SlideShare a Scribd company logo

DDKT-Munich.pdf

GRAPE
GRAPE

Munich

1 of 49
Download to read offline
Incentive separability
Pawel Doligalski (Bristol & GRAPE)
Piotr Dworczak (Northwestern & GRAPE)
Joanna Krysta (Standford)
Filip Tokarski (Stanford GSB)
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon
2020 research and innovation programme (ERC Starting grant IMD-101040122)
1
Efficiency-equity trade-off: distortions are useful for redistribution
• central concept in Public Finance
2
Efficiency-equity trade-off: distortions are useful for redistribution
• central concept in Public Finance
However, sometimes the efficiency-equity trade-off can be avoided:
• Diamond-Mirrlees ’71: distortionary taxes on firms are redundant
• Atkinson-Stiglitz ’76: differentiated consumption taxes are redundant with
a non-linear income tax
• Sadka ’76, Seade ’77: there should be “no distortion at the top”
We study the general logic behind these results using Mechanism Design
2
What we do
We consider a general framework allowing for complex incentive constraints
- private information, moral hazard, voluntary participation etc
We introduce a new notion of incentive separability
- A set of decisions is incentive-separable if perturbing these decisions along
agents’ indifference curves preserves all incentive constraints.
3
What we do
We consider a general framework allowing for complex incentive constraints
- private information, moral hazard, voluntary participation etc
We introduce a new notion of incentive separability
- A set of decisions is incentive-separable if perturbing these decisions along
agents’ indifference curves preserves all incentive constraints.
Main theorem: It is optimal to allow agents to make unrestricted choices over
incentive-separable decisions, given some prices and budgets.
3
What we do
We consider a general framework allowing for complex incentive constraints
- private information, moral hazard, voluntary participation etc
We introduce a new notion of incentive separability
- A set of decisions is incentive-separable if perturbing these decisions along
agents’ indifference curves preserves all incentive constraints.
Main theorem: It is optimal to allow agents to make unrestricted choices over
incentive-separable decisions, given some prices and budgets.
The theorem allows us to
• Extend and unify Atkinson-Stiglitz and Diamond-Mirrlees theorems
- optimal to jointly remove consumption and production distortions
- harmonizing consumption taxes with distorted production—a bad idea
• Propose a novel argument for the optimality of food vouchers
3

Recommended

DDKT-Praga.pdf
DDKT-Praga.pdfDDKT-Praga.pdf
DDKT-Praga.pdfGRAPE
 
DDKT-Southern.pdf
DDKT-Southern.pdfDDKT-Southern.pdf
DDKT-Southern.pdfGRAPE
 
DDKT-SummerWorkshop.pdf
DDKT-SummerWorkshop.pdfDDKT-SummerWorkshop.pdf
DDKT-SummerWorkshop.pdfGRAPE
 
DDKT-SAET.pdf
DDKT-SAET.pdfDDKT-SAET.pdf
DDKT-SAET.pdfGRAPE
 
Options Portfolio Selection
Options Portfolio SelectionOptions Portfolio Selection
Options Portfolio Selectionguasoni
 
Presentation.pdf
Presentation.pdfPresentation.pdf
Presentation.pdfGRAPE
 
Hybrid dynamics in large-scale logistics networks
Hybrid dynamics in large-scale logistics networksHybrid dynamics in large-scale logistics networks
Hybrid dynamics in large-scale logistics networksMKosmykov
 

More Related Content

Similar to DDKT-Munich.pdf

A Generalized Metric Space and Related Fixed Point Theorems
A Generalized Metric Space and Related Fixed Point TheoremsA Generalized Metric Space and Related Fixed Point Theorems
A Generalized Metric Space and Related Fixed Point TheoremsIRJET Journal
 
Machine Learning in Actuarial Science & Insurance
Machine Learning in Actuarial Science & InsuranceMachine Learning in Actuarial Science & Insurance
Machine Learning in Actuarial Science & InsuranceArthur Charpentier
 
