Carbon Pricing Winner and Losers: Workers, Consumers, and Policy Options SEPTEMBER 25, 2020
Marina M. Tavares Research Department IMF | Research
1
Climate Change and Carbon Taxes • Climate Change is the defining issue of our time. • Carbon taxes are the most efficient instrument to fight climate change. • The support for carbon taxes still lacks due to their cost on output and inequality. • Carbon-taxes are known to be regressive because:
IMF | Research
1.
Poor households spend a large share of their income on energy-intensive goods.
2.
Poor households (unskilled) are more likely to work in energy-intensive sectors.
2
Paper Objective • Quantify the impact of $50 dollar per ton of CO2 carbon price on income inequality in the United States, taking into consideration the consumption and employment channel.
• Design complementary measures to neutralize the carbon tax’s negative impact.
IMF | Research
3
Literature Review • There are a large literature that studies the distributional impact of carbon taxes: 1.
Incidence Analysis – IMF(2019), Grainger and Kolstad (2010), Fremstad and Paul (2019), Hassett et al (2007), among many others.
2.
Computable General Equilibrium Model (CGE) – Chateau et al. (2018), Goulder et al. (2019), Rausch et a. (2011).
3.
Mix – Williams et al. (2015), Chen et al. (2020).
4.
General Equilibrium Models – Fullerton and Monti (2013), Aubert and ChiroleuAssouline (2019), Klenert et al. (2018), Fried et al. (2018).
5.
IMF | Research
Empirical Studies – Marin and Vona (2019).
4
Outline 1. Empirical Evidence 2. Model Outline 3. Calibration 4. Results 5. Conclusion
IMF | Research
5
Poor households work more in high-energy intensive sectors and consume more high-energy intensive goods Household Consumption Expenditure
Sectoral Employment Share
(Percent of household total consumption expenditure)
(Share of workers employed in each sector by skill level)
1.00 0.90
0.90 0.80 0.70
0.80
Unskilled
0.70
Skilled
0.60
0.60 0.50
0.50
0.40
0.40
0.30
0.30
0.20
0.20
0.10
0.10
0.00
Q1 Energy
Q2
High-energy intensive good
Source: Consumption Expenditure Survey (CEX) IMF | Research
Q3
Q4
Q5
Low-energy intensive good
0.00
High-Energy Intensive
Low-Energy Intensive
Energy
Source: American Community Survey (ACS) 6
Unskilled workers have a less favorable view regarding environment policies Public opinion at different education levels (Percent)
‌ and occupation types (Percent) 60
80 75
58
70 65
56
60 55
54
50 45
52
40 35 30
50 Skilled
Unskilled
Clean occupation
Dirty occupation
Skilled
Clean occupation
Dirty occupation
Unskilled
Source: World Values Survey, wave 7 (2017–20). IMF | Research
7
Model
IMF | Research
8
Economy • Small open economy model with heterogenous agents • The economy is populated by two types of households: Skilled and Unskilled • Four type of goods are produced by competitive firms: clean energy, dirty energy, high-energy intensive goods, and low-energy intensive goods.
• Low-energy intensive good production uses more skilled labor and less energy. • High-energy intensive good production uses more unskilled labor and more energy. • Energy uses only unskilled labor and capital. • Clean energy is more labor intensive than dirty energy. IMF | Research
9
Households • Households have non-homothetic preferences, and they supply labor. • They face idiosyncratic shocks that can be partially insured using a risk-free bond. • They choose: (1) each good consumption, (2) labor supply, and (3) savings. • Carbon taxes affect households by impacting:
IMF | Research
1.
relative prices of dirty to clean energy (direct)
2.
relative prices of high-energy intensive goods (indirect)
3.
households’ wage (indirect)
10
Final Goods Firms • Firms maximize profit by (1) renting capital, hiring (2) skilled and (3) unskilled, and
purchasing (4) clean, and (5) dirty energy.
• Carbon taxes affect firms by impacting: 1.
relative prices of dirty to clean energy (direct)
2.
relative prices of the final good produced (indirect)
3.
labor markets (indirect)
• High-energy intensive goods are going to be more impacted • More energy intensive firm is also more unskilled intensive.
IMF | Research
11
Government • Government collects a carbon tax đ?œ?đ?œ?.
• Government uses the revenue to finance alternative scheme: 1. wasteful government spending, 2. cash-transfer program (targeted or universal), 3. subsidy to clean energy consumption (“feebateâ€?).
• Government budget constraint is satisfied every period.
IMF | Research
12
Calibration
IMF | Research
13
Calibration • Employment – American Community Survey (ACS) • Three economy sectors (IEA): High-Energy Intensive, Low-Energy Intensive, and Energy • Two workers’ types: skilled (more than high-school) and unskilled (less than high-school)
• Consumption – Consumption Expenditure Survey (CEX) • Three consumption goods: High-Energy Intensive, Low-Energy Intensive, and Energy • Adjust consumption using Personal Consumption Expenditure (PCE) from the BEA
• Main Elasticities are from the Literature • Elasticity of substitution between clean and dirty energy equal to 3 • Elasticity of substitution between energy and capital-labor equal to 0.5 • Elasticity of substitution between skilled and unskilled workers equal to 2
IMF | Research
14
Results
IMF | Research
15
Numerical Exercise • I simulate the impact of $50 dollar per ton of CO2. • The tax is levied on: 1. 2.
