Effect of Recurrent Government Expenditure on Private Investment in Kenya

Effect of Recurrent Government Expenditure on Private Investment in Kenya

Cecilia Njeri Chege¹, Lenity Kananu Maugu², Miriam Thuo³, Onesmus Mbaabu⁴

^1,2,3Department of Social Science, Chuka University, P.O BOX 109, CHUKA

⁴Department of Economics, South Eastern Kenya University, P.O BOX 170-90200, KITUI

DOI: https://dx.doi.org/10.47772/IJRISS.2024.8100250

Received: 11 October 2024; Accepted: 17 October 2024; Published: 21 November 2024

ABSTRACT

The tenacity of the country’s entrepreneurs and the favorable climate has been credited with the rapid expansion of private sector activity in the Kenyan economy. New technology and increased private investment propel the country’s economy toward full employment, where resources are best used and economic development is achieved. Investment levels should be over 32 percent of Gross domestic product, with state investment comprising above 9 percent of Gross domestic product and private investment being 24 percent of Gross domestic product, according to Vision 2030. This objective does not appear to be attainable at the current growth pace. As a result, the goal of this research was to evaluate how recurrent government expenditure (independent variable) affects private investment (dependent variable) in Kenya. To find the casual effect link between the variables, a casual study design was used. From 1980 to 2022, a sample of 42 years from the annual series data was collected from the websites of the Kenya Revenue Authority, Kenya National Bureau of Statistics, and African Development Indicators. ARDL model was employed and the ECM was estimated since there was cointegration. Augmented Dickey Fuller was used in a unit root test to check for data stationarity. Granger Causality test and cointegration were undertaken on the stationary data to ascertain the kind of causality and presence of Long run equilibrium between variables. To confirm that the traditional linear regression assumption has been upheld, diagnostic tests were run. It was found that there was a negative significant effect of recurrent government expenditure both in the short run and in the long run on private investment in Kenya. The coefficients were, -0.363166 and -2.7596 respectively that were statistically significant at 10 percent significance level. This implies that 1 percentage increase in recurrent government expenditure lead to 0.363166 and 2.7596 decrease in private investment. A negative coefficient implies that more government spending on recurring costs may crowd out private investment. Additionally, it may increase the demand for borrowing by the government, which would raise interest rates. This may result in higher borrowing costs for businesses looking to make investments, which would discourage investment from the private sector in Kenya. The study suggests that the government implement changes to lower costs and improve the efficacy of ongoing expenditures. This means reducing bureaucracy, optimizing staffing levels, and streamlining the procurement process. Future studies may be done targeting other variables as well as in other countries.

Key Words: Government Recurrent Expenditure, Private Investment, Gross Domestic Product, Fiscal Policy variables

INTRODUCTION

Most economists such as John Keynes concurred that the amount of both private and governmental investment in an economy is one of the key elements influencing long-term economic success. A short-term improvement in aggregate demand results from increased investment, which also has the ability to expand an economy’s long-term capacity for output. A rise in private sector in both developed and developing countries promotes economic development through job and income creation. Private investment is important in Kenya since it contributes to over 60 percent of overall employment. For instance, in 2020 and 2021, private sector employment in Kenya accounted for 67.8 percent and 68.3 percent respectively of the total employment (Republic of Kenya, 2022). Private sector has played a key role in improving cost effectiveness through modernizing production process. This effectiveness leads to increase in labor efficiency leading to higher productivity and thus allowing for the production of new and improved products. Therefore, private investment is important because it contributes towards the national income and offers a foundation for the effectiveness of fiscal policy (Debrun & Kinda, 2013; Njuru, 2012).

Recurrent government expenditure refers to the continuous, ongoing expenditure of public funds on goods and services that are utilized on a regular basis, such as public wages, infrastructure maintenance, and social welfare programs (Maingi, 2010). A good example of recurrent government expenditure may be huge spending on salaries of public servants (Maina, 2013). The theoretical literature suggests that there is a link between recurrent government expenditure and private investment.  For instance, the Neoclassical theory states that too much government spending takes away valuable economic resources from business (private sector) and in return reduces economic growth (Yergin and Stanislaw, 2002). According to Ochieng (2018), in cases where recurrent government expenditure is financed through local borrowing, private investment may seriously be discouraged. This is known as crowding-out of private investment.

