Волгоград, Российская Федерация
Балацкий Евгений – Центральный экономико-математический институт РАН, Москва, Российская Федерация
Вишневский Валентин – Институт экономики промышленности Национальной академии наук Украины, Донецк, Украина
Гварлиани Татьяна – Сочинский государственный университет, Сочи, Российская Федерация
Герасименко Виктор – Одесский государственный экономический университет, Одесса, Украина
Гунаре Марина– Балтийская международная академия, Рига, Латвия
Динь Чан Нгок Хай – Банковский университет Хошимин Вьетнам - GSIM, Международный университет Японии, Япония
Минакир Павел – Институт экономических исследований ДВО РАН, Хабаровск, Российская Федерация
Крыштановская Ольга – Институт социологии РАН, Москва, Российская Федерация
Папава Владимир – Тбилисский государственный университет имени Иване Джавахишвили, Тбилиси, Грузия
Прокопенко Ольга – Сумский государственный университет, Сумы, Украина
Журнал индексируется в: CrossRef (Великобритания), EBSCOhost Electronic Journals
Service (США), Global Impact Factor (Австралия),Научная электронная библиотека
(Россия), Open Academic Journals Index (Россия), ResearchBib (Япония), ULRICH’s WEB (США).
Статьи, поступившие в редакцию, рецензируются. За достоверность сведений, изложенных в статьях, ответственность несут авторы публикаций.
Мнение редакции может не совпадать с мнением авторов материалов.
Адрес редакции: 831 04, Словакия,
г. Братислава, Нове Место, ул. Стара Вайнорска, 1367/4
Сайт журнала: http://ejournal2.com/ E-mail: email@example.com
Учредитель и издатель: Academic Publishing House Researcher s.r.o.
Дата выпуска 01.09.17 Формат 21 29,7/4.
Заказ № 112
© European Journal of Economic Studies, 2017
al of Econ
C O N T E N T S
Articles and Statements
Current Account Dynamics of Central European Countries
K. Halil Arıç, M. Tuncay, S. Kun Sek ... 78
Does Government Size Affect Economic Growth in Developing Countries? Evidence from Non-stationary Panel Data
M. Cetin ... 85
The Role of Inflation and its Targeting for Low-Income Countries (Lessons from Post-Communist Georgia)
V. Charaia, V. Papava ... 96
Study on Client-Satisfaction Factors in Construction Industry
M. Duljevic, M. Poturak ... 104
Enablers of Successful Knowledge Sharing Behavior: KMS, Environment and Motivation
A. Özlen ... 115
The Relationship between Short-Run Interest Rate and its Economic Determinants: Consumer Price Index, Industrial Production Index, Household Consumption and Exchange Rate. An Empirical Research for the Four Most Developed Countries
opyright © 2017 by Academic Publishing House Researcher s.r.o.
Published in Slovak Republic
European Journal of Economic Studies
Has been issued since 2012. ISSN: 2304-9669
2017, 6(2): 78-84
Articles and Statements
Current Account Dynamics of Central European Countries
Kıvanç Halil Arıç a , *, Merve Tuncay a , Siok Kun Sek b
a Cumhuriyet University, Sivas, Turkey
b Universiti Sains Malaysia, Minden, Penang, Malaysia
"Current account" has been considered as an important variable in forecasting an economic crisis. Therefore; specifying the determinants of current account is a substantial topic for policy makers. The aim of this study is to examine the current account dynamics in the scope of Central European countries (Poland, Hungary, Czech Republic, Slovakia Republic and Croatia) during the 1997 – 2015. Panel data analysis was used in the methodology. According to the analysis results, growth of gross domestic product has no significant effect on current account. Real exchange rate, foreign direct investment and importation affect the current account negatively. However, exportation and government expenditure have positive effects on current account in Central European countries.
Keywords: current account, European Countries, balanced panel data.
Current account represents the situation of macroeconomic policies and the behavior of economic agents. It is considered as an important variable in the international macroeconomics perspective. Current account cannot be considered as a target variable such as unemployment rate and inflation also it cannot be taken as a policy variable such as interest rate and money supply. Mainly, current account has effects on the decision of lenders and borrowers in the global economy. If the current account deficit keeps up in a continuous path, it identifies the inadequate creditworthiness of a country in the global economy context. In this circumstance, the country may face the risk of bankruptcy (Hassan et al., 2015: 190).
In the 1990s financial liberalization expanded in the global level and monetary integration process of Europe was started in 2001 as a currency of euro. After the implementation of the euro in the European countries, it has contributed to the credit expansion and a decrease in private savings, and these factors led to high current account deficits for some European countries (Brissimis et al., 2010). Monetary integration brings different implementations on fiscal policy and units labor costs in Europe and, therefore; current account positions diverge between European
* Corresponding author
E-mail addresses: firstname.lastname@example.org (H.A. Arıç),
countries. For instance Germany and some smaller northern European countries' economy policies generate current account surpluses, however western periphery, eastern and most of the southern countries of Europe exhibit current account deficits (Belke, Schnabl, 2013). Before the 2009 global economic crisis, there had been excessive current account deficits in the periphery of Europe. During the pre-crisis period, these deficit problems led to economic shrinking, depredation sovereign creditworthiness, and problems in banking systems in the periphery regions. Also the decline in aggregated demand and losses on foreign asset holdings in the periphery had negative effects on current account surplus in European countries. In this respect, the control of current account imbalances has been seen as a prior policy for European policymakers (Lane, Pels, 2012).
The aim of this study is to examine the current account dynamics for Central European Countries (Poland, Hungary, Czech Republic, Slovakia Republic and Croatia) in the period of 1997-2015. Panel data was used in the analysis process. There are several studies which examine the current account dynamics in the European countries. These studies selected numerous countries in the analysis process. However; this study focuses on a limited number of countries for the specific results for the Central Europe region.
There are five main sections in this study. First section includes literature review, second section represents data and methodology, third section elaborates on the analysis process, fourth section infer analysis results and finally, fifth section is the conclusion of the study.
2. Literature Review
There are numerous studies in the literature in the respect of current account dynamics. Some of these studies examine this subject for OECD countries (Gosse, Serranito, 2014; Cavdar and Aydin, 2015; Bertola and Prete, 2015; Karras, 2016) and some of them analyze different country groups (Erauskin, 2015; Kim, 2015; Martin, 2016; Tan et al. 2015; Moral-Benito and Roehn, 2016). In order to determine the scope of this study, we particularly examine current account dynamics literature for the European countries.
