Monetary policy and unemployment in Croatia

The extended abstract of this article was originally presented at The 7th International Scientific Conference ‘European Union Future Perspectives: Innovation, Entrepreneurship and Economic Policy’, Pula (Croatia), May 21–23, 2015.

Figures & data

Graph 1. Unemployment, consumer prices, money M1 and the real effective exchange rate of Croatian kuna. Source: Authors’ calculation.

Table 1. Unit root tests.

Variable and test	Level		First difference
	Constant	Constant and trend	Constant	Constant and trend
ADF test	t-stat.
LUNEMP	-2.899280	-3.300217	-2.172292	-2.180518
LCPI	-1.589609	-2.469990	-3.147214	-3.430400
LM1	-1.671564	-0.670032	-5.859515	-6.111687
LREER	-1.399625	-1.926888	-4.227311	-4.195562
PP test	Adj. t-stat.
LUNEMP	-1.679008	-1.805247	-2.348963	-2.354420
LCPI	-1.926064	-1.723374	-6.160648	-6.429134
LM1	-1.435638	-1.050568	-5.887578	-6.057253
LREER	-1.077275	-1.707173	-4.225847	-4.192105
KPSS test	LM-stat.
LUNEMP	0.235018	0.147416	0.123902	0.126338
LCPI	1.058710	0.097381	0.266383	0.073131
LM1	0.935932	0.239876	0.252457	0.092824
LREER	0.719631	0.176629	0.149812	0.151273

Source: Authors’ calculation.

Note: In the analysis, EViews (IHS Global Inc., 2015), Gretl (Cottrell & Lucchetti, 2015) and Microfit (Pesaran & Pesaran, 2009) econometric software were used. ‘L’ indicates logarithm of the variable. For the implementation of ADF test the Schwarz information criterion has been implemented. ADF test critical values (MacKinnon, 1996); constant: 1% level (-3.53), 5% level (-2.91), 10% level (-2.59); constant and trend: 1% level (-4.10), 5% level (-3.48), 10% level (-3.17). PP test critical values (MacKinnon, 1996); constant: 1% level (-3.53), 5% level (-2.91), 10% level (-2.59); constant and trend: 1% level (-4.10), 5% level (-3.48), 10% level (-3.17). KPSS asymptotic critical values (Kwiatkowski et.al., 1992); constant: 1% level (0.739), 5% level (0.463), 10% level (0.347); constant and trend: 1% level (0.216), 5% level (0.146), 10% level (0.119).

Table 2. Testing for existence of a level relationship among the variables in the ARDL model.

F-statistic	95% Lower Bound	95% Upper Bound	90% Lower Bound	90% Upper Bound
6.9387	4.2937	5.3474	3.6399	4.6332
W-statistic	95% Lower Bound	95% Upper Bound	90% Lower Bound	90% Upper Bound
27.7547	17.1750	21.3896	14.5596	18.5327

Source: Research results.

Note: The critical value bounds are computed by stochastic simulations using 20.000 replications.

Table 3. Diagnostic tests of the ARDL (2, 1, 0, 0) unemployment equation.

Test Statistics	LM Version	F Version
Serial Correlation: Lagrange multiplier test of residual serial correlation	CHSQ(4) = 2.2020, Prob. = 0.699	F(4,32) = 0.46177, Prob. = 0.763
Functional Form: Ramsey’s RESET test using the square of the fitted values	CHSQ(1) = 4.2336, Prob. = 0.040	F(1,35) = 3.8902, Prob. = 0.054
Normality: Based on a test of skewness and kurtosis of residuals	CHSQ(2) = 0.37409, Prob. = 0.829	-
Heteroscedasticity: Based on the regression of squared residuals on squared fitted values	CHSQ(1) = 0.49587, Prob. = 0.481	F(1,61) = 0.48394, Prob. = 0.489

Source: Research results.

Table 4. Estimated long-run coefficients of the ARDL (2, 1, 0, 0) unemployment equation.

Dependent variable: LUNEMP
	Coefficient	Std. Error	T-Ratio	Prob.
LCPI	4.7702	1.8629	2.5607	0.013
LM1	-0.18340	0.23650	-0.77545	0.441
LREER	2.4514	0.99624	2.4607	0.017
INPT	-27.0082	9.8698	-2.7365	0.008
TREND	-0.026995	0.012028	-2.2443	0.029

Source: Research results.

Note: ‘L’ indicates logarithm of the variable.

Table 5. Error correction representation of the ARDL (2, 1, 0, 0) unemployment equation.

Dependent variable: dLUNEMP
	Coefficient	Std. Error	T-Ratio	Prob.
dLUNEMP_1	0.68620	0.10266	6.6841	0.000
dLCPI	-0.37643	0.35058	-1.0737	0.288
dLM1	-0.018313	0.020938	-0.87465	0.386
dLREER	0.24479	0.14157	1.7291	0.089
dTREND	-0.0026956	0.0011473	-2.3496	0.022
ecm(-1)	-0.099857	0.026568	-3.7586	0.000
R-Squared = 0.78282		R-Bar-Squared = 0.75518
S.E. of Regression = 0.013850		F-Stat. F(6,56) = 33.0407, Prob. = 0.000
Mean of Dependent Variable = 0.0011893		S.D. of Dependent Variable = 0.027991
Residual Sum of Squares = 0.010550		Equation Log-likelihood = 184.4922
Akaike Info. Criterion = 176.4922		Schwarz Bayesian Criterion = 167.9197
DW-statistic = 1.9293

Source: Research results.

Note: ‘d’ indicates first difference, while ‘L’ indicates logarithm of the variable.

IHS Global Inc. (2015). EViews (8.1) [Computer software]. Irvine, CA. Retrieved from http://www.eviews.com/

 Google Scholar

Cottrell, A., & Lucchetti, R. Gretl. (2015). Gretl (1.10.1) [Computer software]. Retrieved from http://gretl.sourceforge.net/

 Google Scholar

Pesaran, B., & Pesaran, H. M.. (2009). Microfit (5.1) [Computer software]. Oxford University Press.

 Google Scholar

MacKinnon, J. G. (1996). Numerical distribution functions for unit root and cointegration tests. Journal of Applied Econometrics, 11, 601–618. 10.1002/(ISSN)1099-1255

 Web of Science ®Google Scholar