In the blueprint Vision 2030, Kenya’s medium term development program set a goal for the government to achieve a growth rate in the economy of above 10 percent (Republic of Kenya, 2007). To do this, the government recommended regulating government spending to prevent it from driving away private investment. The main goal of the fiscal strategy is to keep total expenditure growth in check while generating fiscal space through rationalization of spending. This space can be used to reallocate funds from non-priority to areas of priority, including spending on the flagship projects that are essential to realizing the 2030 vision. These projects such as development of infrastructures, for example roads complement the private sector. Other interventions to regulate recurrent government expenditure in Kenya include fighting corruption, removing ghost workers and limiting international travel among senior government officers (World Bank, 2020).

The Kenyan government has also identified private investments as a priority area to stimulate sustainable growth. According to sessional paper No. 12 of 2012, investment levels should be higher than 32 percent of GDP, with governmental investment being higher than 9 percent of GDP and investment from the private sector being higher than 24 percent of GDP (Government of Kenya, 2012). However, a slow expansion of private investment has been noted despite the various interventions by the Kenyan government to expand private investment. For instance, private investment has been creeping up to around 13.5 percent of GDP in 2022, which is significantly less than the projected 24.5 percent of GDP for sustainable development in Kenya’s 2030 Vision (KNBS,2023). Despite the various interventions taken to keep the recurrent expenditure at manageable levels to promote investment, this form of expenditure has increased over the years, especially the component of salaries and wages. For instance, an analysis of the Kenyan 2024/2025 budget by KPMG (2024) shows that the public sector wage bill was over 38 percent of revenue in 2021. Therefore, there is a need to empirically examine whether recurrent government expenditure has a significant influence on private investments in Kenya.

Several studies have examined the nexus between government expenditure and private investment (Okisai, 2018; Njuru et al 2014; Bella et al 2013; Joseph et al. 2016). However, there are mixed findings across countries and regions. For instance, some studies find a positive relationship (Njuru et al. 2014), others find an insignificant effect (Beni & Mwakalobo, 2009) while others find a negative relationship (Okasi, 2018; Bello et.al 2013). The contrasting findings could be due to differences in methodology, measurement of variables or country specific differences. Kenya has a dynamic economy with unique characteristics. It is therefore necessary to conduct further study in the country using a robust methodology and recent data. Moreover, there is inadequate empirical evidence on recurrent government expenditure and private investment in the country. Most of the studies done in Kenya only examine government expenditure as an aggregate variable or development government expenditure without examining the component of recurrent expenditure (Nyang’aya, 2019; Mose, 2014; Maingi, 2010). This study contributes to the existing literature by providing an empirical investigation of the effect of recurrent government expenditure on private investment in Kenya using data covering the period of 1980-2022. The study employed the autoregressive distributed lag (ARDL) model, which is important in establishing whether there is a long run or short run relationship.

LITERATURE REVIEW

The linkage between recurrent government expenditure and private investment is elaborated well by the Keynesian crowding-out theory (Traum & Yang, 2015). According to the theory, increased government expenditure on recurring costs (such salaries, maintenance, or other continuous expenditures) may crowd out private investment. Crowding-out effect implies that the economy would operate at its best level throughout time, with full employment leading to supply equaling demand and a state of equilibrium devoid of any remaining capacity. At this stage therefore, savings and investment are interest rate elastic, which implies that a unit change in interest rates leads to higher change in levels of savings and investment since they are highly responsive to interest changes in the long run. According to classical economists, the government’s active involvement in the economy, particularly through the implementation of fiscal expansionary policies, results in lower disposable income, greater interest rates, and wage increases that endanger firm profitability by reducing the profit margin to a point where it is dissuasive to private investment, which in turn results in lower productivity and, ultimately, lower output. Therefore, the classical conclude although government intervention may influence output, its impact is short lived and that in the long run fiscal policies such as government expenditure may crowd out private investment hence being harmful to growth of the economy through making declining firms exit the market.