Aristovnik (2006) investigated the current account dynamics for the Eastern Europe and the former Soviet Union countries in the period of 1992-2003 by using dynamic panel data analysis. He concludes that economic growth has a negative effect on current account balance. This result reflects that the economic growth is relevant with increasing of domestic investment instead of domestic saving. Public budget shocks move together with current account breakdown and this circumstance indicates the twin deficit conditions in the region. Rise in value of the real exchange rate and deterioration of the terms of trade also affects the affect the current account balance negatively.
Gehringer (2015) examines current account dynamics for all European countries, except Luxemburg, during the 1995 and 2010 period by using panel data method. He concludes that excessive private and public consumption cause current account deficits. Additionally, credit variable, growth of GDP per capita, real exchange rate and construction sector variables have negative effects on current account balance in European Union economies.
Bollano and Ibrahimaj's (2015) study on the current account dynamics of Central and Eastern European countries in the period of 2015:1 to 2014:4 by using panel data methodology. They find out that GDP growth and fiscal deficit have a negative effect on current account. However, depreciation of the real effective exchange rate affects current account positively.
Zorzi et al. (2009) make a comprehensive survey about current account benchmarks for Central and Eastern European countries. They use external sustainability approach â la Lane and Milesi-Ferretti (LM) and structural current accounts literature (SCA) which is based on panel data methodology. According to LM approach they provide the importance of sensitivity of outcome to the external indebtedness and the consideration to exclude the foreign direct investment subcomponent from the net foreign assets aggregate. In respect to SCA approach they analyze the sensitivity of outcome to various combinations of fundamentals.
Brissimis et al. (2010) examined the determinants of current account for Greece during the 1960 to 2007 by using co-integration analysis in the long and the short run. They conclude that the current account balance could be established when the ratio of private sector financing to GDP counts as an indicator for financial liberalization in the model.
Accordingly, large current account balances prior to the crisis is the best predictor of a sharp drop in output during the crisis. They suggest supportive macro policies to moderate the adjustment process and to keep overall euro inflation at or above target level are necessary.
Kollmann et al. (2015) investigate the determinants of German’s current account surplus and its effects on Euro Area for the period of 1995 and 2013 in which they find factors like positive shocks to German saving rate, world’s demand for German exports, German labor market reforms and other positive German aggregated supply shocks have effect on German current account surplus and negatively affect Euro Area net exports. The research also discusses that exchange rate regime may have a first order effect on current account dynamics.
3. Data and Methodology
We used current account balance as a percentage of GDP (CA) for dependent variable. Independent variables are GDP growth rate (GDP), real effective exchange rate index (RER), exports of goods and services as a percentage of GDP (EXP), imports of goods and services as a percentage of GDP (IMP), foreign direct investment as a percentage of GDP (FDI) and general government final consumption expenditure growth (GOV) respectively. Data was collected from “World Development Indicators” in the World Bank website. This data was collected in the respect of five Central European Countries (Poland, Hungary, Czech Republic, Slovakia Republic and Croatia) in the period of 1997-2015.
According to theoretical perspective, in the emerging markets, growth of the economy leads to an increasing expectation of incomes and, correspondingly, an increasing on workers’ consumption. Therefore, it can be expected that GDP growth has a negative effect on current account (Zorzi et al., 2009; Bollano and Ibrahimaj, 2015; Gehringer, 2015). Real exchange rate adjustment is the most effective indicator on current account adjustment than other adjustment instruments such as income, output and expenditure. Relative price movements lead to matching expenditure between domestic goods and foreign goods (Gervais et al., 2016). If an appreciation occurs for the real exchange rate, it leads to an increase in the purchasing power of household with respect to imported goods, as well as an increase the in the value of the property assets of domestic agents. Therefore, all these variables lead to increase on consumption and a decrease on the saving tendency. Hence it is expected that the increasing of real exchange rate has a negative effect on current account (Brissimis et al., 2010). Similar results obtained for European countries (Aristovnik, 2006; Gehringer, 2015; Bollano and Ibrahimaj, 2015). Exports indicate demand for a local product and imports reflect supplies from foreign countries to meet local production requirements. Shortly export can be regarded as a credit to local economy whereas import implies a debit for a local economy. From this point it could be expected that export has a positive and import has a negative effect on current account. In the literature, foreign direct investments have a positive spill-over effects on host countries' current account by means of bringing technology and know-how, contributing to development of companies, integration into the global economy and increasing competition (Mencinger, 2008). Generally, it is suggested that government budget deficits leads to current account deficits via redistributing income from future generations to present generations. In this respect of twin deficit hypothesis, government expenditure could be seen as an important factor for budget deficits (Zorzi et al., 2009). Therefore model has been established in the respect of literature as an equation (1).
In this study we used balanced panel data set in the panel data analysis process. Balanced panel data implies that the all year’s data has been obtained for each country and there has not any deficient data. Panel data set in includes of 5 horizontal section units. i symbolizes country and t symbolizes time; i=1-5 (5 countries) and t=1997-2015 (19 years). The total number of observations in data set (i×t = 95) is 95.
4. Analysis Process
examined whether standard error of unit effects is equal to zero (H0: σµ=0). Otherwise, LR test is
used to examine whether standard error of time effects is equal to zero (H0: σλ=0) (Yerdelen
Tatoğlu, 2012: 170). Pooled OLS model can be used, if unit and time effects are not determined in LR test. In spite of this condition, if unit and/or time effects are determined in test results, it can be said that the model is one sided or two sided.
Table 1. LR Test
Unit Effect Time Effect
χ2 47.26 0.62
prob. 0.0000 0.2151
The results of LR test exhibit that there is an only unit effect in the model. Consequently, the model is one sided. Hausman specification test is used to specify whether unit effect is fixed or random.
Hausman test infers that if there is no correlation between error components (ui) and
explanatory variables (xkit), both fixed effects and random effects estimators are appropriate. In any
case, if there is a correlation between error components and explanatory variables, random effects estimator is inappropriate. In Hausman test, null hypothesis implies that there is no correlation between error components and explanatory variables (Hill et al., 2011: 559). It can be said that random effects are appropriate when there is not a correlation between ui and xkit, and fixed effects
are appropriate when there is a correlation between ui and xkit (Gujarati, 2003: 650).