Inference for stochastic differential equations via approximate Bayesian comp...
Inference for stochastic differential equations via approximate Bayesian comp...Inference for stochastic differential equations via approximate Bayesian comp...
Inference for stochastic differential equations via approximate Bayesian comp...Umberto Picchini
 
quebec.pdf
quebec.pdfquebec.pdf
quebec.pdfykyog
 
Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Asma Ben Slimene
 
Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Asma Ben Slimene
 
Predictive Modeling in Insurance in the context of (possibly) big data
Predictive Modeling in Insurance in the context of (possibly) big dataPredictive Modeling in Insurance in the context of (possibly) big data
Predictive Modeling in Insurance in the context of (possibly) big dataArthur Charpentier
 
block-mdp-masters-defense.pdf
block-mdp-masters-defense.pdfblock-mdp-masters-defense.pdf
block-mdp-masters-defense.pdfJunghyun Lee
 
International conference "QP 34 -- Quantum Probability and Related Topics"
International conference "QP 34 -- Quantum Probability and Related Topics"International conference "QP 34 -- Quantum Probability and Related Topics"
International conference "QP 34 -- Quantum Probability and Related Topics"Medhi Corneille Famibelle*
 
Poggi analytics - star - 1a
Poggi   analytics - star - 1aPoggi   analytics - star - 1a
Poggi analytics - star - 1aGaston Liberman
 
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptxsaadurrehman35
 
Econometrics homework help
Econometrics homework helpEconometrics homework help
Econometrics homework helpMark Austin
 
Fuzzy Logic.pptx
Fuzzy Logic.pptxFuzzy Logic.pptx
Fuzzy Logic.pptxImXaib
 
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...Ioannis Partalas
 

Similar to DDKT-Munich.pdf (20)

A Generalized Metric Space and Related Fixed Point Theorems
A Generalized Metric Space and Related Fixed Point TheoremsA Generalized Metric Space and Related Fixed Point Theorems
A Generalized Metric Space and Related Fixed Point Theorems
 
2005 f c49_note02
2005 f c49_note022005 f c49_note02
2005 f c49_note02
 
Slides ensae-2016-9
Slides ensae-2016-9Slides ensae-2016-9
Slides ensae-2016-9
 
Machine Learning in Actuarial Science & Insurance
Machine Learning in Actuarial Science & InsuranceMachine Learning in Actuarial Science & Insurance
Machine Learning in Actuarial Science & Insurance
 
Inference for stochastic differential equations via approximate Bayesian comp...
Inference for stochastic differential equations via approximate Bayesian comp...Inference for stochastic differential equations via approximate Bayesian comp...
Inference for stochastic differential equations via approximate Bayesian comp...
 
Slides angers-sfds
Slides angers-sfdsSlides angers-sfds
Slides angers-sfds
 
Slides ensae 9
Slides ensae 9Slides ensae 9
Slides ensae 9
 
Slides Bank England
Slides Bank EnglandSlides Bank England
Slides Bank England
 
CLIM Fall 2017 Course: Statistics for Climate Research, Statistics of Climate...
CLIM Fall 2017 Course: Statistics for Climate Research, Statistics of Climate...CLIM Fall 2017 Course: Statistics for Climate Research, Statistics of Climate...
CLIM Fall 2017 Course: Statistics for Climate Research, Statistics of Climate...
 
quebec.pdf
quebec.pdfquebec.pdf
quebec.pdf
 
Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...Research internship on optimal stochastic theory with financial application u...
Research internship on optimal stochastic theory with financial application u...
 
Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...Presentation on stochastic control problem with financial applications (Merto...
Presentation on stochastic control problem with financial applications (Merto...
 
Predictive Modeling in Insurance in the context of (possibly) big data
Predictive Modeling in Insurance in the context of (possibly) big dataPredictive Modeling in Insurance in the context of (possibly) big data
Predictive Modeling in Insurance in the context of (possibly) big data
 
block-mdp-masters-defense.pdf
block-mdp-masters-defense.pdfblock-mdp-masters-defense.pdf
block-mdp-masters-defense.pdf
 
International conference "QP 34 -- Quantum Probability and Related Topics"
International conference "QP 34 -- Quantum Probability and Related Topics"International conference "QP 34 -- Quantum Probability and Related Topics"
International conference "QP 34 -- Quantum Probability and Related Topics"
 
Poggi analytics - star - 1a
Poggi   analytics - star - 1aPoggi   analytics - star - 1a
Poggi analytics - star - 1a
 
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx
53158699-d7c5-4e6e-af19-1f642992cc58-161011142651.pptx
 
Econometrics homework help
Econometrics homework helpEconometrics homework help
Econometrics homework help
 
Fuzzy Logic.pptx
Fuzzy Logic.pptxFuzzy Logic.pptx
Fuzzy Logic.pptx
 
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...
Aggressive Sampling for Multi-class to Binary Reduction with Applications to ...
 