•
IMF | Research
Households’ consumption of dirty energy Firms’ consumption of dirty energy
Recycle revenue to finance: 1. Government spending in low-energy intensive goods 2. Universal cash-transfer program 3. Target cash-transfer program 4. Subsidy to clean energy consumption (“feebates”)
16
Aggregate and Distributional Effects Aggregate Impact of a Carbon Tax
Distributional Impact of a Carbon Tax
0.50
2.00
(Percentage Change Relative to no-tax baseline)
Low-Energy Intensive Goods Universal Cash-Transfer Targeted Cash-Transfer
0.00
(Percentage Change Relative to no-tax baseline)
1.50 1.00
Low-Energy Intensive Goods Universal Cash-Transfer Targeted Cash-Transfer
0.50
-0.50
0.00 -0.50
-1.00
-1.00 -1.50
-1.50
-2.00 -2.00
IMF | Research
Output
Consuption
Effective Hours
-2.50
Gini
Skill Premium 17
Non-Environmental Welfare Effects Welfare Impact of a Carbon Tax
(Consumption Equivalent relative to no-tax baseline)
Low-Energy Intensive Goods
Targeted Cash-Transfer
Welfare Decomposition
(Consumption Equivalent relative to no-tax baseline)
0.21
General Equilibrium Effects
0.37
Unskilled Skilled
Universal Cash-Transfer
Aggregate
-1.5
IMF | Research
-1
-0.5
0
0.5
0.42
Income Channel
Consumption Channel
Low-Energy Intensive Good
18
Feebates Distributional Impact of a Carbon Tax
Consumption Expenditure by Quintile
(Percentage Change Relative to no-tax baseline)
(Percentage Change Relative to no-tax baseline)
0.50
2.00
0.00
1.50
-0.50
Low-Energy Intensive Goods Feebates
1.00
-1.00 Low-Energy Intensive Goods
-1.50 -2.00
Feebates
-2.50 -3.00
IMF | Research
0.50 0.00 -0.50 -1.00
Output
Consuption
Effective Hours
Gini
Skill Premium
19
Conclusion • This paper studies the impact of carbon on income distribution, taking into consideration the employment channel.
• A 50 USD per ton Tax on CO2 may reduce income inequality depending on how the revenue is used.
• When the revenue is used to finance a targeted cash-transfer program, the carbon tax impact on the bottom quintile of the income distribution is fully neutralized.
IMF | Research
20
Equilibrium Given a distribution of skilled and unskilled workers đ?œ‡đ?œ‡, carbon tax đ?œ?đ?œ?, interest rates đ?‘&#x;đ?‘&#x;, factor prices
{đ?‘¤đ?‘¤ đ?‘ đ?‘ , đ?‘¤đ?‘¤ đ?‘˘đ?‘˘ , đ?‘?đ?‘?đ?‘™đ?‘™ , đ?‘?đ?‘?đ?‘?đ?‘? , đ?‘?đ?‘?đ?‘‘đ?‘‘ }, and capital depreciation đ?›żđ?›ż, a stationary competitive equilibrium consists of workers’
decision rules, đ?‘?đ?‘? đ?‘—đ?‘—,â„Ž , đ?‘?đ?‘? đ?‘—đ?‘—,đ?‘™đ?‘™ , đ?‘™đ?‘™ đ?‘—đ?‘— , đ?‘’đ?‘’ đ?‘—đ?‘—,đ?‘?đ?‘? , đ?‘’đ?‘’ đ?‘—đ?‘—,đ?‘‘đ?‘‘ , đ?‘Žđ?‘Žâ€˛đ?‘—đ?‘— đ??žđ??žđ?‘—đ?‘— , đ??żđ??żđ?‘—đ?‘—,đ?‘ đ?‘ , đ??żđ??żđ?‘—đ?‘—,đ?‘˘đ?‘˘ , đ??¸đ??¸đ?‘—đ?‘—,đ?‘‘đ?‘‘ , đ??¸đ??¸đ?‘—đ?‘—,đ?‘?đ?‘?
đ?‘—đ?‘—∈{đ?‘˘đ?‘˘,đ?‘ đ?‘ }
, goods firms’ production plan
đ?‘—đ?‘— đ?‘˘đ?‘˘,đ?‘—đ?‘— , energy firms’ production plan đ??žđ??ž , đ??żđ??ż đ?‘—đ?‘—∈{â„Ž,đ?‘™đ?‘™}
Γ(��), such that the following holds:
đ?‘—đ?‘—∈{đ?‘?đ?‘?,đ?‘‘đ?‘‘}
, and the distribution of agents,
1. Given prices and policies, all households solve their recursive problem. 2. Given prices and policies, all firms solve their optimization problem. 3. Government budget balances every period. 4. Labor market clears. 5. Low-energy intensive goods and energy markets clear.