Table 2. Hausman Test
prob. 46.25 0.0000
Hausman test results show that unit effects are fixed. Therefore, analysis is made in accordance with one sided fixed effects model.
After these findings, model was examined in the scope of variation from basic assumptions. One of these assumptions is constant variance (homoscedasticity) assumption. Constant variance assumption implies that while unit values of explanatory variables change, variance of error term remains fixed. If this assumption does not valid, model includes heteroscedasticity (Wooldridge, 2012: 93). Modified Wald Test was used to examine this assumption.
Table 3. Test for Heteroscedasticity
Modified Wald Test
Table 4. Test for Autocorrelation
Modified Bhargava et al. Durbin-Watson Test Baltagi-Wu LBI Test
Another assumption is about correlation between units. In studies such as domestic and regional economies, neighborhood effects can show spill-over in themselves. In such cases, correlations have spatial view rather than temporal view (Greene, 2012: 389). This assumption is tested through Friedman’s Test. According to the Friedman’s test of cross sectional independence test statistics and probability values, there is a correlation between units.
Table 5. Test for Correlation between Units
Friedman’s Test of Cross Sectional Independence
According to the results of analysis, there have been autocorrelation and correlation between units problems in the model. In order to solve these problems, standard errors which are resistant to deviations from assumptions were produced by using method of Driscoll-Kraay estimator.
Table 6. Analysis Results
Explanatory Variables Coef. t-stat. p-value
RER -0.0670 -0.0631 -4.53 -1.27 0.0110.274 **
EXP 0.9172 14.74 0.000*
IMP -0.9372 -11.20 0.000*
FDI -0.0509 -4.46 0.011**
GOV 0.0862 2.41 0.073***
Cons. 4.3684 8.44 0.001*
R2: 0.8658 Prob. 0.0000
Note: (*) significant at %1 level, (**) significant at %5 level, (***) significant at %10 level.
Analysis results show that the GDP variable effects on CA negatively but it is statistically insignificant. RER effects on CA negatively and this result is statistically significant. In this regard one unit appreciation in RER leads to 0.06 % decrease in CA. Coefficient of EXP has a positive and statistically significant impact on CA. It can be described that one unit increase in EXP gives rise to 0.91 % increase in CA. IMP variable has a negative and statistically significant effect on CA. One unit increase in IMP cause 0.93 % decrease in CA. Coefficient of FDI has a negative and statistically significant effect on CA. It implies that one unit increase in FDI leads to 0.05 % decrease in CA. GOV variable has a positive and statistically significant effect on CA. One unit increase GOV gives rise to 0.08 % increase in CA. Effect of RER, EXP and IMP variables on CA are coherent with the theoretical expectations. However FDI and GOV variables effect on CA are not consistent with the theoretical expectations. Also it can be seen that there have been strong effect of EXP and IMP variables on CA.
the control of current account imbalances has been considered as a primary goal for European countries. In this study we analyze current account dynamics for five Central European countries in the period of 1997-2015 by using panel data methodology.
Real exchange rate has a negative and statistically significant effect on current account in Central European countries. This result is coherent with the theoretical expectations, and it can be said that appropriation in the real exchange rate brings to increase purchasing power of household and rising demand for imported goods. Furthermore, increasing value of the property assets of domestic agents also affect current account negatively. Exportation has a positive effect on current account and this result implies that increasing of foreign demand for local products gives rise to foreign currency access to the economy. Conversely the demands for foreign goods have an impact on current account negatively in Central European countries. This is because that importation stated the debt conditions for countries. The results of exportation and importation are also convenient with theoretical perspective. However foreign direct investment has a negative effect on current account in Central European countries and this result contradicts with the theoretical expectations. But it can be said that in the long run the FDI’s positive effects on current account could be turned negative by the way of repatriation of profits to investor country and this negative effect could be extended if the investment funds gain from the host country through credits channel (Moura, Forte, 2010). Government expenditure has positive effects on current account in Central European countries. It can be said that this result is also adverse with theoretical expectations. Theoretically, twin deficit hypothesis implies that if the government expenditure financed by the government incomes, it leads a current account deficit in the economy. However, Finn (1998) asserts that government expenditure on final goods has a positive effect on private sector’s investment and domestic output. In this respect, government expenditure could be financed without government income and, government expenditure could be financed by the increasing private investments. Therefore it can be said that government expenditure impacts on current account positively in Central European countries.
Current account balance is sensitive to international trade movements as respect to import and export. Improving of the policies to increase export and decrease import are important agenda for Central European countries. It can be used regulations on foreign investors to limiting the repatriation of profits to host country. The impact of government expenditure on private investments is positive. Therefore government expenditure does not generate current account deficit in Central European countries.
Hassan et al., 2015 – Hassan, K., Hoque, A., Rao, A. (2015). Sustainability of Current Account Balance in Asean Countries: Evidence from a Panel Error Correction Model. The Journal of Developing Areas, vol. 49, no. 6, pp. 189-204.
Brissimis et al., 2010 – Brissimis, S. N., Hondroyiannis, G., Papazoglou, C., Tsaveas, N.T., Vasardani, M.A. (2010). Current Account Determinants and External Sustainability in Periods of Structural Change. European Central Bank Working Paper SeriesNo: 1243, Germany: European Central Bank.
Belke and Schnabl, 2013 – Belke, A., Schnabl, G. (2013). Four Generations of Global Imbalances. Review of International Economics, vol. 21, no. 1, pp. 1-5.
Lane and Pels, 2012 – Lane, P.R., Pels, B. (2012). Current Account Imbalances in Europe,
IIIS Discussion Paper No: 397. [Electronic resourse]. URL: https://www.tcd.ie/iiis/ documents/discussion/pdfs/iiisdp397.pdf [Accessed: 30.01.2017].
Gosse and Serranito, 2014 – Gosse, J., Serranito, F. (2014). Long-run Determinants of Current Accounts in OECD Countries: Lessons for Intra-European Imbalances, Economic Modelling, vol. 38, pp. 451-462.
Cavdar, Aydin, 2015 – Cavdar, S.C., Aydin, A.D. (2015). Understanding the Factors Behind Current Account Deficit Problem: A Panel Logit Approach on 16 OECD Member Countries,
Procedia Economics and Finance, vol: 30, pp.187-194.
