Inflation and unemployment

Inflation and unemployment

Inflation and unemployment are some of the most challenges that affect many nations. According to Yolanda, 2017; Behera & Mishra, 2017, both inflation and unemployment affect economic activities such as export, investment, saving, economic growth, and poverty. For example, inflation declines the social welfare level. On the other hand, low inflation has the potential of lowering the economic growth rate, decreasing job opportunities, inclining poverty and gradually result in a recession. According to Behera and Mishra (2017), inflation has a positive effect on gross domestic products hence reducing the level of unemployment?  Minisi & Pantina, 2017 state that unemployment rates are influenced by other social factors such as insecurity, economic growth and so forth.

Unemployment continues to be a critical challenge in the economic growth and development of many countries. This has resulted in researchers developing the different frameworks of unemployment evolution and factors that affect (Helpman et al., 2010). Long fluctuation in unemployment is most common in shorter business cycles.  Short-term economic challenges such as unemployment and inflation have been noted to be the most macroeconomic problems that affect the economic growth of developing and developed countries today. The European Commission started to give more focus on inflation after the great global financial crisis of the year 2007/2008.

Inflation refers to the gradual increase in prices of goods and services (Hall, 2009). According to Arnold (2008), an inverse relationship exists between the inflation rate and unemployment rate.  For the last 20 years, inflation rates in Kenya have not been stable.  The level of inflation was high in March 2009 at 17.07% and low January 2011 at 3.93%. In 2019 the Kenya inflation rate was about 5%. The unemployment rate in Kenya between 2007 and 2009 was at 127.7% and thereafter rose to 40% in 2011, 46% in 2013, and 47% in 2014. The level of unemployment reduced to 41% in 2016.  In 2019 the unemployment rate had decreased to 39.1%.  Theoretical relationship between inflation and unemployment is evidence in the short run between the years 2009 to 2011. During this period the unemployment increases as inflation decrease.

According to a survey conducted in 2014, it was observed that only 2.4 million Kenyans are employed in the formal sector while 11.8 million Kenyans are employed in the informal sector (Irungu, 2016).   Informal sectors apply traditional approaches, therefore, resulting in inefficiency and hence causing unemployment. Unemployment theories such as Keynesian and classical theories give different views about unemployment which contradict each other. According to Keynes inflation cause an increase in production hence resulting in a decline in the level of unemployment.

In 1958 the concept of Philips curve was developed. The curve indicates the trade-off between the price of goods and services and the level of unemployment in the economy. The curve indicates that a relationship between inflation and unemployment exist. In the case of lower unemployment level then wages are high as the labor market strive to win the available workforce  (Forder, 2014). On the other side, a higher level of unemployment leads to the excess supply of workforce hence reducing market competition and thus the wage will increase at a slow rate. This means that higher inflation increases the rate of unemployment or vice versa. A dilemma, therefore, exists whether the inflation rate affects the level of unemployment. Phelps and Friedman argue that the government may not find it easy to trade higher inflation for lower unemployment in the long run. The concept of the Philip curve plays a key role in accessing the relationship between inflation and unemployment. The main objective of this research to access the relationship between inflation and unemployment as evidenced in Kenya.

Research method

The study employed an explanatory research design. The main case study for this study was Kenya as there were available data for analysis. The researcher used secondary data from the World Bank publication and the central bureau of statistics. The annual inflation rate of the country is generated from the Central Bank of Kenya and other published annual reports. On the other hand, the unemployment rates are generated from the annual statistical publication and the Central Bureau of Statistics. To be more specific, the researcher relied on the time series data from 1963 to 2017. The period for this analysis is the period under which the Kenya economy has been in the hands of the independent government. The researcher applied the purposive sampling method. This research captures inflation as a percentage change in consumer index and unemployment as the percentage unemployed workforce concerning the labor force in Kenya. In Kenya, unemployment is manifested by technology/ structural advancement. There is a high expectation that technological/structural unemployment will increase productivity. Demand-pull is the dominant manifestation of inflation in Kenya. This means that an increase in will encourage more production hence increasing the level of output. As a result of this it is expected that the increase in inflation would result to increase in production, thus decrease or reduction in the rate of unemployment assuming economic growth and other factors remain constant