Several studies provide empirical evidence on the relationship between government expenditure and private investment. Some researchers studied several countries in a panel setup while others conducted single-country studies. In the context of the former, Joseph et al. (2016) conducted an empirical study to determine the connection between capital development in the private sector and fiscal policy in a number of West African countries. The study employed time series data collected from these countries from 1993 to 2014 with Ordinary Least Squire (OLS).The findings showed that government spending on development and tax revenue support private investment. However, there was a crowding out impact noted in recurring expenses, non-tax revenue, and external debt. Similarly, Okisai (2018) investigated the impact of economic growth on private investment in Kenya for the period 1970-2007. The Multivariate cointegration technique was utilized in the empirical investigation to ascertain the characteristics and correlation between the two variables. The results invalidated the theory that recurring expenditures encourage investment behavior in both people and businesses. Therefore, the current study seeks to contribute and shed more light by using a more robust analyzing technique (ARDL) model and relatively more recent set of data covering the period between 1980 and 2022 that generated reliable findings.

A research conducted in Nigeria on Government spending and private investment, Bello et al. (2013) found that recurring expenditure significantly affected private investments negatively. The research covered a period of 35 years from 1975 to 2019. Increased recurrent government expenditure led to crowding out of private investment in Nigeria. Contrastingly, a study conducted on the impact of government expenditure on private investments in Kenya by Njuru et al. (2014) indicated a positive impact of recurrent expenditure on private investment. The study used the VAR technique for the period covering 1963-2012.The difference in results could be attributed by the fact that the studies were conducted in two different countries. However, the current study sought to add more knowledge on the effect of recurrent government expenditure on private in investment by use of a more advanced technique (ARDL) and using a longer timeframe of 42 years (1980-2022). Using quarterly data from 1960 to 2005, Beni & Mwakalobo (2009) investigated the nature of and link between capital production in the private sector and government spending in South Africa. Without further decomposition, the study’s independent variables were recurring and development spending. The study’s findings did not clearly show a relationship between the two variables.

The contrasting findings from the above studies could be due to differences in methodology, measurement of variables or country specific differences. Kenya has a dynamic economy with unique characteristics. It is therefore necessary to conduct further study in the country using a robust methodology and recent data. This study contributes to the existing literature by providing an empirical investigation of the effect of recurrent government expenditure on private investment in Kenya using data covering the period of 1980-2022. The study employed the autoregressive distributed lag (ARDL) model, which is important in establishing whether there is a long run or short run relationship.

METHODOLOGY

Research Design

In order to determine whether there exists a cause-and-effect relationship between variables related to fiscal policy and private investment, a causal research design was used in the study. A data collection checklist was utilized to gather the data since the study involved secondary data. The Africa Development Bank indicators, Kenya Revenue Authority and Kenya National Bureau of Statistics provided data for this study, covering the years 1980 through 2022. The procedure for gathering data involved issuing download requests for time series historical data on annual public spending, interest rates, corporate taxes, and private investment from the relevant websites.

Data Analysis

In addition to identifying pre-existing patterns of relationships among data groupings (Kothari ,2009), the Augmented Dickey-Fuller (ADF) test was used in the unit root test to determine whether the data were stationary, with additional differencing being done as needed to make the data stationary. The study employed the autoregressive distributed lag (ARDL) model and the ECM was estimated since there was cointegration. The researcher additionally employed Granger Causality to establish the relationship among the variables. A Cointegration test between the variables was used to assess whether a long-term relationship exists or not. In order to confirm that the assumptions of the Classical Linear Regression Model (CLRM) are valid, diagnostic checks were carried out at the end. Multicollinearity, heteroscedasticity, and autocorrelation were examined. After receiving approval from the Chuka University Ethics Committee, research permission was requested from the National Council for Science, Technology and Innovation (NACOSTI) in order to gather the necessary data. Confidentiality and informed consent are ethical considerations.

Theoretical Model

This study’s analytical methodology was designed to align with a flexible accelerator model, which is grounded in Keynesian investment theory (Keynes, 1936).;

\(\Delta I_{t} = \beta \left( I_t^* – I_{t-1} \right) \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.1\)

In this case, I^*_t   represents the desired investment at time t, I_t-1 denotes the actual investments at time t, and β is the partial adjustment coefficient.

In Equation (3.2), it was assumed that the targeted capital K* would be proportionate to the

Projected Output Y_t

\(K_{t}^* = \delta Y_{t} \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.2\)

Where the input-output ratio, which is taken to remain constant, is represented by δ.

