Table 3.

Quartic specification estimation results

Denmark
Scaling factor (a)	0.00003	0.00004	0.0000483	0.00009	1 (original data)	10	100
$\widehat{{\beta }_1}$	−0.2568**	0.3390**	1.4673***	15.0919**	11762.7366	21399.8491	35237.5345
t-value	(−2.0577)	(2.1484)	(8.7833)	(2.2716)	(1.3781)	(1.3682)	(1.3615)
$\widehat{{\beta }_2}$	−0.5644	−1.9360***	−4.2326***	−19.6682*	−1691.6130	−2521.2378	−3515.8986
t-value	(−1.4283)	(−5.7221)	(−3.5257)	(−1.7923)	(−1.3599)	(−1.3536)	(−1.3494)
$\widehat{{\beta }_3}$	0.0968	3.0818**	5.0382*	11.4959	108.1548	132.0462	155.9376
t-value	(0.0931)	(2.1308)	(1.7544)	(1.4912)	(1.3427)	(1.3398)	(1.3378)
$\widehat{{\beta }_4}$	−2.5940	−2.5940	−2.5940	−2.5940	−2.5940	−2.5940	−2.5940
t-value	(−1.3267)	(−1.3267)	(−1.3267)	(−1.3267)	(−1.3267)	(−1.3267)	(−1.3267)
$\widehat{\eta }$	−0.2726	−0.2726	−0.2726	−0.2726	−0.2726	−0.2726	−0.2726
t-value	(−0.0299)	(−0.0299)	(−0.0299)	(−0.0299)	(−0.0299)	(−0.0299)	(−0.0299)
Sweden
Scaling factor (a)	0.00003	0.00004	0.0000483	0.00007	1 (original data)	10	100
$\widehat{{\beta }_1}$	−1.2502***	−1.1679***	0.1306	3.8387***	−8813.4848	−16778.0579	−28482.9102
t-value	(−7.6954)	(−9.5781)	(0.7235)	(2.9922)	(−0.9615)	(−0.9912)	(−1.0116)
$\widehat{{\beta }_2}$	2.5221***	−2.4026***	−4.3097***	−5.0085	1375.3762	2109.5873	2999.7589
t-value	(4.5660)	(−4.1894)	(−9.5357)	(−1.2014)	(1.0153)	(1.0348)	(1.0482)
$\widehat{{\beta }_3}$	7.1166***	4.2958***	2.4470	−1.1914	−94.9991	−117.5767	−140.1543
t-value	(3.0627)	(6.5917)	(1.2162)	(−0.2289)	(−1.0682)	(−1.0777)	(−1.0843)
$\widehat{{\beta }_4}$	2.4513	2.4513	2.4513	2.4513	2.4513	2.4513	2.4513
t-value	(1.1199)	(1.1199)	(1.1199)	(1.1199)	(1.1199)	(1.1199)	(1.1199)
$\widehat{\eta }$	−1.5508	−1.5508	−1.5508	−1.5508	−1.5508	−1.5508	−1.5508
t-value	(−0.2201)	(−0.2201)	(−0.2201)	(−0.2201)	(−0.2201)	(−0.2201)	(−0.2201)

Source: Data from Jaforullah and King (2017)

Notes: The values of the scaling factors are chosen to show the existence of coefficients with the opposite signs and with significant and insignificant values (according to t-values). The estimated elasticities $\widehat{(\eta })$ are calculated at mean values as discussed in Gujarati and Porter (2009), inter alia