More from GRAPE

Tone at the top: the effects of gender board diversity on gender wage inequal...
Tone at the top: the effects of gender board diversity on gender wage inequal...Tone at the top: the effects of gender board diversity on gender wage inequal...
Tone at the top: the effects of gender board diversity on gender wage inequal...GRAPE
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingGRAPE
 
ENTIME_GEM___GAP.pdf
ENTIME_GEM___GAP.pdfENTIME_GEM___GAP.pdf
ENTIME_GEM___GAP.pdfGRAPE
 
Boston_College Slides.pdf
Boston_College Slides.pdfBoston_College Slides.pdf
Boston_College Slides.pdfGRAPE
 
Presentation_Yale.pdf
Presentation_Yale.pdfPresentation_Yale.pdf
Presentation_Yale.pdfGRAPE
 
Presentation_Columbia.pdf
Presentation_Columbia.pdfPresentation_Columbia.pdf
Presentation_Columbia.pdfGRAPE
 
Presentation.pdf
Presentation.pdfPresentation.pdf
Presentation.pdfGRAPE
 
Presentation.pdf
Presentation.pdfPresentation.pdf
Presentation.pdfGRAPE
 
Slides.pdf
Slides.pdfSlides.pdf
Slides.pdfGRAPE
 
Slides.pdf
Slides.pdfSlides.pdf
Slides.pdfGRAPE
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingGRAPE
 
Matching it up: non-standard work and job satisfaction.pdf
Matching it up: non-standard work and job satisfaction.pdfMatching it up: non-standard work and job satisfaction.pdf
Matching it up: non-standard work and job satisfaction.pdfGRAPE
 
Investment in human capital: an optimal taxation approach
Investment in human capital: an optimal taxation approachInvestment in human capital: an optimal taxation approach
Investment in human capital: an optimal taxation approachGRAPE
 
Sponsored content in contextual bandits. Deconfounding targeting not at random
Sponsored content in contextual bandits. Deconfounding targeting not at randomSponsored content in contextual bandits. Deconfounding targeting not at random
Sponsored content in contextual bandits. Deconfounding targeting not at randomGRAPE
 
slides_cef.pdf
slides_cef.pdfslides_cef.pdf
slides_cef.pdfGRAPE
 
Fertility, contraceptives and gender inequality
Fertility, contraceptives and gender inequalityFertility, contraceptives and gender inequality
Fertility, contraceptives and gender inequalityGRAPE
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingGRAPE
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingGRAPE
 
Matching it up
Matching it upMatching it up
Matching it upGRAPE
 
Childbearing and gender norms
Childbearing and gender normsChildbearing and gender norms
Childbearing and gender normsGRAPE
 

More from GRAPE (20)

Tone at the top: the effects of gender board diversity on gender wage inequal...
Tone at the top: the effects of gender board diversity on gender wage inequal...Tone at the top: the effects of gender board diversity on gender wage inequal...
Tone at the top: the effects of gender board diversity on gender wage inequal...
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population aging
 
ENTIME_GEM___GAP.pdf
ENTIME_GEM___GAP.pdfENTIME_GEM___GAP.pdf
ENTIME_GEM___GAP.pdf
 
Boston_College Slides.pdf
Boston_College Slides.pdfBoston_College Slides.pdf
Boston_College Slides.pdf
 
Presentation_Yale.pdf
Presentation_Yale.pdfPresentation_Yale.pdf
Presentation_Yale.pdf
 
Presentation_Columbia.pdf
Presentation_Columbia.pdfPresentation_Columbia.pdf
Presentation_Columbia.pdf
 
Presentation.pdf
Presentation.pdfPresentation.pdf
Presentation.pdf
 
Presentation.pdf
Presentation.pdfPresentation.pdf
Presentation.pdf
 
Slides.pdf
Slides.pdfSlides.pdf
Slides.pdf
 
Slides.pdf
Slides.pdfSlides.pdf
Slides.pdf
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population aging
 
Matching it up: non-standard work and job satisfaction.pdf
Matching it up: non-standard work and job satisfaction.pdfMatching it up: non-standard work and job satisfaction.pdf
Matching it up: non-standard work and job satisfaction.pdf
 
