Full length articleEfficiency wages, unemployment, and environmental policy
Introduction
The effects that environmental policies may have on labor markets, and whether and to what extent they kill jobs or create jobs, is of utmost importance to policymakers. Much popular aversion to environmental regulation comes from its perceived negative impact on jobs. Other distributional impacts of policy, like the sources-side and uses-side incidence, can depend on frictions in the labor market that yield unemployment. It is important for policymakers to understand the effect of environmental policies on unemployment and on both factor and output prices.
There are several ways to go about addressing the general question of how environmental policies affect labor markets and unemployment. Many papers empirically estimate the impact of specific environmental policies on employment, including Martin et al. (2014), Curtis (2018) and Colmer et al. (2018). Other papers use computable general equilibrium (CGE) models to quantify the large-scale effects that policies like an economy-wide carbon tax might have, including Böhringer et al. (2003), Hafstead et al. (2018), and Castellanos and Heutel (2019). A third approach uses analytical general equilibrium modeling, which can shed light on the mechanisms behind the effects that can be quantified through empirical or CGE models. Both Hafstead and Williams (2018) and Aubert and Chiroleu-Assouline (2019) introduce pollution policy and unemployment resulting from labor search frictions into an analytical general equilibrium model. Our paper follows this third approach.
The purpose of this paper is to study the effect of pollution taxes on unemployment and incidence using an analytical general equilibrium model where unemployment is endogenously generated via efficiency wages. Workers' effort is a function of the real wage and the economy's unemployment level. Pollution is modeled as a production input along with capital and labor. We find closed-form analytic solutions for the general equilibrium responses to a change in the pollution tax rate, including expressions for changes in the unemployment rate, factor prices (the sources-side incidence), output prices (the uses-side incidence), and worker effort. The model allows us to decompose the net effects into substitution effects, output effects, and effects from the efficiency wage specification. Lastly, we conduct numerical simulations using calibrated parameter values.
Our modeling approach dates back to the canonical tax incidence modeling of Harberger (1962). Like Agell and Lundborg (1992) and Rapanos (2006), our paper adds an efficiency wage theory of unemployment to the model, though those papers do not model pollution. Like Fullerton and Heutel (2007), our paper adds pollution and pollution taxes to the model, though that paper does not model unemployment.1 We incorporate both efficiency wages and environmental policy into a Harberger-style analytical general equilibrium tax incidence model. Our paper is most similar to Hafstead and Williams (2018) and Aubert and Chiroleu-Assouline (2019), which both also model environmental policy and unemployment in an analytical general equilibrium setting. However, in both of those papers, unemployment arises from Diamond-Mortensen-Pissarides-style search frictions (Pissarides, 2000). In our paper, unemployment arises from efficiency wage theory (Akerlof, 1982; Shapiro and Stiglitz, 1984).2
Our theoretical results add new insights to the tax incidence literature. We identify effects that have been found in previous studies of environmental taxes, like the output and substitution effects. For example, an output effect exists such that the pollution tax disproportionately burdens the factor that is used more intensively in the polluting sector. These effects differ, though, when there is endogenous unemployment generated through efficiency wages. Along with these standard effects, we identify an effect that is new to the environmental tax literature, which we call the efficiency wage effect. The magnitude and direction of this effect depend on the form of the workers' effort function. Generally, the less elastic the workers' marginal effort response to the real wage is, the less burden labor bears, and the smaller increase in unemployment. With more structure on the effort function, we show that the crucial parameters of the effort function are the elasticities of effort with respect to the real wage and to unemployment. When effort responds more strongly to the real wage, then the magnitude of the efficiency wage effect is larger, which alleviates the tax burden on labor. When effort responds more strongly to unemployment, then the magnitude of the efficiency wage effect is smaller.
This key result depends on the efficiency wage specification causing unemployment, and so it is missing from previous studies that model unemployment through other causes. Our efficiency wage specification is general enough to accommodate different causes of efficiency wages and different effort functions. A gift exchange or fair wage efficiency wage model like Akerlof (1982) will lead to effort being very responsive to the real wage, and we find that this implies that the efficiency wage effect will be large. A shirking and firing model like Shapiro and Stiglitz (1984) will lead to effort being very responsive to unemployment, and we find that this implies that the efficiency wage effect will be small. Thus, the structural origin of unemployment fundamentally affects how large of an effect the unemployment friction will have on standard tax incidence outcomes.