Karras 2016 – Karras, G. (2016). Macroeconomic Volatility and the Current Account: Evidence from a Panel of OECD Countries, International Economic Journal, vol. 30, no. 3, pp. 322–338.
Erauskin, 2015 – Erauskin, I. (2015). Savings, the Size of the Net Foreign Asset Position and the Dynamics of Current Accounts. International Review of Economics and Finance, vol. 39, pp. 353-370.
Kim, 2015 – Kim, S. (2015). Country Characteristics and the Effects of Government Consumption Shocks on the Current Account and Real Exchange Rate. Journal of International Economics, vol. 97, pp. 436-447.
Martin, 2016 – Martin, F.E. (2016). Exchange Rate Regimes and Current Account
Adjustment: An Empirical Investigation. Journal of International Money and Finance, vol. 65, pp. 69–93.
Tan et al. 2015 – Tan, Z., Yao, Y., Wei, S. (2015). Financial Structure, Corporate Savings and Current Account Imbalances, Journal of International Money and Finance, vol. 54, pp. 142-167.
Moral-Benito, Roehn, 2016 – Moral-Benito, E., Roehn, O. (2016). The Impact of Financial Regulation on Current Accunt Balances. European Economic Review, vol. 81, pp. 148-166.
Aristovnik, 2006 – Aristovnik, A. (2006). The Determinants and Excessiveness of Current Account Deficits in Eastern Europe and the Former Soviet Union. William Davidson Institute Working Paper No:827, USA: University of Michigan.
Gehringer, 2015 – Gehringer, A. (2015). New Evidence on the Determinants of Current Accounts in the EU, Empirica, vol. 42, pp.769-793.
Bollano, Ibrahimaj, 2015 – Bollano, J., Ibrahimaj, D. (2015). Current Account Determinants in Central Eastern European Countries, Graduate Institute of International and Development Studies International Economics Department Working Paper Series, No: HEIDWP0022-2015, Switzerland.
Zorzi et al., 2009 – Zorzi, M.C., Chudik, A., Dieppe, A. (2009). Current Account Benchmarks for Central and Eastern Europe a Desperate Search? European Central Bank Working Paper Series No: 995, https://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp995.pdf? 9da81a457e30259e36a6df428edf2a51 [Accessed: 30.01.2017].
Kang, Shambaugh, 2016 – Kang, J.S., Shambaugh, J.C. (2016). Current Account Deficits,
Economic Policy, vol. 31, no. 85, pp. 153-199.
Kollmann et al., 2015 – Kollmann, R., Ratto, M., Roeger, W., Veld, J., Vogel, L. (2015). What Drives the German Current Account? And How Does It Affect Other EU Member States? Economic Policy, vol. 30, no. 81, pp. 47-93.
World Bank – World Bank (2016). World Development Indicators. [Electronic resourse]. URL: http://data.worldbank.org/data-catalog/world-development-indicators [Accessed: 15.11.2016].
Gervais et al., 2016 – Gervais, O., Schembri, L., Suchanek, L. (2016). Current Account Dynamics, Real Exchange Rate Adjustment, and the Exchange Rate Regime in Emerging-Market Economies. Journal of Development Economics, vol. 119, pp. 86–99.
Mencinger, 2008 – Mencinger, J. (2008). The ‘Addiction’ with FDI and Current Account Balance, International Center for Economic Research Working Paper Series No: 16/2008, [Electronic resourse]. URL: http://www.biblioecon.unito.it/biblioservizi/RePEc/icr/ wp2008/ICERwp16-08.pdf [Accessed: 30.01.2017].
Yerdelen Tatoğlu, 2012 – Yerdelen Tatoğlu, F. (2012). Panel Veri Ekonometrisi, Stata Uygulamalı, Beta Yayınları, İstanbul.
Hill et al., 2011 – Hill, R.C., Griffiths E.W., Guay. C.L. (2011). Principles of Econometrics, Fourth Edition, Wiley, USA.
Gujarati, 2003 – Gujarati, D.N. (2003). Basic Econometrics, Fourth Edition McGraw-Hill, New York.
Wooldridge, 2012 – Wooldridge, J. (2012). Introductory Econometrics: A Modern Approach, Fifth Edition, South-Western, Cengage Learning, USA.
Greene, 2012 – Greene, W.H. (2012). Econometric Analysis, Seventh Edition, Prentice Hall, USA.
Moura, Forte, 2010 – Moura, R., Forte, R. (2010). The Effects of Foreign Direct Investment on the Host Country Economic Growth-Theory and Empirical Evidence, FEP Working Papers No.390, pp. 1-32.
opyright © 2017 by Academic Publishing House Researcher s.r.o.
Published in Slovak Republic
European Journal of Economic Studies
Has been issued since 2012. ISSN: 2304-9669
2017, 6(2): 85-95
Does Government Size Affect Economic Growth in Developing Countries? Evidence from Non-stationary Panel Data
Murat Cetin a ,*
a Namik Kemal University, Turkey
The Armey curve suggests that there is an inverted U relationship between government size and economic growth. In order to investigate this relationship for 12 developing countries from 1990 to 2012, this study uses panel data methodology including panel unit root, cointegration and causality tests. The results show that i) the series are integrated at order of I(1), ii) there exists a long run equilibrium relationship between the variables, iii) economic growth is positively correlated with the government consumption expenditure, iv) economic growth is negatively correlated with the squares of government consumption expenditure, v) there exists a causality running from the explanotary variables to economic growth in the long run and short run. The study provides an evidence that there exists an inverted U relationship between government consumption expenditure and economic growth implying the validity of Armey curve in these countries. The study may also provide some policy implications.
Keywords: government size, economic growth, Armey curve, panel data analysis.
Economic growth and its determinants has been one of the main topics investigated by theorists and politicians. According to growth literature, there are two fundamental kinds of growth theory. The first is the neoclassical growth theory. It is well known as the exogenous growth model presented by Solow (1956), Swan (1956), and Koopmans (1965). The second is the new growth theory developed by Romer (1986; 1990), Lucas (1988), Barro (1990), Rebelo (1991), Grossman and Helpman (1991), Aghion and Howitt (1992), and Jones (1996). This theory is also known as the endogenous growth model.
