The analysis of dat a of t he st udies is present ed in t his sect ion. It is divided into subsect ion sect ion which out lined t he st at istical t rends of t he variables of t he st udies, t he unit root t est , t he coint egrat ion bound t est , t he long run and t he short run ARDL error correct ion models and t he Granger causalit y.
Unemployment trend in China
The annual unemployment rat e of China is wit hin t he range of 3.76% and 4.89%
in 2007 and 1991 respect ively. This shows t hat out of 100 people who are act ively searching and willing t o work t o receive some income only 4 t o 5 people are unable t o secure employment . This really shows that unemployment rat e in China which falls in t he range of 3.76% t o 4.89% is not high but moderat e and it really shows how t he economy of China is able to cont ain most of t he available human capit al or labour force.
Alt hough, t here is some linear addit ional upsurge in t he rat e of unemployment from 2010 where t he unemployment rat e was 4.12% t o t he current year 2018 where t he unemployment rat e is 4.71%. However, it can be perceived t hat the linear increase in t he unemployment rat e in each year from 2010 t o dat e is reasonable.
The t rend of t he unemployment rat e in China from 1991-2018 is int eresting
rat es of changes in t he unemployment rat e for a period of 27 years. From t he graph below, China experienced it s highest unemployment rat e in 1991 with an unemployment rat e of 4.89% t hus among 100 act ive people who were and willing to work didn’t get employed. From 2010 t he unemployment rate of China has cont inued t o rise from a rat e of 4.2% t o 4.7% which is it s current
Economic growth trend in China
China wit hin t he last four decade, t hat is, 1979 t o 2017 has t ransformed its economy from t he st at e is a poor developing nat ion t o a st at e of being one of t he most power developed market s in t he world. The average annual real product ivity out put of China has grown t o an approximat e of 10% which according t o t he World Bank makes t hem one t he fast er grown, expanded and sust ained economy in t he world. This change in t ransit ion has cause China t o eradicat e about 800 million persons out of povert y (Morrison, 2018).
According t o t he annual t ime series dat a available at World Development Indicat or, t he annual economic growt h (GPD) rat e of China is wit hin t he range of 14.23% and 6.7%which represent s t he highest rat e economic growt h (GPD) and t he lowest economic growt h (GPD) rat e respect ively form 1991-2018. This shows t hat from 1991 t o 2018, a variat ion in economic progress (GPD) rat e is wit hin t he highest 14.23% and lowest at 6.7%.
Though t he rat e of economic growt h (GPD) is changing, in 1992 China had experienced it s highest increase in economic growt h (GPD) rat e wit h a change of 4.93%. This shows t hat there was a lot of labour out put which led to an increase in product ion. Nonet heless, bet ween 2003 and 2004 China experience it s lowest change of economic growt h (GPD) rat e wit h a rat e of 0.07%. Which indicat e t here was lit t le out put which led t o const ant or lit tle addit ion t o 2003 economic growt h (GPD).
The t rend of economic growt h (GDP) rat e in China from 1991-2018 is
were many dissimilar rat es of changes in t he economic growt h rat e (GDP) t he period of 27 years. From t he graph below, in 1991 China experienced 9.3%
economic growt h (GDP) rat e and t hen t he economic growt h (GDP) increase t o 14.22% in 1992 which represent 4.92% which indicat e China highest increase of economic growt h (GDP) rat e up t o dat e but represent t he second highest economic growt h (GDP) rat e.
Economic growt h (GDP) begun t o cont inuously decrease from 1993 wit h t he rat e of 13.86% t o 2001wit h rate of 8.34% and t hen begun t o gain some st rength from 2002 wit h t he rat e of 9.13% t o 2007 wit h t he rat e of 14.23 which is t he highest economic growt h (GDP) rat e up t o dat e. Since 2007, economic growth (GDP) has diminished from 2008 wit h a rat e of 9.65% which represent t he
Analysis of the Unit Root Test
In ot her t o est imat e non-spurious regression result s, I first est imat ed t he st at ionarity of t he variables by employing t he Augment ed Dickey-Fuller (ADF) and t he Phillips and Perron (PP) unit root t est and t he following result s were obt ained.
Table 4.1 below shows t he result s obt ained for t he St ationarity of variables.
variables Level First Difference ADF PP ADF PP
Constant Constant Constant Constant
Inuemp 0.512 0.512 0.003***
0.003***
Ingdp 0.354 0.354 0.028***
0.028***
We t est the null hypot hesis of t he series being non-st at ionary or has a unit root against t he alt ernative hypot hesis of t he exist ence of st at ionarity. Mackinnon (1996) crit ical values was used in reject ing t he null hypot hesis by bot h ADF and PP t est , ***,**,* signifies t he reject ion of t he null hypot hesis of t he exist ence of a unit root at 1%,5%, and 10% significance levels respect ively.
It can be ascended from t able 4.1 t hat, t ests by ADF and PP clearly shows that
st at ionary at level I(0) since ADF t -st atist ics of -1.304 for economic growt h (GDP) and -2.447 for unemployment were less t han t he crit ical values of 1%, 5%, and 10% respect ively and PP t st atistics of 1.304 for economic growt h (GDP) and -2.519 for unemployment were also fewer t han t he crit ical values of 1%, 5%, and 10% respect ively.
Nonet heless, t he variables (economic growt h (GDP) and unemployment ) of t he st udy was st at ionary at first level I(1) since ADF t -st atist ics of -5.707 for economic growt h (GDP) and -6.076 for unemployment were great er t han the crit ical values of 1%, 5%, and 10% respect ively and PP t -st atistics of -5.706 for economic growt h (GDP) and -6.284 for unemployment were also more t han t he crit ical values of 1%, 5%, and 10% respect ively.