The following was an expression of the real private capital stock during time t:

\(K_t = K_{t-1} + I_t – \alpha K_{t-1} \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.3\)

Where α is the rate for depreciation, K_t   is the capital stock for the current period, and K_t-1 is the capital stock for the prior period. It should be highlighted that, by description, gross private investment at a time t is presented as follows in order to understand its implications:

\(I_t = K_t – K_{t-1} + \alpha K_{t-1} \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.4\)

Ensuring a steady state condition and adding the lag operator L, which is expressed as LK_t = K_{t-1 the} Equation (3.4) becomes;

\(I_t = \left[ 1 – (1 – \alpha) \right] L K_t^* \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.5\)

Where I_t^* is the desired investment at time t. When we combine both equation (3.2) and (3.5) we get the following equation (3.6)

\(I_t^* = \left[ 1 – (1 – \alpha) L \right] \delta Y_t \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.6\)

Equation (3.7) below shows the dynamic accelerator model that is generated by substituting equation (3.6) into equation (3.1).

\(\Delta I_t = \beta \left\{ \left[ 1 – (1 – \alpha) L \right] \delta Y_t – I_{t-1} \right\} \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.7\)

The methodology of Blejer and Khan (1984), which permits private investment to fluctuate depending on economic circumstances, was adopted by this study as follows:

\(\beta = f \left[ \frac{\sum_{j=1}^m x_{jt}}{I_t^* – I_{t-1}}, \frac{\sum_{j=1}^k x_{it}}{I_t^* – I_{t-1}} \right] \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.8\)

Where X_jt is a fiscal variable vector, Z_it a vector made up of non-fiscal factors. According to this specification, investment is influenced by the parameters mentioned above through the method of adjusting actual investment to target levels. Equation (3.8) is able to be expressed linearly as follows:

\(\beta = \mu_0 + \frac{1}{I_{t-1}^*} \left[ \mu_i \sum_{i=1}^k x_{it} + \mu_j \sum_{j=1}^m z_{jt} \right] + \epsilon_t \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.9\)

where the error term for white noise is ɛ_t , the intercept is µ₀, and the coefficients of the fiscal variables and non-fiscal variables are µ_i and µ_j, respectively. Equation (3.9) is substituted into (3.1) and it can be solved for to give the following equation:

\(I_t = \mu_0 I_t^* + \mu_i \sum_{i=1}^k x_{it} + \mu_j \sum_{j=1}^m z_{jt} + (1 – \mu_0) I_{t-1} + \epsilon_t \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.10\)

The intended investment I* cannot be observed in equation (3.10). Equation (3.6) can be substituted into equation (3.10) to get the following result:

\(I_t = \mu_0 \left[ 1 – (1 – \alpha) L \right] \delta Y_t + \mu_i \sum_{i=1}^k x_{it} + \mu_j \sum_{j=1}^m z_{jt} + (1 – \mu_0) I_{t-1} + \epsilon_t \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.11\)

Equation (3.11) suggests that a vector of both fiscal and non-fiscal variables also contributes to the explanation of investment, in addition to changes in output as suggested by the classical flexible accelerator model (Shapiro, 1992).

Model Specification

The ARDL model was used and a cointegration test was conducted to establish the nature of the relationship of the variables. A long run relationship was established, which led to the use of the ECM model as specified in the equation 3.12.

\(\Delta PI_t = \beta_0 + \sum_{i=1}^p \beta_1 \Delta PI_{t-1} + \sum_{i=1}^p \beta_2 \Delta GDP_{t-1} + \sum_{i=1}^p \beta_3 \Delta GEX_{t-1} + \lambda U_{t-1} + \epsilon_t \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots \ldots 3.12\)

Where;

\(\beta_0\): Private investment that is not influenced by Recurrent government expenditure

\(\beta_1\): Represents the elasticity parameter of private investment

\(\beta_2\): Represents the elasticity parameter of the recurrent government expenditure

\(\beta_3\): Represents the elasticity parameter of the Interest rate

\(\beta_4\): Represents the elasticity parameter of the Inflation

\(\beta_5\): Represents the elasticity parameter of the dummy variable

\(RGX_{t-1}\): The lagged Recurrent government expenditure

\(RIR_{t-1}\): The lagged Real Interest Rate

\(INF_{t-1}\): The lagged Inflation Rate

\(GDP_{it}\): GDP per capita at time \(t\)

\((D_1)\): Dummy variable to represent Kenya’s economic structural breaks. Years having structural breaks and those without will be denoted by 1 and 0, respectively.

