Fiscal Deficit Episode in Nigeria: What Is the Percentage of Error Correction between Public Revenue and Expenditure?

In the last three decades, almost every year except 1995 and 1996 respectively, fiscal operations of government in Nigeria have been ended in deficits. This paper determines the percentage of error correction between government revenue and expenditure in Nigeria from 1980-2018. Secondary data sourced from Central Bank of Nigeria, 2018, are analysed. Augmented-Dicky-Fuller (ADF) unit root test is used to examine the properties of the variables. Using Engel-Granger 2-step procedure to assess cointegration, error correction technique of estimation employed provides interesting results. The findings reveal that it takes two years and five months, approximately 42% as speed of adjustment, for the variables to be at equilibrium, the development that makes fiscal deficit inevitable in Nigerian economy. To widen the scope of revenue generation, it is, therefore, pertinent for authorities to ensure that investments are undertaken to close the gap between revenue and expenditure within the shortest time possible.


Introduction
Theoretical economic literature on growth and development emboldens authorities of developing economies to embark on fiscal deficits so as to stimulate steady economic growth process and particularly, bail the economy out of recession (Archibald & Greenidge, 2003). However, experience have shown that huge and perpetual fiscal deficits have been accompanied by large debts, high interest rates and inflation in most of transition economies particularly in Latin America and sub-Saharan Africa (Tshiaswak-Kashalala, 2006). Moreover, in recent years, accumulated budget deficits overtime have led to public debts reaching its highest level in economies around the world, making public debts a worldwide issue. The development situates economies under serious financial pressure and threatens the fiscal sustainability of many nations, including Nigeria (Oyeleke & Ajilore, 2014). Fiscal deficit occurs when revenue generated by government of a country falls short of its expenditure in a fiscal year.
In the last three decades, almost every year except 1995 and 1996 respectively, fiscal operations of government in Nigeria have been ended in deficits. For instance, in 1980, while government revenue was N12.993 billion, expenditure stood at N14.923 billion, making deficit of N1.975 billion. Also, in 1990, when total revenue generated in the economy of Nigeria, including oil sales, was N38.152 billion, the expenditure incurred was N60.268 billion and the deficit had increased to N22.116 billion. As at the beginning of another decade, i.e. year 2000, the discrepancy between government revenue and expenditure has become worrisome in Nigeria. In the same year 2000, while government proposed to spend N597.282 billion, N701.059 was estimated to cover both recurrent and capital expenditure. Lastly, in 2018, total government was N4185.64 trillion and expenditure was N7813.74 trillion, leaving the balance of N3628.10 trillion as deficit.
The recurring experience has been the product of dwindling revenue generation as opposed to perpetual increase in expenditure profile of government, thus, making the incidence of fiscal deficits unavoidable (Ezeabasili & Mojekwu, 2011). Fiscal deficit is becoming problematic in Nigeria as private investors are no longer willing to hold nation's debts when they become more increasingly convinced that government may either default or monetise the debts by printing more money (seignoirage) to reduce the existing real debt's value. Then, the development portends strong macroeconomic instability for the economy. Therefore, with the existing debt profile of Nigeria, economists and policy makers have begun to exercise worry over the persistent decline in revenue generation and upsurge in public spending in the economy of Nigeria.
Although, studies have investigated causality between government revenue and expenditure to assist the policy makers in ascertaining which variable particularly causes the other. Again, studies have investigated the sustainability of fiscal policy in Nigeria to determine if the economy has violated intertemporal budget constraint equation (see Oyeleke & Ajilore, 2014;Oshikoya & Tarawale, 2010). Also, to determine whether weak or strong, some research works have verified the strength of relationship between public revenue and expenditure in Nigeria. However, as its objective, no empirical work known to us has determined the number of years or the speed of adjustment to revert to equilibrium whenever there is shock to either public revenue or expenditure. This has been the focus of this paper. The rest of the paper is organised as follows: empirical review is covered in section two, the data and model specifications are presented in section three, while estimation techniques in section four. Results and discus-sion are covered in section five and conclusion of the study is presented in section six.

Empirical Review
Most existing empirical studies concentrate on fiscal policy sustainability for various countries. For example, Nwakwe (2008) analysed the government finances for Italy to determine if the economy's fiscal policy satisfies the Inter-temporal Budget Constraint (IBC) since post-Maastricht period. The author tested for stationarity, cointegration and structural breaks in the debt to GDP annual series, government revenue and expenditure quarterly series for the pe- After allowing for structural breaks, there was evidence that South Africa revenue and expenditure were cointegrated and that the economy was in the long-run weakly sustainable. Kuncoro (2011)

Gap in Existing Literature
It is obvious, given the empirical literature reviewed above, that there is possibly no empirical studies devoted to investigating the percentage of error correction between government revenue and expenditure in Nigeria, which is the objective of this study. Knowing the percentage of error correction with which government revenue variable is adjusting to attain equilibrium with government ex-penditure would provide the policy makers with information on how economy is fast or slow to move out of current debts profile. With this objective, the study employs Error Correction Model (ECM) to determine the percentage of error correction between government revenue and expenditure in Nigeria.