The neoclassical theory of growth generally focuse on capital accumulation and its relation to savings and population growth. It suggests that in the long run economy will reach a steady state where per capita output is constant. It also suggests that there is a linear relationship between a number of variables and economic growth in the long-run. According to this theory, government policy cannot influence the steady-state growth rates. As a result, the impact of government policy on the long run growth has not been investigated in this model.
The new growth theory suggests that both transition and steady state growth rates are endogenous and there are several determinants of long run growth. Here, long run growth rates can differ across countries and convergence in income per capita cannot occur. However, according
to this theory government policy can affect economic growth either directly or indirectly. In this model there are three basic fiscal instrument affecting the long run growth rates: expenditure, taxation and the aggregate budgetary balance. Firstly, these instruments affect the efficiency of resource use and the rate of factor accumulation. These developments influences a country’s long-run growth performance (Barro, 1989; 1990; Brons et al. 1999).
A part of the new growth theory focuses on the relationship between government size and economic growth. The literature on public expenditure and economic growth stresses on the presence of a historical relationship between government size and GDP growth. This is called as the Armey curve (Armey, 1995), Rahn curve (Rahn and Fox, 1996) or BARS curve (Barro, 1989; Armey, 1995; Rahn and Fox, 1996; Scully, 1994). This literature uses the form of an inverted U-shaped curve. The Armey curve is based on the law of diminishing factor returns and implies the idea that there is a positive correlation between public expenditure and GDP up to a certain point. After that the correlation becomes negative. In other words, after this point an increase in public expenditure leads to a decrease in GDP. So, Armey curve exhibits a relationship similar to that of Kuznets’ curve. According to Armey curve, the government size and economic growth may be modelled by using a quadratic function (Vedder and Gallaway, 1998).
Barro (1990) investigates the impact of different sizes of government on economic growth. According to Barro, an increase in taxes decreases economic growth, while an increase in government expenditure raises marginal productivity of capital. So, economic growth increases. If the government is small, the second force dominates. If the government is large, the first force dominates. The study’s main finding reveals that the relation between government expenditure and economic growth is non-monotonic.
The Armey curve can be formulized in different shapes in order to test whether an “inverted U” relationship exists between public expenditure and economic growth. The empirical research on this topic aims to test the presence of this relationship in diﬀerent countries by using several econometric techniques. Examples are given by Miller and Russek (1997), Vedder and Gallaway (1998), Kneller et al. (1999), Folster and Henrekson (2001), Pevcin (2004), Chen and Lee (2005), Angelopoulos et al. (2008), Herath (2010), Magazzino and Forte (2010), Afonso and Furceri (2010), Wu et al. (2010), Ijeoma and O’Neal (2012), Roy (2012), and Altunc and Aydın (2013). But, the empirical literature provides inconclusive findings regarding the relationship between government expenditure and economic growth.
Miller and Russek (1997) investigate the link between government expenditure and economic growth in both developed and developing countries. The results indicate that debt-financed increases in government expenditure slow economic growth and tax-financed increases enhance economic growth for developing countries. The results also indicate that there is no relation between debt-financed increases in government expenditure and economic growth and there is negative link between tax-financed increases and economic growth for developed countries.
Vedder and Gallaway (1998) test the validity of the Armey curve in the cases of United States, Sweden, Denmark, Canada, Britain and Italy over the period 1947-1997. The results show that there is empirical evidence supporting the validity of the Armey curve for all these countries. Employing panel data for 22 OECD countries, Kneller et al. (1999) show that productive government expenditure increases economic growth, while non-productive government expenditure does not.
Folster and Henrekson (2001) investigate the impacts of expenditure and fiscal measures on economic growth for rich countries over the period 1970-1995. The study finds a strong negative relationship between public expenditure and economic growth. Using panel data regression analysis based on five-year arithmetic averages, Pevcin (2004) examines the relationship between government expenditure and economic growth for European countries. The empirical findings support the presence of the Armey curve over the period.
Angelopoulos et al. (2008) analyze the relation between government spending and economic growth in developed and developing countries. Using a panel OLS and 2SLS, they find evidence that there is a nonlinear link between government expenditure and economic growth. The results show that an efﬁcient public sector has a positive impact on economic growth.
Herath (2010) investigates the relationship between government expenditure and economic growth in the case of Sri Lanka by using second degree polynomial regressions. The findings show that there is a positive relation between the variables. The findings also support the Armey’s idea of aquadratic curve for Sri Lanka.
Magazzino and Forte (2010) investigate the existence of Armey curve for the EU countries in the period 1970-2009 by using time-series and panel data techniques. The study provides empirical evidences generally supporting the presence of Armey curve.
Afonso and Furceri (2010) analyze the impacts of size and volatility of government revenue and spending on economic growth in OECD and EU countries by applying panel regression analyses. The findings suggest that both variables are harmful to economic growth. In particular, the results show that government consumption and investments have a negative effect on economic growth.
Wu et al. (2010) examine the causal relation between government spending and economic growth by using the panel Granger causality method presented by Hurlin (2004) and panel data set from 1950 to 2004. The study finds evidence of a positive relation between government spending and economic growth. The sudy also finds bi-directional causality between the variables for the different sub samples of countries.
Ijeoma and O’Neal (2012) examine the impact of government expenditure on economic growth for Nigerian economy from 1980 to 2011. Using ARDL bounds testing approach, the results indicate that government recurrent and capital expenditures are positively correlated with economic growth in the short-run. In the long run there is a positive relation between government recurrent expenditure and economic growth, while government capital expenditure is negatively linked to economic growth in Nigeria.
Using time-series data covering the period 1950-2007, Roy (2012) analyses the relationship between government size and economic growth in the United States. The study particularly investigates the impacts of government consumption and government investment expenditures on US economic growth. Based on the results of a simultaneous-equation model, government consumption expenditure decreases economic growth, while government investment expenditure increases economic growth in the United States. So, the study shows that the overall impact of total government spending on economic growth is uncertain.
Altunc and Aydın (2013) examine the presence of Armey curve for Turkey, Romania and Bulgaria by using ARDL bounds testing approach to cointegration from 1995 to 2011. This study finds an empirical evidence that the Armey curve is valid for Turkey, Romania and Bularia.
