On the Link Between Policy Uncertainty and Domestic Production in G7 Countries: An Asymmetry Analysis

Abstract

Previous studies have assessed the impact of policy uncertainty on consumption and investment in G7 countries. In this study, we assess its impact on domestic output in the same countries. Furthermore, we argue that its impact could be asymmetric, implying that increased uncertainty affects domestic output at a different rate than decreased uncertainty. Unlike consumption and investment, we find the unanimous outcome in all G7 countries that increased uncertainty hurts domestic output and decreased uncertainty boosts it, though significant long-run asymmetric evidence was found only in the cases of Canada, Japan, and the U.S. Thus, any policy aimed at reducing uncertainty will be growth-enhancing.

KEYWORDS:

• Policy uncertainty
• asymmetry
• domestic output
• G7

JEL CLASSIFICATION:

• f40

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Al-Thageb and Algharabali (2019) is a review article on policy uncertainty.

2 Miles, Hahm and Steigerwald, Banks, Blundell, and Brugiavini, Guariglia and Rossit, Menegatti, Menegatti, and Roberto and Riveiro are some old studies that have assessed impact of income uncertainty on consumption and saving decisions.

3 Bahmani-Oskooee and Ghodsi (2018) have demonstrated that the t-test here is exactly the same as the t-test that is used to test significance of lagged error-correction term in the Engle and Granger (1987) setting. Note also that estimate of θ₀ must also be negative if variables are to converge.

4 Note that Pesaran et al.’s (2001) critical values are for large samples. However, in light of their footnote 25, we can still use their critical values since we have more than 100 observations in each case. Furthermore, by using the AF test, we made sure that there are no I(2) variable.

5 See Shin et al. (2014, p. 291). Furthermore, specifications like Equation (4) corrects for weak exogeneity of explanatory variables and choice of optimum lags will make the residuals autocorrelation free (Shin et al., 2014, p. 290).

6 For some other application of these methods in this journal, see Bahmani-Oskooee and Maki-Nayeri (2018b) and Bahmani-Oskooee and Arize (2019).

7 By meaningful we mean cointegration is supported either by the F- or the t-test, reported in Panel C.

8 Other diagnostics are similar to those of the linear model and need no repeat.

9 All diagnostics are similar to those of Canada and need no repeat.

10 For some other applications of these methods, see Apergis and Miller (2006), Halicioglu (2007), Nusair (2012, 2017), Gogas and Pragidis (2015), Durmaz (2015), Aftab et al. (2017), Arize et al. (2017), and Hajilee and Niroomand (2019).

Additional information

Notes on contributors

Mohsen Bahmani-Oskooee

Mohsen Bahmani-Oskooee is a University Distinguished Professor and Patricia and Harvey Wilmeth Professor of Economics at the University of Wisconsin-Milwaukee. He is also the director of the Center for Research on International Economics. He has published extensively in international finance and open economy macroeconomics including articles in the Journal of Political Economy, Review of Economics and Statistics, Journal of Development Economics, Economic Development and Cultural Change, etc.

Amirhossein Mohammadian

Amirhossein Mohammadian is an Assistant Professor of Economics at the College of Business and Economics at the University of Hawaii at Hilo. His research interests are international finance, open economy macroeconomics, and applied econometrics. His articles have appeared in such journals as Australian Economic Papers, International Review of Applied Economics, Emerging Markets Finance and trade, Empirica, etc.