David Baqaee and Ben Moll on What if Germany is cut off from Russian oil and gas?

Video & Timestamps

On Thursday, April 7, David Baqaee and Ben Moll joined Markus’ Academy for a lecture on What if Germany is cut off from Russian oil and gas? Baqaee is an assistant professor of economics at UCLA, a research affiliate of the CEPR, and a Faculty Research Fellow at the NBER. Moll is a professor of economics at London School of Economics and Political Science.

Watch the full presentation below. You can also watch all Markus’ Academy webinars on the Markus’ Academy YouTube channel.

Timestamps:

[0:00] Introductory remarks

[16:16] Goals of the research

[20:37] German energy shock

[38:36] Keys to basic macro modeling

[50:16] Richer models with supply chains and international trade

[1:04:46] Order of Magnitudes calculation

[1:15:41] Other missing effects?

[1:28:06] Policy recommendations

Executive Summary

• [0:00] Introductory remarks. Coal, oil, and natural gas sanctions are very different. Coal and oil are part of the world market, whereas gas is more local; hence, the West can replace Russian coal and oil from other suppliers, but Russia can also sell their energy to other customers, undermining the effectiveness of sanctions. For natural gas the situation is different. In terms of timing, “blitz sanctions” are imposed cold turkey and are more effective though costly, while sustained sanctions are long lasting preparing for a long confrontation. The costs depend crucially on how easily one can substitute natural gas and gas dependent products with alternatives. Estimating the elasticity of substitutability is tricky, as there are (non-local) non-linearities – small shocks are easier to deal with compared to large shocks. For an example, see Jim Hamilton’s webinar for the 1973 OPEC shock. Furthermore, production chains further complicate matters and many macro-models do not take physical transportation constraints fully into account. Finally, financial frictions can further amplify the effects.
• [16:16] Goals of the research: analyzing the economic consequences for Germany of some supply shock, caused by an embargo or stop of deliveries from Russia. Would this lead to a small GDP decline, a depression, or something in between? The paper estimates that GDP decline will be between 0.5% and 3%, making an import stop less severe than a Covid recession, where Germany was able to provide insurance and socialize costs. 

• [20:37] German energy shock. German energy is largely sourced from Russia– 34% of oil, 55% of gas, and 25% of coal– the paper focuses primarily on gas, due to the challenges in global transportation. Substitution of gas is only to some degree possible, perhaps from Norway or the Netherlands. It would also be possible to use lignite, hard coal, or nuclear technology, which could mean that the gas shock would essentially be 30% instead of 55%; this equates to an 8% decrease in total energy from gas, oil, and coal. Consumption of gas, oil, and coal comprise 4% of Gross National Expenditure (GNE), the imports make up 2.5% of GNE, and consumption of gas is 1.2% of GNE. Yet, amplification could make these effects quite large. Hardest hit sectors could see a difference comparable to the Covid crisis in Gross Value Added, wages, and employees. The paper argues that the distribution of costs should be relatively equal across the income distribution.

• [38:36] Substitutability is key to basic macro modeling. Modeling that assumes perfect or no substitutability, i.e. GDP drop will be proportional with gas decline, is ill-founded. The simplest model is an aggregate CES production function, which has output as a function of gas, capital, labor, and more; the keys here are the elasticity of substitution and the gas share. Elasticities of substitution are very time dependent (relevant horizon of next winter), and understanding the macro effects of cross-production processes is necessary. Supply chains and substitution via imports are other considerations. Output losses for different elasticities will be very different.

• [50:16] Richer models with supply chains and international trade. The richer analysis based on Baqaee & Fahri that takes supply chains into account and distinguishes between German real consumption and real production. However, it assumes perfect substitutability within each subsector and abstracts away from within-sector transportation frictions of gas. Changes in real consumption can be approximated (to a first and second order) through changes in imports, exports expenditures, and employment. Non-local effects are not included. 

• [1:04:46] Order of Magnitudes calculation. By comparing to the oil shock in the 1970s, we can estimate how much the energy share changes. Assuming that expenditure share quadrupled, as they did in the 1970s, we would see a 1% loss in GNE. All models suggest a GNE loss under 2.3%, which would suggest a per capita per year shift of 400-900 euros. However, these calculations may miss some business cycle amplification effects, such as the real and nominal frictions contracting aggregate demand. To deal with this, pessimistic calibration was used, but the numbers should be lower than 3%, if the estimates are off.
• [1:15:41] Other missing effects? One thing left out was Keynesian aggregate demand amplification; Bayer et al. try to focus on heterogeneous households and use Keynesian demand functions with a TFP drop of 2.2 % (from this paper) and 3% capital obsolescence due to a depreciation shock. This analysis assumes that inflation expectations remain well anchored. The German Council of Economic Experts had a variety of scenarios, and all GDP deduction estimates were somewhere under 2.5%. They also clarified that a full import stop would mean a 3-5% GDP loss in the first year (more information here). Krebs (2022) believed that there should have been separate elasticities of substitution for different industries. Others claimed that the gas shock would hit harder, perhaps closer to 55% than 30%. 

• [1:28:06] Policy recommendations. Need people to substitute, and prevent the shock from falling entirely on the industry or household. Price adjustments are assumed to lead to appropriate substitution. Monetary policy will have to be ready to raise interest rates to control inflation, fiscal policy should not tax subsidies on energy. Using policies applied in Covid will help socialize losses, and substantial inflation effects might require adjustment of tax/transfer schedules.