Following the empirical lietrature, this study’s main aim is to investigate wether the Armey curve (the inverted U relationship between government size and economic growth) exists in developing countries over the period 1990-2012. In this purpose, we employ panel unit root tests developed by Maddala and Wu (1999), Hadri (2000), and Im et al. (2003). We also employ the cointegration methods developed by Kao (1999) and Maddala and Wu (1999) to examine the long-run relationship between the variables. Long-long-run estimation is conducted by panel OLS method. Finally, the long run and short run causality between the variables is investigated by panel vector error correction model (PVECM).
The remainder of this study is organized as follows. Section 2 describes the model and data of the empirical analysis. Section 3 presents the empirical methodology. Empirical results are reported in Section 4. Section 5 concludes the study with some policy implications.
2. Model and Data
government size and economic growth (Armey curve), the following quadratic function presented by Vedder and Gallaway (1998) can be used
it it it
where GDP, GOV and GOV2 represent per capita real income, government consumption
expenditure as a percentage of real GDP and square of government consumption expenditure as a percentage of annual real GDP, respectively. So, government consumption expenditure is used as an indicator of government size. The data are transformed to natural logarithm because log-linear form provides a better result. α1 and α2 are the slope coefficients and the sign of the coefficients is
expected to be positive and negative, repectively (Vedder and Gallaway, 1998; Herath, 2010; Altunc and Aydın, 2013). εt is the error term assumed to be normally distributed with zero mean and
constant variance. Table 1 presents the descriptive statistics of the variables employed in the analysis. Figure 1 shows the plots of the series.
Table 1. Descriptive statistics
Balanced panel: N=12, T=23, Observations=276
Variable Unit Mean Median Std. Dev. Min. Max.
LNGDP GDP per capita, 2005=100, $ 8.308 8.410 0.535 6.137 9.053 LNGOV Goverment consumption
expenditure/GDP, 2005=100, $ 2.616 2.579 0.297 2.080 3.177 LNGOV2 Square of LNGOV 6.935 6.651 1.573 4.328 10.094
6 7 8 9 10 1 - 90 1 - 00 1 - 10 2 - 97 2 - 07 3 - 94 3 - 04 4 - 91 4 - 01 4 - 11 5 - 98 5 - 08 6 - 95 6 - 05 7 - 92 7 - 02 7 - 12 8 - 99 8 - 09 9 - 96 9 - 06 1 0 - 9 3 1 0 - 0 3 1 1 - 9 0 1 1 - 0 0 1 1 - 1 0 1 2 - 9 7 1 2 - 0 7 LNGDP 2.0 2.2 2.4 2.6 2.8 3.0 3.2 1 - 90 1 - 00 1 - 10 2 - 97 2 - 07 3 - 94 3 - 04 4 - 91 4 - 01 4 - 11 5 - 98 5 - 08 6 - 95 6 - 05 7 - 92 7 - 02 7 - 12 8 - 99 8 - 09 9 - 96 9 - 06 1 0 - 9 3 1 0 - 0 3 1 1 - 9 0 1 1 - 0 0 1 1 - 1 0 1 2 - 9 7 1 2 - 0 7 LNGOV 4 6 8 10 12 1 - 90 1 - 00 1 - 10 2 - 97 2 - 07 3 - 94 3 - 04 4 - 91 4 - 01 4 - 11 5 - 98 5 - 08 6 - 95 6 - 05 7 - 92 7 - 02 7 - 12 8 - 99 8 - 09 9 - 96 9 - 06 1 0 - 9 3 1 0 - 0 3 1 1 - 9 0 1 1 - 0 0 1 1 - 1 0 1 2 - 9 7 1 2 - 0 7 LNGOV2
Fig. 1. The plots of LNGDP, LNGOV and LNGOV2 series
3. Econometric Methodology
3.1 Panel Unit Root Tests
Im et al. (2003) provides a very simple panel unit root test which is well known as IPS test. They employ a separate ADF regression as follows:
∑1 , 1 , i p j it j t i ij t i i i
where i = 1, . . .,N and t = 1, . . .,T
The test allows for a heterogeneous coefﬁcient of yit-1 and bases on averaging individual unit
root test statistics. In this test, the null and alternative hypotheses are as follows:
for all i (3)
for i = 1, 2, ….. N1
i 0for i = N1+1, ….. N (5)
The IPS t-bar statistic indicates an average of the individual ADF statistics and is estimated as follows:
N i i NT
where t𝞺i is the individual t-statistic for testing H0 hypothesis. In case the lag order is always
zero, IPS provides simulated critical values related with t-bar for different number of cross-sections
N and series lenght T. IPS reveals that standardized t-bar statistic exhibits an asymptotic N(0,1) distribution.
The unit root test developed by Maddala and Wu (1999) uses the Fisher (p) test. Under cross-sectional independence of the error terms εit, the joint test statistic can be expressed as follows:
N i i
In this procedure, the null and alternative hypotheses are similar to IPS’s hypotheses. Using the ADF estimation equation in each cross-section, this test computes the ADF t-statistic for each individual series. So, the Fisher-test statistics are calculated and are compared with the appropriate
χ2 critical value.
Hadri (2000) presents a panel version of the Kwiatkowski et al. (1992) test. In this procedure, the null hypothesis implies that there exists stationarity in all units. The null hypothesis is tested against the alternative of a unit root in all units. The test is based on Langrange multiplier test and the residuals are obtained from the following regression:
it mt mi
, m = 2, 3 for i = 1, …… N. (8) The test statistic is then given by
T t ei it N i LM
H1 2 2 1 2
3.2 Panel Cointegration Tests
Kao (1999) suggests several residual-based panel tests and they have parametric properties. In these tests, the null hypothesis implies that there exsists no cointegration. In this procedure, the DF and ADF unit root tests are added to panel cointegration analyses. The main feature of these tests is that they base on the spurious least squares dummy variable panel regression equation as follows:
it it i
, i = 1,……N; t = 1,…….T (10) in which
are restricted to be atmost I(1) with uit∼ (0, 2
and εit∼ (0,
2) i.i.d.. The ADF type panel statistic developed by Kao bases on the following AR (p)
regression itp p t i p t i t i
it pe e e v
Kao (1999) formulates the ADF panel test statistic as follows:
2 0 2 2 2 0 0 1 ' 1 '
(v v v v v v sv N
i i i i
i i i i
where ' 1 '
(ip ip ip ip
Xipindicates a matrix of observations on the
NT v s N i T t itp v
1 1 2
2 ˆ (13)
ˆitpimplies the estimate of
vitp. The panel ADF test has a asymptotically N(0,1)
Hence, in addition to the Kao test, we also employ Fisher’s test to aggregate the p-values of the individual Johansen maximum likelihood cointegration test statistics. In the Fisher procedure which is a non-parametric test the homogeneity in the coefficients are not assumed (Maddala and Kim, 1998; Maddala and Wu, 1999).