IMF | Research
6. The distribution Γ(��) is stationary.
21
Household Maximization Problem 𝑉𝑉𝑗𝑗 𝑎𝑎, 𝑧𝑧 =
max
𝑐𝑐 𝑙𝑙 ,𝑐𝑐 ℎ ,𝑒𝑒,𝑎𝑎′𝑗𝑗
{𝑢𝑢 𝑐𝑐 𝑙𝑙 + 𝑐𝑐̅𝑙𝑙 , 𝑐𝑐 ℎ , 𝑒𝑒 − 𝑒𝑒,̅ 𝑙𝑙 + 𝛽𝛽𝛽𝛽(𝑉𝑉𝑗𝑗 (𝑎𝑎′ , 𝑧𝑧′))
𝑒𝑒 = 𝜇𝜇𝑒𝑒 𝑒𝑒
𝑐𝑐 𝜌𝜌𝑒𝑒
+ 1 − 𝜇𝜇𝑒𝑒 𝑒𝑒
1/𝜌𝜌𝑒𝑒 𝑑𝑑 𝜌𝜌𝑒𝑒
𝑐𝑐 ℎ + 𝑝𝑝𝑙𝑙 𝑐𝑐 𝑙𝑙 + 𝑝𝑝𝑐𝑐 𝑒𝑒 𝑐𝑐 + 1 + 𝜏𝜏 𝑝𝑝𝑑𝑑 𝑒𝑒 𝑑𝑑 + 𝑎𝑎′ ≤ 𝑤𝑤 𝑗𝑗 ℎ 𝑗𝑗 𝑧𝑧𝑧𝑧 + 1 + 𝑟𝑟 𝑎𝑎 + 𝑇𝑇(𝑤𝑤 𝑗𝑗 ℎ 𝑗𝑗 𝑧𝑧𝑧𝑧) ln 𝑧𝑧 ′ = 𝜌𝜌 ln 𝑧𝑧 + 𝜖𝜖𝜖 where 𝜖𝜖~𝑁𝑁(0, 𝜎𝜎ϵ2 )
c ℎ ≥ 0, c 𝑙𝑙 ≥ 0, 𝑒𝑒 𝑐𝑐 ≥ 0, 𝑒𝑒 𝑑𝑑 ≥ 0, 𝑎𝑎′ ≥ 0, 0 ≤ 𝑙𝑙 𝑗𝑗 ≤ 1 and 𝑗𝑗 ∈ {𝑠𝑠, 𝑢𝑢} IMF | Research
22
Final Goods Sector Technology • Final good sector with energy intensity j:
đ?‘—đ?‘—
đ?‘Œđ?‘Œ = đ??´đ??´
đ?‘—đ?‘—
đ?œ‡đ?œ‡đ?‘—đ?‘—
đ??žđ??ž
�� �� ��
đ??żđ??ż
đ?‘—đ?‘— đ?‘—đ?‘— 1−đ?›źđ?›ź
đ?œ‚đ?œ‚đ?‘—đ?‘—
+ (1 − đ?œ‡đ?œ‡đ?‘—đ?‘— ) đ??¸đ??¸ đ?‘—đ?‘—
đ?œ‚đ?œ‚đ?‘—đ?‘—
1 đ?œ‚đ?œ‚đ?‘—đ?‘—
where đ?‘—đ?‘— ∈ {â„Ž,l} and đ?œ‡đ?œ‡â„Ž > đ?œ‡đ?œ‡đ?‘™đ?‘™
IMF | Research
23
Final Goods Sector Technology • Energy is the combination of clean 𝐸𝐸 𝑐𝑐 and dirty energy 𝐸𝐸 𝑑𝑑 : 𝐸𝐸 𝑗𝑗
=�
𝜌𝜌𝑗𝑗,𝑒𝑒 𝑣𝑣𝑗𝑗 𝐸𝐸𝑗𝑗𝑐𝑐
+ 1 − 𝑣𝑣𝑗𝑗
1/𝜌𝜌𝑗𝑗,𝑒𝑒 𝑑𝑑 𝜌𝜌𝑗𝑗,𝑒𝑒 𝐸𝐸𝑗𝑗 �
• Labor is the combination of skilled labor 𝐿𝐿𝑠𝑠 and unskilled labor 𝐿𝐿𝑢𝑢 : 𝐿𝐿𝑗𝑗
= �
𝜌𝜌𝑗𝑗,𝑙𝑙 𝜇𝜇𝑗𝑗 𝐿𝐿𝑗𝑗𝑠𝑠
where 𝑗𝑗 ∈ {ℎ,l} and 𝜈𝜈ℎ < 𝜈𝜈𝑙𝑙 and 𝜇𝜇ℎ < 𝜇𝜇𝑙𝑙
IMF | Research
+ 1 − 𝜇𝜇𝑗𝑗
1/𝜌𝜌𝑗𝑗,𝑙𝑙 𝑢𝑢 𝜌𝜌𝑗𝑗,𝑙𝑙 𝐿𝐿𝑗𝑗 �
24