Therefore we can conclude t hat t he variables (economic growt h (GDP) and unemployment ) of t he st udy were not st at ionary at level but t hey were all st at ionary at first difference which means t hat the dat a of variables of t he st udy are good t o use in t his st udy.
Bounds test
Bound t est expedit es t o check t he long run relat ionship bet ween t he variables (economic growt h (GDP) and unemployment ) of st udy. More so, it was expedient t o t he posit ion of t he F-st at istics wit hin t he crit ical value bound of significance.
Table 4.2 result s of bounds t est
Test ing for t he exist ence of a long run relat ionship among t he variables in t he ARDL
K 95% lower bound 95% upper bound 90% lower bound 90%
upper bound
1 4.49 5.73 4.04 4.78
Model Calculat ed F-st at istics Inference
Ingdp(Inuemp) 6.499611** Coint egrat ion
From t able 4.2 above, t he calculat ed F-St at istics (6.499611) is bigger t han both t he 95% and 90% upper bound confidence level of 5.73 and 4.78 respect ively form t he equat ion. Therefore, from t his finding, it is post ulat ed t hat t here is coint egrat ion among t he dependent variable (unemployment ) and t he independent variables (economic growt h (GDP)). This is in confirmat ion with t he st udies of Soylu, Cakmak and Okur (2017), Banda, Ngirande and Hogwe (2016) and Mosikari (2013).
Results of the Long Run Unemployment
The long-run relat ionship bet ween t he dependent variable (unemployment) and independent variable (economic growt h (GDP)) was est imat ed by t he ADRL. The long-run elast icit y is represent ed by t he coefficient s of t he dependent variable (economic growt h (GDP)).
Table 4.3 Est imated long run inflat ion model
Dependent variable: Inuemp
Regressors Autoregressive Distributed Lag Model Long run Elasticity t-statistics
Ingdp -0.320 ( -3.878) **
C 2.173 (12.045)
***, **, * denot es significance level at 1%, 5% and 10% respect ively. Values in parent hesis are t -st atistics. ARDL (1,4) was based on t he Swchwarz Bayesian crit erion
From t able 4.3 above, t he long run elast icit y coefficient of t he dependent variable (economic growt h (GDP)) is adverse and st at istically significant at t he 10% error level. Wit h respect t o t he coefficient , a one per cent upsurge in economic growt h will cause a 0.32% decrease in t he unemployment rat e. This confirms t he assert ions by Hua-chu (2008), Li and Liu (2012) and Karabulut and Gokhan (2010) t hat , a long run relat ionship exit bet ween economic growth (GDP) and unemployment . Accordingly, t he null hypot hesis t hat t here is no long-run relat ionship bet ween unemployment and economic growt h in China is reject ed.
Results of Short Run Error Correction Model
The error correct ion model t ries t o provide a remedy by t he reconciliat ion of short -run behaviour of a variable wit h t he long run behaviour. It becomes mandat ory t o est imat e t he short -run error correct ion when t here is a long run relat ionship among t he variables (economic growt h (GDP) and unemployment ). Thus it measures t he dynamics of t he short run model capt ured by t he ECM and t he coefficient help wit h t he speed wit h which t he model adjust t o an equilibrium whenever t here is a shock. This model is represent ed by t he first difference as seen in t able 4.4
Table 4.4 Est imated short run error correct ion model using t he ARDL Approach
Dependent variable: Inuemp
Regressors Autoregressive Distributed Lag Model Short Run Elasticity t-statistics
Ingdp -0.333 (-16.640) ***
C 0.484 (3.167)
***, **, * denot es significance level at 1%, 5% and 10% respect ively. Values in parent hesis are t -st at istics. ARDL (1,4) was based on t he Swchwarz Bayesian crit erion
From t able 4.4 above, t he short run elast icit y coefficient s of economic growth
above, economic growt h (GDP) has a negat ive short -run relat ionship with unemployment and a one per cent increase in economic growt h will cause a 0.33% decrease in t he unemployment rat e. This confirms t he assert ions by Makaringe & Khobai (2018), Drit sakis & St amatiou (2016) and Lam (2014) t hat, a short run relat ionship exit bet ween economic growt h (GDP) and unemployment . Accordingly, t he null hypot hesis t hat t here is no short -run relat ionship bet ween unemployment and economic growt h in China is reject ed.
Results of Granger Causality
The unit root by ADF and PP clearly shows t hat economic growt h rat e (GDP) and unemployment rat e are st at ionary at first difference I(1). Therefore, I employed t he first log difference bet ween t he variables in conduct ing t he Granger Causalit y t est .
Table 4.5: Result s of t he Granger causalit y t est
Null hypothesis F-statistics Prob.
lnuemp does not Granger cause lngdp 1.17741 0.3460
lngdp does not Granger cause lnuemp 1.81217 0.1811
From t he t able 4.5 above, unemployment does not Granger cause economic growt h (GDP) in t he long run and short run because t he F-st at istics of 1.177 is insignificant and also economic growt h (GDP) does not Granger cause unemployment in t he long run and short run because t he F-st at istics of 1.812 is insignificant . This confirms t he assert ions by Mosikari (2013) t hat , t here is no causal relat ionship bet ween economic growt h (GDP) and unemployment .
Though t here is a negat ive relat ionship bet ween economic growt h (GDP) and unemployment in t he long run and t he short run, t hey do not have a causal relat ionship and t hat other fact ors might cause t heir relat ionship in bot h long run and short run and not necessary t he variables t hemselves.
CONCLUSION AND RECOMMENDATIONS