\(EC_{t-1}\): Error correction term

\(\lambda\): Denotes the elasticity parameter of the Error correction term

\(\varepsilon_t\): The error term representing other factors that affect private investment other than the independent variable above that have not been captured in the model in time \(t\)

RESULTS AND FINDINGS

Descriptive Statistics and Normality Test

The descriptive statistics analysis summarizes the variable data in terms of standard deviation, minimum, maximum and mean values. The mean provides an overview of the data distribution, and for the data to be statistically reliable, the mean and standard deviation should be within a close range. Normally distributed data enables drawing viable conclusions and recommendations. The variable data analysis findings are presented in Table 4.1.

Table 4.1:  Descriptive Statistics and Normality Test

Source: Authors’ Computation

The mean and standard deviation of recurrent government expenditure were 15.32326 and 2.412057, respectively. The low standard deviation implies that the recurrent expenditure values were within a close range around the mean. Based on the recommended threshold of 25% for recurrent expenditure, the mean value was lower, suggesting that the government’s recurrent spending was below the desired level. Private investment had a mean of 9.307907 and a standard deviation of 3.915811. Compared to the recommended threshold of 24% for private investment, the mean was lower, implying limited access to financing and structural issues hindering private sector investment. The standard deviation indicates moderate variations in private investment levels during the period.

The real interest rate (RIR) had a mean of 7.133076 and a standard deviation of 6.639414. The high standard deviation suggests significant variations in real interest rates over the period, which could be attributed to volatile monetary policies. The lowest value was -10.096, indicating periods of negative real interest rates, while the highest was 21.09633. Gross Domestic Product (GDP) showed a mean of 803.5191 and a standard deviation of 592.8528. The large standard deviation indicates high variability in GDP over the period. The minimum GDP value was 226.5211, while the maximum reached 2099.302, reflecting substantial economic growth or fluctuations during the study period. Inflation had a mean of 11.42649 and a standard deviation of 8.337009. The high standard deviation suggests considerable variability in inflation rates. The minimum inflation rate was 1.554328%, while the maximum reached a high of 45.97888%, indicating periods of both low and very high inflation. Regarding normality, the skewness values for all variables were within the ±3 threshold, and the kurtosis values were also within the ±2 threshold, indicating that the variables followed a normal distribution.

Unit Root Test

The study used Augmented Dickey Fuller Test to detect presence of unit root. The test examines whether the variables are non-stationary (possess unit roots) or stationary (lack unit roots). The null hypothesis of the ADF test postulates that the variable has a unit root (non-stationary), while the alternative hypothesis posits that the variable is stationary. Table 4.2 presented ADF Unit Root Test.

Table 4.2: Unit Root Test Results at Level

Source: Author’s Computation

The results of the ADF test at levels, presented in the in table 4.2 indicate that inflation and real interest rate were stationary at level while all other variables were not stationary in their level forms when the ADF test was performed. The ADF statistics for inflation and real interest rate were -3.536 and -4.609 respectively which are less than the critical value at 5% significance level showing the absence of unit root and therefore stationary at level.

Table 4.3: Unit Root Test Results at First Difference

Source: Author’s Computation

However, after first differencing, the test statistics for dRGX (-6.414), dPI (-7.378), and dGDP (-4.278) were less than the critical value at 5 per cent significance level, that is, -2.955 as indicated in Table 4.3. This result leads to the rejection of the null hypothesis of non-stationarity, implying that the variables become stationary after first differencing.

Granger Causality Test

Granger causality is a situation where previous values of a given time series is capable of predicting future values of a different time series (Granger, 1969). The model assumes existence of previous information of one variable that enhances the prediction of another variable thereby causality relationship (Shojaie & Fox, 2022). Granger causality was used in the study to determine presence of relationship between the study variables (Ceesay et al., 2019). The Granger causality tests were conducted to examine whether the past values of one variable can help predict the current values of another variable, thereby indicating a causal relationship between the variables as demonstrated in Table 4.4.