The calibrated numerical simulation results, based on a $40 per ton carbon tax, provide further insights into these effects. The disproportionate burden of the tax on labor from substitution effects is offset by the disproportionate burden on capital from the efficiency wage effect. Ignoring the efficiency wage effect, as in previous environmental tax incidence models, thus gets the sign of the sources-side incidence wrong. Because of the efficiency wage effect, the carbon tax burdens capital disproportionately higher than labor. The tax increases the unemployment rate by just under 1%; this effect is mainly driven by a substitution effect from the larger, untaxed clean sector, rather than substitution within the smaller, taxed dirty sector. Sensitivity analyses show that the effects on unemployment and on sources-side incidence depend on the effort function elasticities, the effect on sources-side incidence also depends on production elasticities, and the effect on uses-side incidence is relatively unaffected by these parameters. Both the analytical and the numerical results highlight the important role of the efficiency wage effect and the form of the effort function in the analysis of pollution tax incidence and unemployment effects.
The paper is organized as follows. Section 2 presents the model and derives a system of linearized equations. Section 3 presents and interprets the general solution, decomposing the net effect into separate effects. Section 4 calibrates and numerically simulates the model. The last section concludes.
Section snippets
Model
Our model is a two-sector, two-factor incidence model, in the spirit of Harberger (1962), with the addition of involuntary unemployment through an efficiency wage as in Agell and Lundborg (1992) and Rapanos (2006), and with the addition of pollution as in Fullerton and Heutel, 2007, Fullerton and Heutel, 2010. The model is linearized so that it can be solved analytically. We consider a competitive two-sector economy using two factors of production: capital and labor. Both factors are perfectly
Solution
Our focus is on incidence and unemployment effects, so we are most interested in solutions for changes in factor prices ( and ), output prices ( and ), and unemployment .
We present three closed-form solutions. The first is , which is the change in unemployment or the unemployment rate.14
Numerical simulations
Here we numerically simulate the model by assigning parameter values calibrated from data and taken from the previous literature. Ours is a simple two-sector, two-input model, not a CGE model, so the purpose of these simulations is not to pin down plausible quantitative values for the magnitudes of these effects. Rather, the purpose is to explore how the net effects are decomposed into the effects identified in the previous section and how sensitive the magnitudes are to various parameter
Conclusion
We use an analytical general equilibrium model with unemployment generated through efficiency wages to analyze the effect of a pollution tax on unemployment and on sources-side and uses-side incidence. We decompose the general equilibrium impact of the tax on unemployment and incidence into several effects, including an output effect, substitution effects, and an effect that we call the efficiency wage effect, which has not been previously identified in the environmental tax incidence
Credit author statement
Garth Heutel: Conceptualization, Methodology, Validation, Formal analysis, Writing- Original Draft, Writing – Review & Editing, Supervision.
Xin Zhang: Conceptualization, Methodology, Software, Validation, Formal analysis, Writing- Original Draft, Writing – Review & Editing, Visualization.
References (32)
- et al.
Fair wages, involuntary unemployment and tax policies in the simple general equilibrium model
J. Public Econ.
(1992) - et al.
Environmental tax reform and income distribution with imperfect heterogeneous labour markets
Eur. Econ. Rev.
(2019) - et al.
Can carbon taxes be progressive?
Energy Econ.
(2014) - et al.
The general equilibrium incidence of environmental taxes
J. Public Econ.
(2007) - et al.
Can pollution tax rebates protect low-wage earners?
J. Environ. Econ. Manag.
(2013) Distributional effects of carbon taxes: The case of Mexico
Energy Econ.
(2012)- et al.
Unemployment and environmental regulation in general equilibrium
J. Public Econ.
(2018) General equilibrium models with Morishima elasticities of substitution in production
Econ. Model.
(2016)- et al.
The impact of a carbon tax on manufacturing: Evidence from microdata
J. Public Econ.
(2014) Another possible source of wage stickiness
J. Macroecon.
(1979)