An Empirical Investigation of the Impact of Monetary Policy on Economic Growth in Zambia

Effective monetary policy plays a very significant role in achieving macroeconomic stability of any given economy. A critical and most important aspect of the monetary policy transmission process is the pass-through which is defined as the magnitude (degree) and speed with which a change in the monetary policy is passed on to the real economy (Aydin, 2007). In this regard, the impact of monetary policy on the real sector variables of the economy such as output and prices becomes a remarkable issue in discussions surrounding monetary policy transmission process.

Despite the changing financial and monetary environment in Zambia, the degree and speed of monetary policy pass-through to the real sector has not been adequately investigated. This paper intends to fill this gap by building on Ngoma and Chanda (2022) to assess the efficacy of monetary policy on economic growth in Zambia. Specifically, the study intends to answer the following research questions:

i. What is the magnitude of monetary policy transmission to economic growth in Zambia?

ii. How effective are the various channels of monetary policy transmission in Zambia?

The study is significant to the subject of monetary policy transmission as it offers the central bank off Zambia and other key stakeholders an in-depth understanding of the impact of monetary policy on the real economy. This helps to determine the monetary policy mix or the appropriate monetary policy stance that could be implemented at a particular time in order to achieve the intended goal of low and stable inflation necessary for sustainable macroeconomicgrowth.

The key theory underpinning this study is the monetarist view. The Monetarist view asserts that there is a direct link between the monetary sector and the real sector of the economy. They argue that “money matters”, thereby advocating for the use of monetary policy in influencing economic growth. Thus, monetary policy through changes in money supply thus functions by stimulating interest-responsive components of aggregate demand, primarily investment spending (Friedman, 1974). It is argued that lower interest rates have a positive impact on aggregate demand thereby increasing output.

The impact of monetary policy on the real economy has gained prominence in macroeconomic analysis over the years. This has resulted into a number of empirical studies being undertaken in both developed and developing economies. In the developed world, the early attempt to explore monetary transmission on the real sector is by Bernanke et. al (1995) in the US and this was followed by Hayo and Uhlenbrock (1999) in Germany, Farès and Srour (2001) in Canada ,Raddatz and Rigobon (2003) in the United states of America, Tena and Tremayne (2006) in the United Kingdom, among others. In the developing world, empirical evidence on the real sector impact of monetary policy is documented by Alam and Waheed (2006) in Parkstan, Sing and Rao (2014) in India, Javanovic et. al (2015) in Macedonia, Ezeaku et.al (2018) in Nigeria, Omolade and Ngalwa (2016), Moussir and Chatri (2017) in Morrocco, and Nampewo et. al (2016) in Uganda, and in Zambia by Simatele, (2004), Mwange, (2018), Zgambo, (2008), Zgambo and Chileshe, (2014), Mwange, et al., (2022), Kaponda, et al., (2022), Mwange, (2017), Kaponda, et al., 2022). The results of these studies are mixed. Most empirical evidence for developed countries suggest that monetary policy has a strong influence on economic growth while it some suggests that it is limited in developing countries (Mishra and Montiel, 2012; Ngoma and Chanda, 2022; Mwange and Meyiwa, 2022; Mwange, et.al.,2022); and Kaponda, et al., (2022).

Similar to Das (2015) and Mbowe (2015), this study employs a Johansen cointergration approach. This estimation strategy also makes it possible to estimate all co-integrating vectors when there are more than two variables and facilitates the estimation of an error correction model (ECM).

Theoretically, the relationship between monetary policy and the real sector can be investigated using Vector Auto Regressive (VAR) models estimated as follows:

Assuming k=2 and p=1, this gives:

$\left[\begin{array}{l}Y_{1 t} \\ Y_{2 t}\end{array}\right]=\left[\begin{array}{l}C_1 \\ C_2\end{array}\right]+\left[\begin{array}{ll}\alpha_{11} & \alpha_{12} \\ \alpha_{21} & \alpha_{22}\end{array}\right]\left[\begin{array}{l}Y_{1, t-1} \\ Y_{2-1}\end{array}\right]+\left[\begin{array}{l}\varepsilon_{1 t} \\ \varepsilon_{2 t}\end{array}\right]$

Explicitly, this can be rewritten as follows:

The vector Y in this study contains six endogenous variables whose vector is specified as:

It is expressed as:

Taking the natural logarithm of the equation and assuming linearity among the variables gives:

Where: $G D P$ is Gross Domestic Product, alr is average lending rate, $d p r$ is 6 months deposit rate, exr is the USD/ZMW exchange rate, inf is the inflation rate and pcr is credit to the private sector.

The Granger representation theorem allows the estimation of ECM if cointegration is established in the model (Engle and Granger, 1987). The ECM is specified as:

where:

$e c t_{t-1}$ is the error correction term measuring period _t-1 deviation from the long-run stationary relationship through coefficient $\Psi_0 ; \Delta$ is the difference operator.

The study employs quarterly data collected from the Bank of Zambia and Zambia Statistical Agency over the period 2001q1 to 2021q2 (Table 1 and appendix 1).

Table 2 presents descriptive statistics of the data set.