Investment in human capital: an optimal taxation approach
Investment in human capital: an optimal taxation approachInvestment in human capital: an optimal taxation approach
Investment in human capital: an optimal taxation approach
 
Sponsored content in contextual bandits. Deconfounding targeting not at random
Sponsored content in contextual bandits. Deconfounding targeting not at randomSponsored content in contextual bandits. Deconfounding targeting not at random
Sponsored content in contextual bandits. Deconfounding targeting not at random
 
slides_cef.pdf
slides_cef.pdfslides_cef.pdf
slides_cef.pdf
 
Fertility, contraceptives and gender inequality
Fertility, contraceptives and gender inequalityFertility, contraceptives and gender inequality
Fertility, contraceptives and gender inequality
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population aging
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population aging
 
Matching it up
Matching it upMatching it up
Matching it up
 
Childbearing and gender norms
Childbearing and gender normsChildbearing and gender norms
Childbearing and gender norms
 

Recently uploaded

The Case for Commodities as an Asset Class.ppt
The Case for Commodities as an Asset Class.pptThe Case for Commodities as an Asset Class.ppt
The Case for Commodities as an Asset Class.pptJinzhiFeng1
 
99MW Nearshore Wind Farm in Colombia.pdf
99MW Nearshore Wind Farm in Colombia.pdf99MW Nearshore Wind Farm in Colombia.pdf
99MW Nearshore Wind Farm in Colombia.pdfVictoriaGaleano
 
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptx
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptxSlideshare - ONS Economic Forum Slidepack - 19 February 2024.pptx
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptxOffice for National Statistics
 
Shrambal_Distributors_February_Newsletter_2024.pdf
Shrambal_Distributors_February_Newsletter_2024.pdfShrambal_Distributors_February_Newsletter_2024.pdf
Shrambal_Distributors_February_Newsletter_2024.pdfvikashdidwania1
 
TIM GROUP FY '23 Preliminary Results.pdf
TIM GROUP FY '23 Preliminary Results.pdfTIM GROUP FY '23 Preliminary Results.pdf
TIM GROUP FY '23 Preliminary Results.pdfGruppo TIM
 
National Income National Income National Income
National Income National Income National IncomeNational Income National Income National Income
National Income National Income National Income19920606201903202019
 
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...Työeläkeyhtiö Elo
 
India Equity Strategy Presentation (2016) - HSBC
India Equity Strategy Presentation (2016) - HSBCIndia Equity Strategy Presentation (2016) - HSBC
India Equity Strategy Presentation (2016) - HSBCDimitarStoyanov46
 
E-learning Personal Finance Management - with design v2 .pptx
E-learning Personal Finance Management - with design v2 .pptxE-learning Personal Finance Management - with design v2 .pptx
E-learning Personal Finance Management - with design v2 .pptxakinsumboayomide
 
Stock Market Brief Deck for 2:22 youtube vid.pdf
Stock Market Brief Deck for 2:22 youtube vid.pdfStock Market Brief Deck for 2:22 youtube vid.pdf
Stock Market Brief Deck for 2:22 youtube vid.pdfMichael Silva
 
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdfAdnet Communications
 
Stock Market Brief Deck from youtube 2/16
Stock Market Brief Deck from youtube 2/16Stock Market Brief Deck from youtube 2/16
Stock Market Brief Deck from youtube 2/16Michael Silva
 
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptx
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptxAGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptx
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptxVISHALI SELVAM
 
Stock Market Brief Deck for 2/21/24 .pdf
Stock Market Brief Deck for 2/21/24 .pdfStock Market Brief Deck for 2/21/24 .pdf
Stock Market Brief Deck for 2/21/24 .pdfMichael Silva
 
2023_-_Master_Trust___GPP_Defaults_Report (2).pdf
2023_-_Master_Trust___GPP_Defaults_Report (2).pdf2023_-_Master_Trust___GPP_Defaults_Report (2).pdf
2023_-_Master_Trust___GPP_Defaults_Report (2).pdfHenry Tapper
 
business-model-canvas for collaborative business
business-model-canvas for collaborative businessbusiness-model-canvas for collaborative business
business-model-canvas for collaborative businessramijadpro
 
