Social safety nets and US climate policy costs

Abstract

The debate surrounding US climate policy has been strongly influenced by concerns about the impact that placing a price on GHG emissions could have on potentially vulnerable populations, including senior citizens. This article shows that seniors would be one of the best protected groups under the climate policies recently debated in the US Congress. Moreover, low-income groups would also be protected generally under these proposals. This is due to a combination of existing government transfer programmes that automatically adjust benefit payments to account for inflation and provisions in the proposals.

Résumé

Le débat autour le la politique climatique des Etats-Unis a été fortement influence par des préoccupations sur l'impact potentiel de l'imposition d'un prix des émissions de GES sur des populations vulnérables, telles que les personnes âgées. Cet article montre que les personnes âgées seraient un des groupes les mieux protégés sous les politiques climatiques récemment débattues au sein du congrès américain. En outre, les groupes à faibles revenus seraient eux aussi généralement protégés sous ces propositions. Ceci est dû à une combinaison de programmes gouvernementaux existants de transferts qui ajustent automatiquement les paiements d'aides en prenant compte de l'inflation et des provisions contenues dans les propositions.

Keywords:

• allowance allocation
• cap-and-trade
• demographics
• distributional effects
• social impacts

Mots clés:

• allocation de quotas
• plafonnement échange
• démographique
• effets de distribution
• impacts sociaux

Notes

The approach in Waxman–Markey is largely mirrored in the Senate proposal by Kerry-Boxer (S. 1733: Clean Energy Jobs and American Power Act).

Fullerton and Rogers (1993) is the seminal study in this area; they looked at a range of taxes and constructed a sophisticated measure of lifetime income using panel data. Most studies lack such data and must therefore construct proxies in some other way (e.g. Walls and Hanson, 1999).

The EIA (2009) and EPA (2009) comprehensive assessments of the impacts of the Waxman–Markey legislation on the US economy also accounted for the indexing of taxes and benefits, in particular in their assessments of the costs to the government sector. These studies, however, did not look at the incidence across income or age groups.

Inflation would affect other programmes, including food stamps and Supplemental Security Income. However, these were excluded either because the magnitude of the benefit is small or its distribution cannot be extracted from the CE data. An increase in military pensions could be more important, but this was also excluded because of insufficient data.

The increase in benefits is fully offset by an increase in the payroll tax of 0.08% of eligible income.

On average, the proportional increase in income tax payments offsets the decreased burden that results from the automatic indexation of tax brackets. Consequently, although it has an important effect on some households, its effects by income group are small.

Alaska and Hawaii were excluded because they are not part of the Haiku electricity market model (discussed below) that is used here to estimate electricity prices.

In the data period, direct energy consumption accounts for 48% of CO₂ emissions associated with household expenditures (excluding government emissions). The CE data are comprehensive with respect to ‘out of pocket’ expenditures by households, but differ from the national income accounts because they do not include other types of expenditures (such as employer-provided benefits).

Baseline emissions intensities are taken from Burtraw et al. (2009) and Hassett et al. (2009).

Own-price elasticities and are taken from Burtraw et al. (2009).

This value pertains to all sectors except residential electricity and natural gas, which each have their own scalers. In the electricity sector the scaling parameter varies with the policy scenario.

This is essentially a consumer surplus loss measurement, scaled so that the burden equals the abatement cost of EIA.

If these investments are expected to benefit households in the long run, they should affect the long-run allowance price and thereby affect mitigation activity and contributions to the allowance bank in the near term. This inter-temporal effect is not accounted for here. This is analogous to the ‘pessimistic’ scenario described in Blonz et al. (2011), which also considers an ‘optimistic’ scenario in which the investments have an assumed benefit. These pessimistic and optimistic cases do not affect the general results regarding the distribution of net effects by age and income groups reported in this study.

In our interpretation of the Waxman–Markey scenario, some of the collected allowance value is assumed to be directed to programmes that do not provide any direct benefit to households. Under the cap-and-dividend scenario, the revenue is directed to households, who are assumed to use it in productive ways. In both cases, raising revenue takes funds away from households, causing a reduction in indirect spending, and expenditures associated with both programmes will lead to new indirect spending throughout the economy. Neither of the changes in indirect spending appears in the calculations.

See Blonz et al. (2011), who explore alternative assumptions.

‘Merchant coal units’ are independent electricity generators that sell power, usually into deregulated electricity markets. ‘Long-term contract generators’ refer to electricity generators who have entered into long-term contracts to sell power at a fixed price. ‘Domestic fuel producers’ are domestic fuel refineries that make gasoline for consumption.

Allocation to trade-exposed domestic industries leads to a higher allowance price than would otherwise occur without this allocation. This is because the allocation constitutes an output subsidy to those industries and they emit more CO₂ emissions than they otherwise would. Hence, greater reductions must occur elsewhere in the economy. This is captured in the marginal abatement cost (MAC) curve in the EIA modelling and thus in the allowance price.

Blonz et al. (2011) have found that a more optimistic set of assumptions regarding the effectiveness of spending on technology measures and allocation to LDCs would reduce the average burden of the policy, but that the distributional effects across households would be similar to the findings reported here.

Wage effects and different returns on investments are not included in the model. Thus, Table 4 shows the largest effects highlighted in this article. Some other effects, such as veterans’ benefits, the EITC and indexing of tax credits, are included in the calculations, but they are not shown separately because their effects on the distribution across age groups are small. The second-largest programme is the EITC, but this does not show much of an age effect because it is an income-based programme.