Unit root tests were carried out using the Augmented Dickey-Fuller (ADF) test to ensure that spurious relationships, usually associated with time series data are avoided (Table 3).

Based on the Akaike Information Criterion (AIC) and the Shwarz Bayesian Information Criterion (SBIC), two lags were appropriate for the model of the study (Table 4).

Having established that all variables are integrated of the same order I (1), a Johansen cointegration test was undertaken. The Trace Statistic reveals the existence of a cointegrating (long-run) relationship among the variables (Table 5)

The long-run passthrough results from monetary policy variables to GDP are shown in table 6.

The results showed that in the long run, average lending rate $(a l r)$, inflation rate (lninf) and credit to the private sector (lnpcr) had a significant negative impact on gross domestic product (lngdp) while exchange rate (exr) and deposit rate ( $d p r$ ) were found to have a significant positive impact. This implies that one percentage increase in lending rate, inflation rate and private sector credit would lead to a percentage decrease of GDP by 0.013, 0.070 and 0.334, respectively. The result on lending rate is consistent with theory in that contractionary monetary policy through a rise in lending interest rate constrains economic growth by decreasing money supply in the economy for real sector investments. Empirically, this has been supported by the work of Alavinasab (2016) in Iran, Imoughele and Ismaila (2016) in Nigeria and Onderi and Njuru (2015) in Kenya who found contractionary monetary policy to have a negative influence on economic growth.

Looking at inflation, economic theory suggests that if inflation becomes too high, it erodes the purchasing power funds which subsequently reduces the value of real sector investments. This is in line with the empirical findings of Andres and Hernando (1997) who established that too much inflation had a negative impact on economic growth of the OECD economies.

The result on the impact of the private sector credit to growth is however against the apriori expectation. There is, however, need to validate this using a different data set and different methodology. Empirically, this adds to the inconclusive debate on either private sector credit leads to economic growth or not and is consistent with other studies such as Shan and Jianhong (2006) in China and Hondroyiannis et al. (2005) in Greece that found a negative relationship between economic growth and private sector credit.

On the other hand, exchange rate and deposit rate had a significant positive influence on economic growth. Thus, holding all other things constant, a percentage increase in exchange rate and deposit rate would result into a percentage increase in economic growth of 0.016 and 0.011, respectively. This results into a long run passthrough elasticity of 2.2% from exchange rate and 2.5% from deposit rate. This suggests that exchange rate appreciation of the local currency increases the value of the country’s export earnings. Further, expansionary monetary policy associated with a rise in deposit rates encourages real sector investments since there is an assuranceofpreservingthevalueoftheinvestmentreturns.Thisisconsistentwith the findings of Imoughele and Ismaila (2016), who found that deposit interest rates and exchange rate had a positive relationship with economic growth in Nigeria.

A separate short run dynamic equation which included the error correction term extracted from the long run was estimated in order to allow for the disclosure of additional information that might have not been revealed in the long run (Table 7).

In the short run, only inflation had an impact on economic growth with a lag one. Thus, ceteris paribus, a percentage rise, a percentage increase in inflation would result into a 0.038 percentage decline in output. As opposed to the long-run dynamics, all the other variables were found to have a statistically insignificant relationship with economic growth.

The coefficient of the error correction term (ect) has a negative sign and is statistically significant. The absolute coefficient value of the ect of 0.2409, indicates that the speed of adjustment towards the long run equilibrium given that a shock occurs in the economy is about 24% in each given quarter.

In order to establish the possible channels of monetary policy transmission in the real sector, a pairwise granger causality test was undertaken between GDP and each monetary policy variable (Table 8).

A unidirectional causation from exchange rate to GDP was established while a bi-directional causality was established between private sector credit and GDP. These results suggest that the exchange rate and credit channels of monetary policy are essential in transmitting monetary policy impulses into the real economy in Zambia. These results are similar to the findings of Nampewo et.al (2013) who established the effectiveness of exchange rate and credit monetary policy channels in Uganda, a developing country like Zambia. Chileshe et.al (2015), using a different dataset, also established the effectiveness of the exchange rate channel Zambia, mostly due to Zambia’s dependence on copper exports.

In order to assess the adequacy of the model, LM residual correlation, VEC stability and normality tests were undertaken (Table 9).

The model passed serial correlation and stability tests, but it failed normality test. However, this is a common phenomenon for time series data due to its nature, as such, the employment of the Johansen cointegration approach becomes appropriate due its robustness (Diof, 2007).

This study investigated the impact of monetary policy on economic growth in Zambia over the period 2001q1 to 2021q2 based on the Johansen cointegration approach and granger causality test.

Monetary policy was proxied by average lending rate, 6 months deposit rate, exchange rate, private sector credit while inflation rate was used a control variable. The study results suggest that monetary policy has an impact on the Zambian economy via interest rates (lending and deposit), exchange rate and private sector credit. In addition, inflation had a significant influence on economic growth. This confirms the presence of the interest rates, exchange rate and credit channels of monetary policy transmission in Zambia. In terms of policy, there is need to enhance the effectiveness of interest rates channel and credit channel by improving the operations of the financial markets to enable them to transmit central bank actions in the real sector at the right time and appropriate magnitudes.