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...Adrian Sanabria
 
letter to the EF and Premiership about pension liabilities
letter to the EF and Premiership about pension liabilitiesletter to the EF and Premiership about pension liabilities
letter to the EF and Premiership about pension liabilitiesHenry Tapper
 

Recently uploaded (20)

The Case for Commodities as an Asset Class.ppt
The Case for Commodities as an Asset Class.pptThe Case for Commodities as an Asset Class.ppt
The Case for Commodities as an Asset Class.ppt
 
99MW Nearshore Wind Farm in Colombia.pdf
99MW Nearshore Wind Farm in Colombia.pdf99MW Nearshore Wind Farm in Colombia.pdf
99MW Nearshore Wind Farm in Colombia.pdf
 
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptx
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptxSlideshare - ONS Economic Forum Slidepack - 19 February 2024.pptx
Slideshare - ONS Economic Forum Slidepack - 19 February 2024.pptx
 
Shrambal_Distributors_February_Newsletter_2024.pdf
Shrambal_Distributors_February_Newsletter_2024.pdfShrambal_Distributors_February_Newsletter_2024.pdf
Shrambal_Distributors_February_Newsletter_2024.pdf
 
Veritas Financial statement presentation 2023
Veritas Financial statement presentation 2023Veritas Financial statement presentation 2023
Veritas Financial statement presentation 2023
 
TIM GROUP FY '23 Preliminary Results.pdf
TIM GROUP FY '23 Preliminary Results.pdfTIM GROUP FY '23 Preliminary Results.pdf
TIM GROUP FY '23 Preliminary Results.pdf
 
National Income National Income National Income
National Income National Income National IncomeNational Income National Income National Income
National Income National Income National Income
 
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...
Elo’s result 2023: Return on investment increased to 6 per cent and cost effi...
 
India Equity Strategy Presentation (2016) - HSBC
India Equity Strategy Presentation (2016) - HSBCIndia Equity Strategy Presentation (2016) - HSBC
India Equity Strategy Presentation (2016) - HSBC
 
Monthly Economic Monitoring of Ukraine No.229, February 2024
Monthly Economic Monitoring of Ukraine No.229, February 2024Monthly Economic Monitoring of Ukraine No.229, February 2024
Monthly Economic Monitoring of Ukraine No.229, February 2024
 
E-learning Personal Finance Management - with design v2 .pptx
E-learning Personal Finance Management - with design v2 .pptxE-learning Personal Finance Management - with design v2 .pptx
E-learning Personal Finance Management - with design v2 .pptx
 
Stock Market Brief Deck for 2:22 youtube vid.pdf
Stock Market Brief Deck for 2:22 youtube vid.pdfStock Market Brief Deck for 2:22 youtube vid.pdf
Stock Market Brief Deck for 2:22 youtube vid.pdf
 
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf
20240220 Calibre Q4 and FY 2023 Conference Call Present.pdf
 
Stock Market Brief Deck from youtube 2/16
Stock Market Brief Deck from youtube 2/16Stock Market Brief Deck from youtube 2/16
Stock Market Brief Deck from youtube 2/16
 
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptx
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptxAGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptx
AGRICULTURAL PRICE POLICY -AGRICULTURAL ECONOMICS.pptx
 
Stock Market Brief Deck for 2/21/24 .pdf
Stock Market Brief Deck for 2/21/24 .pdfStock Market Brief Deck for 2/21/24 .pdf
Stock Market Brief Deck for 2/21/24 .pdf
 
2023_-_Master_Trust___GPP_Defaults_Report (2).pdf
2023_-_Master_Trust___GPP_Defaults_Report (2).pdf2023_-_Master_Trust___GPP_Defaults_Report (2).pdf
2023_-_Master_Trust___GPP_Defaults_Report (2).pdf
 
business-model-canvas for collaborative business
business-model-canvas for collaborative businessbusiness-model-canvas for collaborative business
business-model-canvas for collaborative business
 
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...
Indistinguishable from Magic: How the Cybersecurity Market Reached a Trillion...
 
letter to the EF and Premiership about pension liabilities
letter to the EF and Premiership about pension liabilitiesletter to the EF and Premiership about pension liabilities
letter to the EF and Premiership about pension liabilities
 

