Investment incentive reduced by climate damages can be restored by optimal policy

Abstract

Increasing greenhouse gas emissions are likely to impact not only natural systems but economies worldwide. If these impacts alter future economic development, the financial losses will be significantly higher than the mere direct damages. So far, potentially aggravating investment responses were considered negligible. Here we consistently incorporate an empirically derived temperature-growth relation into the simple integrated assessment model DICE. In this framework we show that, if in the next eight decades varying temperatures impact economic growth as has been observed in the past three decades, income is reduced by ~ 20% compared to an economy unaffected by climate change. Hereof ~ 40% are losses due to growth effects of which ~ 50% result from reduced incentive to invest. This additional income loss arises from a reduced incentive for future investment in anticipation of a reduced return and not from an explicit climate protection policy. Under economically optimal climate-change mitigation, however, optimal investment would only be reduced marginally as mitigation efforts keep returns high.

Fig. 1. Illustration of the investment effect.

Climate change reduces productivity, which translates into direct income losses (blue boxes). The prospect of reduced investment returns in the future renders investment less attractive. Accordingly, economically optimal investment is reduced and less production enhancing capital is accumulated. As a result, economic growth slows down and yields a future of persistently lowered income. This effect arising through reduced investment incentives is here referred to as the additional investment effect (depicted by red boxes).

Fig. 2. Schematic representation of the iterative procedure.

The estimated change in the annual growth rate due to temperature increase (a) is disentangled into (b) the respective temperature-sensitive productivity function and into (c) its associated optimal investment response in the business-as-usual scenario, which is characterised by inaction of climate policy. d The iterated growth rate converges towards the estimated growth rate after ~200 iterations. Source data are provided as a Source Data file.

Fig. 3. Comparison of the unadjusted and optimal investment rate.

The optimal investment rate is significantly lower than the unadjusted version. Note that the unadjusted investment rate is the same for a scenario with climate change and one without. Source data are provided as a Source Data file.

Fig. 4. The growth effects in the absence of climate policy (a, b) and under economically optimal mitigation of emissions (c, d).

a Unadjusted investment behaviour and particularly optimal investment lead to cumulative investment gaps through the climate effect on growth and the additional investment effect, respectively. b The income losses that occur for unadjusted investment behaviour (direct damage costs and the hereby induced growth effects) and for optimal investment (additional investment effect). c, d Economically optimal climate policy diminishes the climate effect on growth and renders the additional investment effect to be insignificant for (c) cumulative investment and for (d) income losses. Source data are provided as a Source Data file.

Fig. 5. The influence of social preferences on the growth effects and income losses.

Depicted are (relative to the income without climate change) the shares of (a) the total income losses; b the income losses caused through the growth effects; and (c) the income losses induced by the additional investment effect. Their magnitude depends on the social preferences. Solutions for alternative social preferences are illustrated by the grey area around the baseline solutions (red curves). The parameters are chosen uniformly within the unhatched area in Fig. 6. The shade of grey indicates the frequency with which the solution occurs. Source data are provided as a Source Data file.

Fig. 6. The effect of alternative social preferences.

a The ratio of the investment gaps $\frac{\left(I_{\mathrm{opt}}^{\mathrm{nocc}}-I_{\mathrm{opt}}^{\mathrm{cc}}\right)}{\left(I_{\mathrm{opt}}^{\mathrm{nocc}}-I_{\mathrm{unadj}}^{\mathrm{cc}}\right)}$ in 2100. b Difference between the temporally averaged unadjusted and optimal investment rates of the years 2010–2100. c Difference in the optimal investment rate between 2100 and 2010. The unhatched area depicts the range as commonly used in the economic literature and the white marker indicates the baseline calibration. Source data are provided as a Source Data file.

Fig. 7. Optimal mitigation and its effects.

a Optimal emission reduction rates are almost identical for the two assumptions of investment behaviour. b Economically optimal mitigation limits warming to almost 2 °C by 2100. c With unmitigated climate change, temperature-sensitive productivity decreases to ~86% by 2100, while economically optimal mitigation protects the economy from major productivity losses. Source data are provided as a Source Data file.