Table 4.4: Granger Causality Test Outcomes

Source: Author’s Computation

From the findings above, a neutral causality exists between Private investment and Recurrent government expenditure as shown by the P-values that are greater than the Critical value of 0.05 (P=0.722>0.05 and P=0.905>0.050). The null hypothesis of no Granger causality between Private investment and Recurrent government expenditure is rejected since the variables shows that they are statistically independent.

Gross domestic product shows a unidirectional relationship with Private investment. This is because by the probability for GDP is (0.004>0.05) which is less than the Critical value of 0.05. Therefore, GDP Granger cause Private investment. For both relationship Inflation and Real interest rate with Private investment, the P-values exceeded the Critical value 0.05 (P=0.249>0.05, P=0.079>0.05, P=0.653>0.05 and P=0.698>0.05) respectively.

Test for Cointegration

The bounds test for cointegration was employed to examine if there exists a long-run equilibrium relationship among the variables under study. The null hypothesis suggests the absence of a cointegrating equation, and it was rejected if the test statistic value exceeds the upper bound critical values. The results of the cointegration test are presented in Table 4.5

Table 4.5: Model Cointegration Test Results

Source: Author’s Computation

5 Pesaran/Shin/Smith (2001) ARDL Bounds Test

H0: no levels relationship F = 12.537

Because the F-statistic (12.537) is greater than the upper bounds for all the levels of significance; 3.35, 3.79, 4.18 and 4.68, the study therefore rejects the null hypothesis and concludes that there is cointegration.

Correlation Analysis

Before carrying out the estimation of the models a correlation analysis was conducted to determine the relationship among the study variables and the direction of the relationship. The results are presented in the table 4.6.

Table 4.6: Correlation Analysis

Source: Author’s Computation

This correlation matrix provides insights into the relationships between various economic variables in the study. The analysis reveals that Private Investment (dpi) has minimal correlations with most other variables. It shows a slight positive relationship drgx), inflation (dinf) and GDP growth (dgdp), while exhibiting minor negative real interest rate (drir). Recurrent Government Expenditure (drgx) shows positive correlations with inflation (dinf), and real interest rate (drir), while negatively correlating with GDP growth (dgdp).

Results

This study examined the relationships between Recurrent Government Expenditure and Private Investment (DPI) from 1983 to 2022. It also included control variables which are Real Interest Rate (RIR), Gross Domestic Product (GDP), Inflation (INF) and a Dummy variable (D1) representing the economical structural breaks for the period. The results are presented in table 4.7.

Table 4.7: Regression Results for Recurrent Government Expenditure on Private Investment

Source: Author’s Computation

The model was presented as follows;

PI= 0.9520 – 0.3631D(RGX) + 0.1550RIR- 0.1425NF + 0.0037D(GDP) – 0.7990D1 – 1.3160ECT

The model demonstrates a moderate fit, with an R-squared of 0.7760, indicating that it explains a 77.60% of the variance in Private Investment. The P value was 0.0341< 0.1 showing that the model is fit. The constant coefficient was 0.9586 and not statistically significant representing private investment when all other variables are held constant. The model was fit shown by the P-value of 0.0341 which is less than 10% significance level. The coefficient of the Private Investment (DPI) is -1.3160, with a high level of statistical significance (p<0.001). This coefficient signifies a negative adjustment mechanism, implying that deviations from the equilibrium in the previous period are strongly corrected in the current period.

In the long run, the coefficient on Recurrent Government Expenditure (DRGX) is -2.7596, which is statistically significant (p=0.0325), indicating a significant long-term  negative effect on Private Investment. This negative result agrees with a study conducted by Wang (2005) which stated that private investment maybe boosted or impended depending on the nature of government spending where recurrent government expenditure affects private investment negatively (crowd out private investment). Contrastingly, Monadjemi (1996) found that government spending had significant effect on private investment in the United States of America. Real Interest Rate (RIR) exhibits a coefficient of -0.08903, which is not statistically significant (p=0.101), suggesting no significant long-term relationship between Real Interest Rate and Private Investment. The coefficients for Gross Domestic Product (DGDP) (0.0028, p=0.466), Inflation (INF) (-0.0122, p=0.751), and the dummy variable (DI) (-0.0671, p=0.406) are also not statistically significant, implying no meaningful long-term effects on Private Investment.