The sample size for this study comprised of 20 years that is from 1997 to 2017. Time series data was applied as it involved the gathering of data recorded in the Kenya central bureau of statistics and World Bank publication in the last 20 years. The stationary of the data was tested with the help of SPSS, STAT, and E-views. Vector Autoregression and Granger causality were used to analyze the data. Before running the Vector Autoregression model and Granger Causality, the researcher first ran the lag length and stationary. The study used the following models for the test statistic and the stationary test (Brook)

ΔYt = ϕYt – 1 + ϕt;                            (1)

Test statistic =ϕ/S.Eϕ                      (2)

The lag length test was then done. Different approaches are used in selecting the optimal lag length, such as Schwarz Information Criterion, Akaike Information Criterion, Final Prediction Error and Likelihood Ratio (Rosadi 2012). For this study, the researcher applied the Akaike Information Criterion (AIC). In this case, the optimal lag is obtained when the AIC value is minimum.

Model Specification

Philip’s curve was used to explain the causal relationship between inflation and unemployment in the economy. According to Philip’s curve theory increase in output results in a decrease in the unemployment rate. The model shown below was used for this research.

ΔLog𝑈𝑁𝑅_t−i = β_O + β₁ ΔLog(𝐼𝑁𝐹𝐿)_t−i + ε_i………………………………………………3.1

Where

Β₀:  The unemployment level that does not depend on the factors under consideration.

β1: The inflation rate coefficients used in determining how changes in the rate of unemployment as a result of changes in the respective factors by one unit.

εi: it represent other factors other than inflation that influence unemployment in Kenya

ΔLog INFL: These represent Logarithms of inflation rate

ΔLogUNR: These represent Logarithms of the unemployment rate

The researcher also applied linear regression to explain the relationship between inflation and unemployment in Kenya.  The casual research design was employed to capture the effect of inflation on the unemployment rate. The research adopted Okun’s law as a theoretical basis to explain the relationship between inflation and unemployment rate in Kenya. Okun’s law state a deleterious relationship exists between inflation and unemployment. The study employed the Aliyu (2012) model;

Y = β_O + β₁ U_t + β₂ U_t2+  ε_i………………………………………………3.2

Where Y is the inflation, U is the unemployment rate

The model is can be modified as shown below

Unemployment = β_O + β₁ inflation + ε_i………………………………………………3.3

ΔLog𝑈𝑁𝑅_t−i = β_O + β₁ ΔLog(𝐼𝑁𝐹𝐿)_t−i + ε_i………………………………………………3.4

Granger Causality Test

To examine the causality between the variables the researcher employed Granger.

∆LGDP = α + β 1 ∆ 1 𝑖=1 𝐿𝐺𝐷𝑃𝑡−𝑖 + 𝛽2𝐿𝑢𝑛𝑒𝑚𝑝𝑡−𝑖 1 𝑖=1 + ѱECMt−1 + εt … … … … . .17

3.5

From the equation over the null hypothesis

HO: inflation rate does not granger cause unemployment

HO: unemployment does not Granger cause inflation rate

4. Results and Discussions

4.1 Stationary Tests

The stationary statistics of different variables in this research are shown in table 1. The research findings indicate that unemployment ADF statistics of 2.493. This shows a greater level compared to the Mackinnon critical value of -1.96 at a 5% significance level. This means that the researcher should accept the null hypothesis, hence implying that the rate of unemployment isn’t stationary in the level form and thus there is a need for more differencing. Based on the research findings the ADF statistic at level form is -1.554 which is more compared to Mackinnon’s critical value of -1.96. This implies that the rate of inflation is not stationary at the level form thus a stationary difference is observed between inflation rate and unemployment and therefore no need for further differencing.

4.2 Lag selection

Table 2: Lags for Long Run and Short-Run Model

The table above shows the findings of selecting the appropriate lag for the short-run and long-run model. The maximum number of lags for this study is from lag 0 to lag 4. Doing a comparison of the AIC and the SBIC results in the long-run model, the findings reveal that the minimum number of AIC lag is -1.62466 while the value of lag SBIC is -1.37573. The study implication hence indicates that the long run lag was more than 1. The minimum number of lag according to AIC and SBIC is lag 2 for short term model. The most appropriate for short term model in this study is lag length 2

the Co-integration between the expletory and explained variables is shown in table three. The research findings indicate that ADF statistics from lag 0 to lag 2 are less compared to the Mackinnon critical values at a 5% significance level.  Table 3 shows that at lag 2 the ADF statistic is -3.210 which is less than -1.96, also at lag 1 and 0, the lag is -3.372 and -7.434 respectively which are similarly less than Mackinnon critical value. The null hypothesis is therefore rejected, thus meaning the co-integration between there is co-integration between the dependent and independent variables under consideration.