DDKT-Munich.pdf

  • 1. Incentive separability Pawel Doligalski (Bristol & GRAPE) Piotr Dworczak (Northwestern & GRAPE) Joanna Krysta (Standford) Filip Tokarski (Stanford GSB) This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (ERC Starting grant IMD-101040122) 1
  • 2. Efficiency-equity trade-off: distortions are useful for redistribution • central concept in Public Finance 2
  • 3. Efficiency-equity trade-off: distortions are useful for redistribution • central concept in Public Finance However, sometimes the efficiency-equity trade-off can be avoided: • Diamond-Mirrlees ’71: distortionary taxes on firms are redundant • Atkinson-Stiglitz ’76: differentiated consumption taxes are redundant with a non-linear income tax • Sadka ’76, Seade ’77: there should be “no distortion at the top” We study the general logic behind these results using Mechanism Design 2
  • 4. What we do We consider a general framework allowing for complex incentive constraints - private information, moral hazard, voluntary participation etc We introduce a new notion of incentive separability - A set of decisions is incentive-separable if perturbing these decisions along agents’ indifference curves preserves all incentive constraints. 3
  • 5. What we do We consider a general framework allowing for complex incentive constraints - private information, moral hazard, voluntary participation etc We introduce a new notion of incentive separability - A set of decisions is incentive-separable if perturbing these decisions along agents’ indifference curves preserves all incentive constraints. Main theorem: It is optimal to allow agents to make unrestricted choices over incentive-separable decisions, given some prices and budgets. 3
  • 6. What we do We consider a general framework allowing for complex incentive constraints - private information, moral hazard, voluntary participation etc We introduce a new notion of incentive separability - A set of decisions is incentive-separable if perturbing these decisions along agents’ indifference curves preserves all incentive constraints. Main theorem: It is optimal to allow agents to make unrestricted choices over incentive-separable decisions, given some prices and budgets. The theorem allows us to • Extend and unify Atkinson-Stiglitz and Diamond-Mirrlees theorems - optimal to jointly remove consumption and production distortions - harmonizing consumption taxes with distorted production—a bad idea • Propose a novel argument for the optimality of food vouchers 3
  • 7. Literature review • Redundancy (or not) of commodity taxes with nonlinear income tax Atkinson & Stiglitz (1976); Christiansen (1981); Cremer, Gahvari & Ladoux (1998); Laroque (2005); Kaplow (2006); Gauthier & Laroque (2009); Cremer & Gahvari (1995); Cremer, Pestieau & Rochet (2001); Saez (2002); da Costa & Werning (2002); Golosov, Kocherlakota & Tsyvinski (2003) • Optimal in-kind redistribution Nichols & Zeckhauser (1982); Currie & Gahvari (2008); Condorelli (2012); Akbarpour r ○ Dworczak r ○ Kominers (2023) • Optimality of production efficiency Diamond & Mirrlees (1971); Stiglitz & Dasgupta (1971), Naito (1999), Hammond (2000) 4
  • 8. Outline of the talk 1. Framework 2. General results 3. Applications 5
  • 10. • A unit mass of agents with types θ ∈ [0, 1] ≡ Θ, uniformly distributed • Agents’ utility over a vector of K decisions x ∈ RK + U (x, θ) - x contains e.g. consumption of goods, labor supply, effort, ... - U continuous in x, measurable in θ - we make no ”single crossing” assumptions 6
  • 11. • A unit mass of agents with types θ ∈ [0, 1] ≡ Θ, uniformly distributed • Agents’ utility over a vector of K decisions x ∈ RK + U (x, θ) - x contains e.g. consumption of goods, labor supply, effort, ... - U continuous in x, measurable in θ - we make no ”single crossing” assumptions • An allocation rule x : Θ → RK + • associated aggregate allocation: x = ´ x(θ) dθ • associated utility profile Ux , where Ux (θ) = U(x(θ), θ) 6