3.3 Panel Granger Causality Test
it it q
k it ik
k it ik
k it ik
it LNGDP LNGOV LNGOV ECT
1 1 1
where and q represent the first difference operator and the lag length, respectively. ECT
denotes the error-correction term which contains estimated residuals from the cointegration regression (Eq. 1). μ is the serially uncorrelated error term. γ reflects the long-run equilibrium relationship among the variables. If θ2 or θ3 is not equal to zero, it is determined to be a short run
causal relationship. If γ is not equal to zero, it is determined to be a long run causal relationship. If γ and θ2 or θ3 are not equal to zero, it is determined to be a joint causal relationship.
4. Empirical Findings
Table 2 reports panel unit root test results. The findings indicate that the series are not stationary in level. After taking the first difference, the series are stationary. So, it is concluded that all variables are integrated at order of I(1). These results enable us to apply the cointegration tests.
Table 2. Panel unit root test results
Notes: The optimal lag lengths are selected automatically using Akaike information criteria (AIC). The LLC test uses Newey-West bandwidth selection with Bartlett kernel. a denotes significance at
the 1 % level. p-values are given in parentheses.
Table 3 presents the results of Johasen-Fisher and Kao cointegration tests. Fisher statistics estimated from trace and maximum eigen tests indicate that there are two cointegration vectors implying the presence of a long-run relationship between the variables at the %1 level. Kao test results indicate the existence of a long-run relationship between te variables. All the findings provide an evidence that there is a cointegration relationship between per capita real income, government consumption expenditure and square of government consumption expenditure over the period.
test statistics test statistics PP-Fisher Hadri test statistics
Panel A: Level
LNGDP 4.576 5.851 5.293 11.799a0.000
LNGOV -0.805 29.220 21.627 8.704a0.000
LNGOV2 -0.727 28.713 21.981 8.784a0.000
Panel B: First difference
122.139a0.000 133.135a0.000 0.645
128.351a0.000 145.549a0.000 0.092
Table 3. Panel cointegration test results
Notes: The optimal lag length is selected using AIC. a and b denote significance at the 1% and 5% level,
respectively. The values in parenthesis are p-values.
The estimations of long-run parameters are conducted by using panel pooled OLS method. The results are presented in Table 4. Diagnostic tests show that there are the problems of serial correlation and heteroscedasticity in the model. We apply the processes of AR(1) and White cross-section to resolve these problems. The results show that economic growth is positively correlated with the government consumption expenditure. This indicate that an increase in government size can enhance economic growth. The results also show that economic growth is negatively correlated with the square of government consumption expenditure. These findings provide an evidence supporting the presence of an inverted U shaped relationship between government size and economic growth.
Table 4. Panel regression estimation results
(Dependent variable: LNGDP, Method: Pooled panel OLS)
Variables Coefficients t-statistics Standart errors
LNGOV 1.084 3.252 a0.001 0.333
LNGOV2 -0.267 -4.114 a0.000 0.065
Constant 8.158 18.202 a0.000 0.448
AR(1) 0.968 251.622 a0.000 0.003
F-statistic 22628.85 a0.000
Durbin-Watson statistic 1.420
LMh (2) statistic 262.852 a0.000
Baltagi-Lee (2) statistic 385.831 a0.000
Notes:a denotes significance at the 1% level. The values in parentheses are p-values
Table 5 reports the results of the long-run, short-run and joint Granger causality. The results suggest that the lagged error correction term is negative and statistically significant at 5 % level as expected. This implies a causality running from government consumption expenditure and the squares of government consumption expenditure to economic growth in the long run. It is found that there exists a causal relation running from government consumption expenditure and the squares of government consumption expenditure to economic growth in the short run. It is also found that there exists a joint causal relation running from the explanatory variables to economic growth. The Granger causality findings provide an evidence that government consumption expenditure (government size) causes economic growth in developing countries over the period.
Cointegration tests Fisher statistics
(from trace test) (from max. eigen test) Fisher statistics
Panel A: Johansen-Fisher
None 255.6a0.000 195.8a0.000
At most 1 121.0a0.000 120.0a0.000
At most 2 32.880.106 32.880.106
Panel B: Kao ADF statistics
Table 5. Panel Granger causality test results (Dependent variable: LNGDP)
Series Short run
F-statistic Long run ECT(-1) Joint (Short run and Long run) F-statistic
ECTit-1 -0.020 a0.000
LNGOV/ECT 6.061 a0.000
LNGOV2/ECT 6.116 a0.000
Notes: The optimal lag length is selected using AIC. a and b denote significance at the 1 % and 5 % level,
respectively. The values in parentheses are p-values.
5. Conclusion and Policy Implication
The determinants of economic growth have been discussed by theorists and econometricians for a long time. Growth literature presents two fundemental models: exogenous growth model and endogenous growth model. The first model suggests that there is a linear relationship between a number of variables and economic growth in the long-run. In this model, government policy cannot influence the steady-state growth rates. The second model is well known as new growth theory. In this model government policy can affect economic growth either directly or indirectly. In this contex, a fundemental strand of the new growth theory concentrates on the inverted U relationship between government size and economic growth. This is generally called as Armey curve.
The study investigates the cointegration and causal relationship between the government consumption expenditure and economic growth in the context of Armey curve. We employ panel data covering 1990-2012 for 12 developing countries. Panel unit root tests indicate that the series are integrated at order of I(1) implying that we can apply the cointegration tests. Panel cointegration tests reveal that there exists a lon run relationship between the variables. Panel pooled OLS estimations suggest that the coefficients of government consumption expenditure and the squares of government consumption expenditure are positive and negative, respectively as expected. Granger causality test based on VECM shows that there exists a causal relation running from government consumption expenditure and the squares of government consumption expenditure to economic growt in the long run and short run. All the empirical findings reveal that there exists an inverted U-shaped relationship between government consumption expenditure and economic growth. So, the study provides an empirical evidence that the Armey curve is valid for developing countries over the period.