In the short run, Recurrent Government Expenditure (DRGX) displays a coefficient of -0.3631, which is statistically significant (p=0.0316), suggesting a significant short-term effect on Private Investment. This implies that 1% increase in recurrent government expenditure lead to 0.36 decrease in private investment. The study agrees with Wu and Zhang (2009) from China, were the findings indicated that government spending crowds out private sector in the short run.  For Real Interest Rate (RIR) the coefficient is 0.0037 and significant (p=0.003), implying a one percent  increase in Real Interest Rate may increase Private Investment by approximately 0.0037  in the short run. The inflation (INF) coefficient is -0.1425 and significant at 1 % significance level (P=0.002) which implies that one percent increase in Inflation leads to a decrease of 0.1425 of private investment. The Gross Domestic Product (DGDP), and the dummy variable (D1) show no significant short-term effects (-0.0030, p=0.605; and -0.2961, p=0.754 respectively).

CONCLUSION

The purpose of the study was to establish the effect of recurrent government expenditure on private investment in Kenya for the period 1980 to 2022. Granger Causality test, Cointegration and linear regression tests was performed. Recurrent government expenditure exhibits negative coefficient both in the short run and in the long run, -0.363166 and -2.7596 respectively that were statistically significant at 10% significance level. This implies that 1% increase in recurrent government expenditure lead to 0.363166 and 2.7596 decrease in private investment. With an R-squared of 0.7760, the model shows a decent fit and accounted for 77.60% of the variance in private investment. When all other factors are held constant, the constant coefficient, which represents private investment, was 0.9586133 and not statistically significant. The P-value of 0.0341, which is less than the 10% significance level, indicates that the model was fit.

Based on the findings, recurrent government expenditure exhibited a negative and significant effect on private investment both in the short run and in the long run.  A negative coefficient implies that more government spending on recurring costs (such salaries, maintenance, or other continuous expenditures) may crowd out private investment. Additionally, it may increase the demand for borrowing by the government, which would raise interest rates. This may result in higher borrowing costs for businesses looking to make investments, which would discourage investment from the private sector in Kenya. This result agrees with the theory of crowding out by John Keynes which emphasized that government expenditure eventually crowd-out investment due to high interest rate.

The Government should put reforms into place to increase the effectiveness of ongoing expenses. To cut costs, this entails streamlining the procurement process, optimizing staffing levels, and cutting bureaucracy. By implementing strict budgetary controls and monitoring systems, it can improve the fiscal discipline. To be able to stop resource wastage and abuse, it should enforce budgetary caps and guarantee accountability and transparency in financial administration. The government should improve the effectiveness and accountability public financial management systems when handling ongoing expenses. It should also take care when managing governmental debt to prevent it from stifling private investment. The government should also keep an eye on debt levels and focus its spending on long-term sustainability and good economic rewards.

Valuable suggestions for further research arise from this study. For instance, the study was based on time series data between 1980-2022 that involved recurrent government expenditure and private investment. Future studies should consider other variables such as development expenditure, public debt, taxes and determine how they affect private investment. Other future studies can consider using different methodology technique to determine the effect of recurrent government expenditure on private investment in Kenya. This study was also localized in Kenya. Other studies may be done within the East African region to compare the effects of recurrent government expenditure on private investment.

