Page 53 - Azerbaijan State University of Economics
P. 53
Javid Seyfullali: Tax Revenue and Economic Growth in Resource-Rich Country:
Empirical Evidence from Azerbaijan
METHOD AND DATA
Autoregressive Distributed Lag Bounds Testing (ARDLBT) presented by Pesaran et
al. (2001) is used to analyze the relationships between variables. The advantages of
the ARDLBT method include the ability to apply to small samples, I(1) and I(0) series
simultaneously and to estimate long-run and short-run coefficients (Oteng et al. 2006;
Sulaiman et al.,2010).
We use 2 models in this paper, the purpose of the Model 1 and 2 are to find out the short-
term and long-term impact of government spending on general economic growth and non-
oil economic growth, respectively. We will use real non-oil GDP growth as a dependent
variable in the second model while using total real GDP growth in the first model.
The 3 control variables will be added to the model. Capital formation and trade
openness have been used widely in previous research papers (Christie (2012);
Mendonca and Caicedo (2014); Asimakopoulos and Karavias (2015); Olaoye et al.
(2020)) on this topic as control variables. Additionally, the oil price change variable
will be added to the models as oil price changes significantly impact non-oil GDP
(Aliyev and Nadirov, 2016) and obviously total GDP in Azerbaijan. The list of
variables is presented below:
Table 1: Information on variables
Variable Symbol Model Definition
inclusion
Real GDP RGDPG 1 CPI (consumer price index) is used
growth to convert nominal GDP to real GDP
Real non-oil RNOGDPG 2 CPI is used to convert nominal non-
GDP growth oil GDP to real non-oil GDP
Real CPI is used to convert nominal
government RGOVREV 1,2 government expenditure to real
revenue government expenditure
CPI is used to convert nominal
Real capital CAPINV 1,2 capital investments to real capital
investment
investments
Real total trade is calculated as the
Real total RTOTTRADE 1,2 sum of imports and exports, CPI is
trade used to convert nominal total trade to
real total trade
Oil price OILPRCG 1,2, quarterly data of Brent Crude price
change (FRED, 2023)
Notes: Real government expenditure, real capital investments, and real total trade
variables are included to the Model 1 and 2 in the logarithmic forms.
Source: Compiled by author
52