The empirical results also imply that there is an optimal level of government consumption expenditure. Therefore, governments should avoid excessive consumption expenditure. Otherwise, these excessive expenditure hamper to economic growth. On the other hand, this study can be repeated by considering different kinds of government spending. This empirical study may also bring about new empirical studies. In this respect, a further empirical research may include the individual countries or the sub groups of the panel.
Afonso, Furceri, 2010 – Afonso, A., Furceri, D. (2010). Government size, composition, volatility and economic growth. European Journal of Political Economy, 26, 517-532.
Altunc, Aydın, 2013 – Altunc, O. F., Aydın, C. (2013). The relationship between optimal size of government and economic growth: Empirical evidence from Turkey, Romania and Bulgaria.
Procedia-Social and Behavioral Sciences, 92, 66-75.
Angelopoulos, Philippopoulos, Tsionas, 2008 – Angelopoulos, K., Philippopoulos, A., Tsionas, E. (2008). Does public sector efﬁciency matter? Revisiting the relation between fiscal size and economic growth in a world sample. Public Choice, 137, 245-278.
Armey, 1995 – Armey, D. (1995). The Freedom Revolution, Regnery Publishing Co., Washington, D.C.
Barro, 1989 – Barro, R. J. (1989). A cross country study of growth, saving, and government,
Barro, 1990 – Barro, R.J. (1990). Government spending in a simple model of endogenous growth. Journal of Political Economy, 98(5), 103-125.
Brons, de Groot, Nijkamp, 1999 – Brons, M., de Groot, H.L.F., Nijkamp, P. (1999). Growth effects of fiscal policies. Tinbergen Discussion Paper, Amsterdam: Vrije Universiteit.
Chen, Lee, 2005 – Chen, S.T., Lee, C.C. (2005). Government size and economic growth in Taiwan: A threshold regression approach. Journal of Policy Modeling, 27, 1051-1066.
Folster, Henrekson, 2001 – Folster, S., Henrekson, F. (2001). Growth effects of government expenditure and taxation in rich countries. European Economic Review, 45(8), 1501-1520.
Grossman, Helpman, 1991 – Grossman, G.M., Helpman, E. (1991). Quality ladders in the theory of growth. Review of Economic Studies, 58(1), 43-61.
Hadri, 2000 – Hadri, K. (2000). Testing for stationarity in heterogeneous panel data.
Econometric Journal, 3, 148-161.
Herath, 2010 – Herath, S. (2010). The size of the government and economic growth: An empirical study of Sri Lanka. SRE-Discussion Papers, 2010/05. WU Vienna University of Economics and Business, Vienna.
Ijeoma, O’Neal, 2012 – Ijeoma, E.K., O’Neal, E.J. (2012). Government expenditure and economic growth in Nigeria, 1980-2011. International Journal of Academic Research, 4(6), 205-209.
Im, Pesaran, Shin, 2003 – Im, K.S., Pesaran, M.H., Shin, Y. (2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115, 53-74.
Jones, 1995 – Jones, C.I. (1995). Time series tests of endogenous growth models. Quarterly Journal of Economics, 110(2), 495-525.
Kao, 1999 – Kao, C. (1999). Spurious regression and residual-based tests for cointegration in panel data. Journal of Econometrics, 90, 1-44.
Kneller, Bleaney, Gemmell, 1998 – Kneller, R., Bleaney, M.F., Gemmell, N. (1998). Growth, public policy and the government budget constraint: Evidence from OECD countries. Discussion Paper No.98/14, School of Economics, University of Nottingham.
Koopmans, 1965 – Koopmans, T.C. (1965). On the Concept of Optimal Economic Growth. The Econometric Approach to Development Planning, North-Holland, Amsterdam.
Kwiatkowski, Phillips, Schmidt, Shin, 1992 – Kwiatkowski, D., Phillips, P. C. B., Schmidt, P., Shin, Y. (1992). Testing for the null of stationarity against the alternative of a unit root. Journal of Econometrics, 54, 159-178.
Lucas, 1988 – Lucas, R. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22, 2-42.
Maddala, Wu, 1999 – Maddala, G.S., Wu, S. (1999). A comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and Statistics, 61, 631-52.
Maddala, Kim, 1998 – Maddala, G.S., Kim, I.M. (1998). Unit Roots, Cointegration, and Structural Change. Cambridge University Press, Cambridge.
Magazzino, Forte, 2010 – Magazzino, C., Forte, F. (2010). Optimal size of government and economic growth in EU-27. MPRA Paper, No. 26669.
Miller, Russek, 1997 – Miller, S.M., Russek, F.S. (1997). Fiscal structures and economic growth: International evidence. Economic Inquiry, 35(3), 603-613.
Pevcin, 2004 - Pevcin, P. (2004). Does Optimal Spending Size of Government Exist?. EGPA-European Group of Public Administration, Katholieke Universiteit Leuven, Belgium, Retrieved August 18, 2008, from http://soc.kuleuven.be/io/egpa/fin/paper/slov2004 /pevcin.pdf
Rahn, Fox, 1996 – Rahn, R., Fox, H. (1996). What Is the Optimum Size of Government?. Vernon K. Krieble Foundation.
Rebelo, 1991 – Rebelo, S. (1991). Long-run policy analysis and long-run growth. Journal of Political Economy, 99(June), 500-521.
Romer, 1986 – Romer, P.M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94(5), 1002-1037.
Romer, 1990 – Romer, P.M. (1990). Endogenous technological change. Journal of Political Economy, 98(5), 71-102.
Scully, 1994 – Scully, G.W. (1994). What is the Optimal Size of Government in the US?. National Center for Policy Analysis, Policy Report 188.
Solow, 1956 – Solow, R. (1956). A contribution to the theory of economic growth. Quarterly Journal of Economics, 70, 65-94.
Swan, 1956 – Swan, T. (1956). Economic growth and capital accumulation. Economic Record, 32(Nov), 334-361.
Vedder, Gallaway, 1998 – Vedder, R.K., Gallaway, L.E. (1998). Government size and economic growth. paper prepared for the Joint Economic Committee, Retrieved September 11, 2008, from http://www.house.gov/jec/growth/function/function.pdf
World Bank, 2016 – World Bank, (2016). World Development Indicators. http://data.worldbank.org/data-catalog/world-development-indicators