  • 12. • Social planner with objective function W (Ux , x), - W continuous and weakly decreasing in x - dependence on Ux : redistributive (welfarist) objective - dependence on x: preferences over allocations beyond agents utilities - e.g. (opportunity) cost of resources, preference for tax revenue, aggregate constraints 7
  • 13. • Social planner with objective function W (Ux , x), - W continuous and weakly decreasing in x - dependence on Ux : redistributive (welfarist) objective - dependence on x: preferences over allocations beyond agents utilities - e.g. (opportunity) cost of resources, preference for tax revenue, aggregate constraints • Planner chooses allocation x : Θ → RK + subject to incentive constraints x ∈ I ⊆ (RK +)Θ - no a priori restrictions on I - I can represent IC constraints due to private info, moral hazard, etc. • An allocation x ∈ I is called feasible 7
  • 14. • Social planner with objective function W (Ux , x), - W continuous and weakly decreasing in x - dependence on Ux : redistributive (welfarist) objective - dependence on x: preferences over allocations beyond agents utilities - e.g. (opportunity) cost of resources, preference for tax revenue, aggregate constraints • Planner chooses allocation x : Θ → RK + subject to incentive constraints x ∈ I ⊆ (RK +)Θ - no a priori restrictions on I - I can represent IC constraints due to private info, moral hazard, etc. • An allocation x ∈ I is called feasible • Notation: For a subset of decisions S ⊆ {1, ..., K} : xS = (xi )i∈S x−S = (xi )i̸∈S x = (xS , x−S ) 7
  • 15. Incentive separability Def: Decisions S ⊆ {1, ..., K} are incentive-separable (at feasible alloc x0) if {(xS , x−S 0 ) : U(xS (θ), x−S 0 (θ), θ) = U(x0(θ), θ), ∀θ ∈ Θ} ⊆ I. Decisions S are incentive-separable (IS) if all perturbations of xS that keep all types indifferent satisfy incentive constraints 8
  • 16. Incentive separability Def: Decisions S ⊆ {1, ..., K} are incentive-separable (at feasible alloc x0) if {(xS , x−S 0 ) : U(xS (θ), x−S 0 (θ), θ) = U(x0(θ), θ), ∀θ ∈ Θ} ⊆ I. Decisions S are incentive-separable (IS) if all perturbations of xS that keep all types indifferent satisfy incentive constraints IS is a joint restriction on preferences and incentives. Examples: 1. Voluntary participation: all decisions are IS 2. Private information: decisions S are weakly-separable ⇒ S are IS - S are weakly-separable when U(x, θ) = Ũ(v(xS ), x−S , θ) 3. Moral hazard: hidden action a affects the distribution of observed state ω - U(x, θ) = −c(a) + P ω P(ω|a)uω(yω), yω ∈ RL + - Decisions within each state ω (namely, yω) are IS 4. Easy to extend to combinations of frictions, dynamic private info, aggregate states, ... 8
  • 17. Preliminaries Consider a feasible allocation x0. We will keep x−S 0 and Ux0 fixed, and re-optimize over the allocation of incentive-separable goods xS : Uθ(xS (θ)) := U(xS (θ), x−S 0 (θ), θ) R(xS ) := W (Ux0 , (xS , x−S 0 )) Assumptions: 1. Uθ is locally nonsatiated, for all θ 2. U(xS 0 (θ)) ≥ U(0), for all θ 3. For any x0, there exists integrable x̄(θ) such that Uθ(y) = Uθ(xS 0 (θ)) =⇒ y ≤ x̄(θ) 9
  • 19. Suppose that decisions S ⊆ {1, ..., K} are incentive-separable Def: A feasible allocation x0 is S-undistorted if xS 0 solves max xS R(xS ) s.t. Uθ(xS (θ)) = Uθ(xS 0 (θ)), ∀θ ∈ Θ 10
  • 20. Suppose that decisions S ⊆ {1, ..., K} are incentive-separable Def: A feasible allocation x0 is S-undistorted if xS 0 solves max xS R(xS ) s.t. Uθ(xS (θ)) = Uθ(xS 0 (θ)), ∀θ ∈ Θ Lemma 1 (Optimality) Any feasible x0 that is not S-undistorted can be improved upon. 10
  • 21. Suppose that decisions S ⊆ {1, ..., K} are incentive-separable Def: A feasible allocation x0 is S-undistorted if xS 0 solves max xS R(xS ) s.t. Uθ(xS (θ)) = Uθ(xS 0 (θ)), ∀θ ∈ Θ Lemma 1 (Optimality) Any feasible x0 that is not S-undistorted can be improved upon. Proof: Since x0 is not S-undistorted, there exists allocation x∗ = (xS ∗ , x−S 0 ) s.t. 1. Utility levels are unaffected: Uθ(xS ∗ (θ)) = Uθ(xS 0 (θ)), 2. Planner’s objective strictly improves under x∗ relative to x0. Since goods S are incentive-separable, 1. implies that x∗ is also feasible. □ 10