Normality Test of Residuals

The researcher also did the normality test to test the distribution of the data. The Jarque-Bera Chi-square value obtained was 3.676 at 2-degree freedom of 2 while the probability of 0.15918. was obtained. The probability value is more than the significant value of 5% thus meaning that the null hypothesis should be rejected.  This indicates that the residual is distributed normally. Concerning the data skewness, a test value of 0.88719 was observed. The researcher also observed the profitability value of 0.089535 and the chi-square value of 2.886, degree of freedom of 1. The skewness value was greater than 0.005 thus the null hypothesis was rejected, hence the disturbance is systematically as well as normally distributed.  On the other hand, the researcher observed kurtosis of 3.928 and 0.789 at the degree of freedom of 1 while the probability statistic of 0.3743 was observed. The p-value of 0.3743 exceed 0.05, thus the researcher rejected the null hypothesis, indicating a normally distributed residuals.

Regression Analysis

Long Run Regression Model

𝑌 = −0.195541 − 0.003836𝑋1 + 0.1233𝑑1

The researcher used the long-run regression model to assess the relationship between the two variables. The research findings indicate that coefficient, standard error, t-value, and the probability was obtained as follows – 0.1955, as 0.8073, -0.242, and 0.0412 respectively.  The coefficient correlation between the inflation rate and unemployment rate was found to be -0.0038, on the other hand, the standard error, t-value, and t-probability was found to be 0.0336, -0.114, and 0.0310, respectively. The researcher found the overall model R-squired was 0.878 (87.8%). The statistics Durbin Watson was obtained to be 2.14 hence no problem of autocorrelation was found among error terms. A p-value of 0.000 F statistics were found to be 64.2.  Compared to the significance p-value is less than 0.05 hence the model is statistically significant.

𝑌 = 0.00328844 − 0.0158615𝑋1 + 𝜀𝑖

The short-run regression is shown in the table above. The coefficient of the constant 0.003288 was obtained. The researcher also found the standard error of constant to be 0087 while t-value and t-probability were found to be 0.377, and 0.0317 respectively. The coefficient correlation between inflation was found at 0.0158. The research found standard error, t-value, and t-probability to be 0.0115, -1.38, and 0.0211 respectively. The R-squire for the short-run model was found to be 0.8712, implying that the explanatory variable was used to explain 87.12% variation in the unemployment rate. No autocorrelation problem among error terms was found since the Durbin Watson statistic was 2.26. A p-value of 0.000 the F statistic was 16.86. This indicates that the overall model was statistically significant since the p-value is less than 0.05.

Conclusion

Based on research finding a significant relationship was found between inflation rate and unemployment both in the long run and short run where the coefficient relationship was  -0.00383 and 0.01586 respectively. A negative relationship was therefore found both in the long run and short run.  This indicates that as the inflation rate increases the unemployment rate decrease and vice versa. Based on the research findings a negative significant effect of the inflation rate on the rate of unemployment in Kenya was evidence in the short term. In the long run, a positive significant effect was observed. The null hypothesis of the study should, therefore, be rejected, implying that a significant relationship between inflation rate and level of unemployment in Kenya is evidenced in two periods considered. A high inflation rate encourages investors to invest in the private sector due to high profits hence increase the workforce required to produce returns and hence reducing the level of unemployment in the country.

The main of this study was to evaluate the relationship between inflation and unemployment both in the long run and in the short run.  An increase in the inflation rate means that the level of unemployment in the country will decrease significantly. Consequence the study found out that the Philip curve concept does not apply in the Kenyan economy and other developing countries. The null hypothesis was therefore rejected. In conclusion, high inflation will increase the price of goods and services hence increasing the supply of goods and services to the economy. As a result, more labor will be required to produce goods and services hence reducing the level of unemployment in the countries.