Assessing the Macroeconomic Effects of COVID-19 in an Empirical SFC Model for Denmark

Abstract

The COVID-19 pandemic has tossed the world into a state of uncertainty, not only from a health point of view but also from an economic perspective. Most countries have adopted a strategy of strict lockdowns, leading to a fall in production and a rise in unemployment. From an economic perspective, this lockdown has erupted in supply-and-demand shocks in the economy. In order to investigate the potential impact of these shocks, we develop a post-Keynesian empirical Stock-Flow Consistent (SFC) model for Denmark. The impact of these shocks is analyzed in three different scenarios. Overall, our simulation results indicate that temporary (1 year) shocks in the economy will lead to a fall in output but a quick recovery, whereas persistent shocks can lead to a full-blown economic crisis. We find that the impact of these shocks can be lowered by public intervention in the form of fiscal stimulus, which can reduce the impact of these shocks and also improve the speed of economic recovery. We also address the economic consequences of financing this stimulus, and whether such a policy response is feasible. Our main conclusion is that fiscal stimulus seems to be an effective and affordable policy option available to Denmark.

Keywords:

• COVID-19
• Denmark
• empirical stock-flow consistent models
• open economy

JEL CLASSIFICATIONS:

• B22
• E12
• E37
• E62

Acknowledgement

The authors would like to thank Finn Olesen, Mogens Ove Madsen, and the two anonymous reviewers for their constructive feedback.

Notes

1 This is a standard example of how supply-side shocks can trigger demand-side shocks.

2 A part of the financing of these expenditures comes from the EU Recovery Fund, where Denmark is expected to receive 11.6 bill. Krones. Sixty percent of these funds are allocated to initiatives related to the climate.

3 Discussing the structure of these models is beyond the scope of this article.

4 The analysis is based on the assumption that the COVID-19 is under control in the near future.

5 This particular feature sets post-Keynesian SFC models apart from mainstream macro models, where balance sheets structures are usually very simplified.

6 The decision to use net stocks data is due to lack of counterpart information. Using net stocks retains the basic accounting principles, ensuring that there are no leakages and that the sum of financial stocks across the sectors equals zero, as shown by the rows in Table 1.

7 Nominal variables in the model are denoted by capital letters whereas real variables are denoted by small letters.

8 All β-parameters are determined by estimation of the behavioral equations, as presented in Table 2.

9 This is in line with the model proposed in Zezza (2008).

10 In the estimation equations, the dummy variable for a specific year is denoted by D and a subscript representing the year.

11 On the revenue side, the government sector also receives a share of the gross operating surplus ($B_2^G$) from the production sector, which is exogenous in the model.

12 To estimate the equations, in most cases, we start our estimation by including two lags due to small sample. We then follow general-to-specific methodology and fit a parsimonious model. We also test for unit roots and account for any significant structural breaks in our estimations. We do not rule out the possibility that some of our estimates maybe biased due to various issues related to endogeneity or small-sample biases. Unfortunately, due to data limitations, we cannot investigate these issues in depth and therefore rely on the OLS estimation technique.

13 This comparison is only presented for certain key variables of interest in the Appendix in order to reserve space. However, more results can be made available upon request for interested readers.

14 However, we do not strictly bind ourselves by the aforementioned criteria and, in some cases when a variable shows a mean reverting tendency, we either keep its value constant or zero, depending on how far it has been oscillating from zero.

15 Note that the baseline assumes no crisis when the economy is growing at approximately 1% per year.

16 That is, consumption is exogenously reduced by 5% and investment is lowered by 5%.

17 $u_t$ is assumed to be a white-noise process.

18 Public consumption is chosen as an example of fiscal stimuli. In reality, public investment may also play a central role in this fiscal stimulus but, within the setting of this model, a change in public government and public investment is assumed to work through the same channels.

Additional information

Notes on contributors

Mikael Randrup Byrialsen

Mikael Randrup Byrialsen is an associate professor of economics at Aalborg University Business School, Aalborg University, Denmark. Mikael can be contacted via email: [email protected].

Hamid Raza

Hamid Raza is an associate professor of economics at Aalborg University Business School, Aalborg University, Denmark. Hamid can be contacted via email: [email protected].