CHAPTER THREE
RESEARCH METHODOLOGY
3.1 Introduction
This chapter presents the methodology adopted for examining the effect of government revenue on
economic growth in Nigeria. It describes the research design, sources of data, model specification,
measurement of variables, estimation techniques, and diagnostic tests employed in the study. The
methodology is intended to ensure that the results obtained are reliable and suitable for achieving the
study's objectives.
3.2 Research Design
This study adopts the ex-post facto research design. This design is appropriate because it relies on
historical data that have already been generated and cannot be manipulated by the researcher. The
study uses annual time-series data on government revenue and economic growth in Nigeria.
The ex-post facto design is widely used in macroeconomic studies because it enables researchers to
investigate relationships among economic variables using existing data obtained from official
publications.
3.3 Sources of Data
The study relies exclusively on secondary data.
The data will be obtained from:
Central Bank of Nigeria (CBN) Statistical Bulletin
National Bureau of Statistics (NBS)
Federal Inland Revenue Service (FIRS)
World Bank World Development Indicators (WDI)
The study covers the period 1990–2024, subject to data availability.
3.4 Model Specification
The model is adapted from previous studies on government revenue and economic growth in Nigeria.
The functional relationship is specified as:
RGDP = f(OR, NOR, TR)
Where:
RGDP = Real Gross Domestic Product
OR = Oil Revenue
NOR = Non-Oil Revenue
TR = Tax Revenue
The econometric model is expressed as:
RGDPₜ = β₀ + β₁ORₜ + β₂NORₜ + β₃TRₜ + μₜ
Where:
β₀ = Constant
β₁–β₃ = Parameters to be estimated
μ = Error term
t = Time period
3.5 Description of Variables
Variable
Measurement
Expected Sign
Real GDP
Proxy for Economic Growth
Dependent Variable
Oil Revenue
Government oil earnings
Positive (+)
Non-Oil Revenue
Revenue from non-oil sources
Positive (+)
Tax Revenue
Total tax collections
Positive (+)
3.6 A Priori Expectation
The study expects all explanatory variables to have a positive relationship with economic growth.
Specifically:
β₁ > 0
β₂ > 0
β₃ > 0
This implies that increases in oil revenue, non-oil revenue, and tax revenue are expected to increase
Nigeria's Real GDP.
3.7 Estimation Technique
The study employs the Autoregressive Distributed Lag (ARDL) estimation technique.
The ARDL approach is appropriate because:
It accommodates variables integrated at I(0) and I(1).
It estimates both short-run and long-run relationships.
It performs well with relatively small sample sizes.
It provides efficient estimates when variables are cointegrated.
Before estimating the ARDL model, the following tests will be conducted:
Descriptive Statistics
Unit Root Test (ADF)
ARDL Bounds Cointegration Test
Long-run Estimation
Error Correction Model (ECM)
Diagnostic Tests
Stability Tests (CUSUM and CUSUMSQ)
3.8 Decision Rule
The hypotheses will be tested at the 5% level of significance (0.05).
Decision rule:
If Probability (p-value) < 0.05, reject the null hypothesis.
If Probability (p-value) > 0.05, fail to reject the null hypothesis.
3.9 Diagnostic Tests
To ensure the reliability of the estimated model, the following diagnostic tests will be conducted:
Breusch-Godfrey Serial Correlation Test
Breusch-Pagan-Godfrey Heteroskedasticity Test
Jarque-Bera Normality Test
Ramsey RESET Test
CUSUM Stability Test
CUSUMSQ Stability Test
These tests help verify that the estimated model is statistically valid and satisfies the assumptions of
regression analysis.
3.10 Summary of Methodology
This chapter described the research methodology adopted for the study. It explained the research
design, data sources, model specification, variable measurements, estimation techniques, and diagnostic
procedures. The next chapter will present the data analysis, estimation results, interpretation, and
discussion of findings.