Data and Model Specification
The data used in this study are annual time series comprising of government revenue and government expenditure expressed as fractions of GDP for the period 1980 to 2018. The data are sourced from Central Bank of Nigeria Statistical Bulletin, 2018 issue. This study adopts transversality condition equation used in Oyeleke & Ajilore (2014) to model equilibrium condition, though with different objective. To specify the budget constraint which is the equilibrium condition for revenue and expenditure variables, we begin the model as; If limit term in Equation (1) approaches zero, we obtained the equation below, thus, the model is specified in form: where R t is the government revenue, α is the intercept which shows the degree of autonomous drift in the parameters. β represents the slope of the equation that shows the extent to which variations in government expenditure affects the value of government revenue, G t is the government expenditure and t  is the error term.

Unit Root Test
To determine the percentage of error correction back to equilibrium whenever there is shock to either government revenue or government expenditure over the period under review, which is the focus of this paper, the following steps are taken: we first test for the stationarity and order of integration of government revenue and government expenditure to ensure that either of the variables is not I(2). Thereafter, we test for cointegrating relationship between the variables and finally estimate the short-run nexus between government revenue and expenditure, using error correction model (ECM) to determine the speed of adjustment between the two variables.
In the first step, we employ only one technique of unit root test of Augumented Dickey & Fuller (1979) to ascertain the stationarity and order of integration of government revenue and expenditure. Table 1 presents the results of unit root tests.
From Table 1, results show that the variables are I(1) process, hence we move on to testing for long run relationship between government revenue and expenditure variables. To achieve this aim, Engel-Granger 2-step approach is adopted, since the variables are just two in the model.

Cointegration Test
The next step that follows is cointegration test, using Engle-Granger 2-step method (Engle & Granger, 1987 (1) is re-specified in the model below: To determine if there is cointegration between government revenue and expenditure, having generated the residuals from OLS regression estimates in Table 2, the residuals are tested for stationarity or presence of unit root, using ADF unit root test. The results are pasted in Table 3. Results from both tests confirm that the residuals are stationary at level, implying that There is cointegration or long-run relationship between government revenue and expenditure in Nigeria over the period review.

Results and Discussion
Finally, short-run association between government expenditure and revenue for the period under review is estimated with error correction model. Since cointegration is established between government revenue and expenditure, short-run analysis is undertaken to determine the percentage of error correction between the variables of interest. It is equally important to note that the technique (ECM) is capable of presenting corresponding error correction representation. Also, the technique helps to confirm the existence of long-run association between the variables because variations in public revenue is not influenced by only shocks from public expenditure, but from its own shocks as well and the magnitude of volatility between the levels of both variables. We, therefore, transform Equation (3) into first difference of the variables and re-specify it to include error correction term, hence, the Equation (3) becomes: where λ is the coefficient of the error correction term that combines reaction in association between government revenue and expenditure. It, therefore, indicates whether the model is convergent or divergent towards equilibrium. It also denotes the speed of adjustment to equilibrium after the shocks to the system.
 is the first difference of government expenditure/GDP and  is the first difference of government revenue. ECM −1 is the first-order error correction (residuals) generated from OLS regression estimated in Equation (3). The results of short run regression are presented in Table 4. The ECM has the correct sign-negative, and it is statistically significant at 1% significance level, given the value of its t-statistic (−3.039) and the probability (0.005) respectively. This indicates that there exists approximately 42% i.e. (41.78%) of error correction back to the equilibrium whenever there is distortion to government revenue in the economy of Nigeria during the period under review. Converting 42% of error correction to yearly period, it takes government revenue in Nigeria two years and five months to catch up with the movement of government expenditure before there could be equilibrium, i.e. long run relationship between the variables. Though the adjustment is not too slow, yet, government revenue has not been spontaneously adjusting back to its track whenever there is shock to variable over time. This development has been accounted for persistence in deficit finances of government of Nigeria during the period

Conclusion
This paper examined the speed of adjustment between government revenue and expenditure in Nigeria from 1980 to 2018 with sole aim of determining the percentage of error correction with which public revenue was moving behind public expenditure. After examining the stationarity properties of the series, long-run relationship between the variables of government revenue and expenditure was established. To analyse the data, error correction technique of estimation was employed to determine the percentage of error correction between government revenue and expenditure. The findings showed that approximately 42% (−0.417%) was the speed of adjustment with which public revenue was trailing public expenditure for them to be at equilibrium. The implication was that the recovery process between government revenue and expenditure was somehow slow for them to be at equilibrium whenever there was distortion between government revenue and expenditure in the economy of Nigeria during the period under review. From the findings, converting the percentage of error correction into yearly period, it would take government revenue two years and five months to catch up with government expenditure, for both variables to be in equilibrium in Nigeria.

Conflicts of Interest
The authors declare no conflicts of interest regarding the publication of this paper.