REFERENCES

1. Bello, M. Z., Nagwari, A. B., & Saulawa, M. A. (2013). Crowding in or crowding out? Government spending and private investment: The case of Nigeria. European Scientific Journal, 8(28) 9 – 22.
2. Beni, A., & Mwakalobo, S. (2009). Economic Reforms in East African Countries: The Impact on Government Revenue and Public Investment. Open Access Dissertations, 66.
3. Blejer, M.S & Khan, M.S. (1984). Government policy and private investment in developing countries. IMF Staff Papers, 31(2), 379-403.
4. Ceesay, E. K., Belford, C., Fanneh, M. M., & Drammeh, H. (2019). Relationship between export, imports and economic growth: an export-led growth strategy for the Gambia using the granger causality test. International Journal of Social Sciences Perspectives, 4(2), 38-47.
5. Debrun, X., & Kinda, T. (2013). That Squeezing Feeling: The Interest Burden and Public Debt Stabilization. IMF Working Paper No. WP/13/125
6. Gitahi, S.W. and Mwilaria, S.M. (2015), “The effects of public debt on private investments in Kenya (1980-2013)”. International Journal of Development and Sustainability, Vol. 4, Issue 8. www.isdsnet.com/ijds.
7. Government of Kenya (2012), Sessional on Kenya Vision 2030. Ministry of State for Planning, National Development and Vision 2030. https://vision2030. go.ke/publication/sessional-paper-no-of-2012-on-kenya-vision-2030/.
8. Granger CWJ. 1969. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37:424–38
9. Hermes, N., & Lensink, R. (2001). Capital flight and the uncertainty of government policies. Economics letters, 71(3), 377-381.
10. Kamundia, S. W. (2015). The effects of public debt on private investments and economic growth in Kenya (1980-2013). Unpublished master of economics project, Kenyatta University, Nairobi.
11. Keynes, J.M. (1936). The general theory of employment, interest and money. New York: Oxford university press
12. KMPG. (2024, June). 2024/25 Budget Brief.  https://assets.kpmg.com/content/ dam/kpmg/ke/pdf/tax/ 2024/Kenya%202024-2025%20Budget%20Brief.pdf
13. Kenya National Bureau of Statistics (2023) Economic Survey, 2023. KNBS, Nairobi.
    https://www.knbs.or.ke/economic-survey-2023/
14. Kothari, C.R. (2009). Research Methodology Methods and Techniques. New Age Publications (Academic).
15. Maina, R.W. (2013): Does the composition of Public Expenditure affect economic growth? Evidence from Kenya, an unpublished MA thesis, University of Nairobi, Nairobi
16. Maingi, J.N. (2010). The impact of government expenditure on economic growth in Kenya 1963-2008. School of Economics Kenyatta University.
17. Monadjemi, M. S. (1996). Public Expenditure and Private Investment: A Study of the UK and the USA. Applied Economics Letters, 3(10), 641-644.
18. Mose, N.G. (2014). Effect of government expenditure on economic growth in East Africa. A disaggregated model 1980-2010. School of Economics Egerton University.
19. Njuru, S. G., Ombuki, C., Wawire, N., & Okeri, S. (2014). Impact of Government Expenditure on Private Investment in Kenya. Research Journali’s Journal of Economics, 2(8), 1-19
20. Njuru, S. G. (2012). Effects of fiscal policy on private investment in Kenya 1964–2010. School of Economics Kenyatta University.
21. Nyang’aya, T.G. (2019). Effect of public investment on private investment in Kenya 1980-2017. School of Economics University of Nairobi.
22. Ochieng, E. (2018). Effects of government borrowing on financial development in Kenya 2008-2017. School of Business University of Nairobi.
23. Okisai, E. (2018). Impact of economic growth on private investment in Kenya. VI(5),361–372. Fiscal policy and private investment in selected West African countries, CBN Journal of Applied Statistics, ISSN 2476-8472, The Central Bank of Nigeria, Abuja, Vol. 07, Iss. 1, pp. 277-309
24. Republic of Kenya. (2022), “Economic Survey”, Nairobi: Government Printers. Retrieved  from, https://www.knbs.or.ke/download/economic-survey-2022/
25. Republic of Kenya. (2007). Kenya Vision 2030: A globally competitive and prosperous Kenya. Nairobi: Government Printer
26. Shojaie, A., & Fox, E. B. (2022). Granger causality: A review and recent advances. Annual Review of Statistics and Its Application, 9, 289-319.
27. Wang, B., (2005). Effects of government expenditure on private investment: Canadian empirical evidence. Empirical Economics, 30(2), 493-504.
28. World Bank. (2020). Kenya public expenditure review. Washinton DC: World Bank
29. World Bank. (2018). Survey of the Kenyan private equity and venture capital landscape. Washington DC: World Bank
30. Wu, D., & Zhang, Z. (2009, March). Effects of government expenditure on private investment-China empirical evidence. In 2009 First International Workshop on Education Technology and Computer Science (Vol. 3, pp. 86-89).IEEE.