  • 22. Simplifying Assumption: R(xS ) is linear in xS with coefficients −λ ≪ 0 - Interpretation: constant marginal costs of producing xS - Too strong, we will relax it in a moment 11
  • 23. Simplifying Assumption: R(xS ) is linear in xS with coefficients −λ ≪ 0 - Interpretation: constant marginal costs of producing xS - Too strong, we will relax it in a moment Lemma 2 (Decentralization) Any feasible S−undistorted allocation x0 can be decentralized with prices λ. That is, there exists a budget assignment m : Θ → R+ such that, for all θ ∈ Θ, xS 0 (θ) solves max xS ∈R |S| + Uθ(xS ) subject to λ · xS ≤ m(θ). 11
  • 24. Simplifying Assumption: R(xS ) is linear in xS with coefficients −λ ≪ 0 - Interpretation: constant marginal costs of producing xS - Too strong, we will relax it in a moment Lemma 2 (Decentralization) Any feasible S−undistorted allocation x0 can be decentralized with prices λ. That is, there exists a budget assignment m : Θ → R+ such that, for all θ ∈ Θ, xS 0 (θ) solves max xS ∈R |S| + Uθ(xS ) subject to λ · xS ≤ m(θ). Proof: x0 is S-undistorted if xS 0 solves max xS −λ · xS s.t. Uθ(xS (θ)) = Uθ(xS 0 (θ)), ∀θ ∈ Θ It is a collection of expenditure minimization problems. The claim follows from consumer duality (Mas-Colell et al. 1995, Prop 3.E.1). □ 11
  • 25. Main theorem (simplified) Combining Lemmas 1 and 2, we arrive at: Theorem Starting at any feasible allocation x0, the planner’s objective can be (weakly) improved by allowing agents to purchase incentive-separable goods at prices λ subject to type-dependent budgets. • There should be no distortions between IS goods • Agents can trade IS goods freely given prices and budgets • With production, prices of IS goods should be equal to marginal costs • Result silent about non-IS goods 12
  • 26. Main theorem (general) Assumption (R has bounded marginals) There exist scalars c̄ > c > 0 such that, for all y ∈ RK , k ∈ {1, ..., K}, ϵ > 0 c̄ ≥ |R(y + ϵek ) − R(y)| ϵ ≥ c. Theorem 1 Starting at any feasible allocation x0, the planner’s objective can be (weakly) improved by allowing agents to purchase incentive-separable goods at undistorted prices subject to type-dependent budgets. • Undistorted prices: prices that decentralize an S-undistorted allocation • R has bounded marginals → the planner will always select an allocation implying strictly positive prices 13
  • 27. Main theorem (general) Assumption (R has bounded marginals) There exist scalars c̄ > c > 0 such that, for all y ∈ RK , k ∈ {1, ..., K}, ϵ > 0 c̄ ≥ |R(y + ϵek ) − R(y)| ϵ ≥ c. Theorem 1 Starting at any feasible allocation x0, the planner’s objective can be (weakly) improved by allowing agents to purchase incentive-separable goods at undistorted prices subject to type-dependent budgets. • Undistorted prices: prices that decentralize an S-undistorted allocation • R has bounded marginals → the planner will always select an allocation implying strictly positive prices The results can be easily extended to allow for: • the set of IS goods to depend on type θ and initial allocation x0 • additional constraints faced by the planner: x ∈ F 13
  • 29. APPLICATION 1: ATKINSON-STIGLITZ Atkinson-Stiglitz model: - θ: privately observed productivity - c ∈ RK−1 + : vector of consumption of goods - produced with const marginal costs λ - L: labor in efficiency units - consumption is weakly separable: U(v(c), L, θ) → consumption goods c are incentive-separable 14
  • 30. APPLICATION 1: ATKINSON-STIGLITZ Atkinson-Stiglitz model: - θ: privately observed productivity - c ∈ RK−1 + : vector of consumption of goods - produced with const marginal costs λ - L: labor in efficiency units - consumption is weakly separable: U(v(c), L, θ) → consumption goods c are incentive-separable - Planner’s objective is (α is the marginal value of public funds): W (Ux , x) = V (Ux ) + α(L − K−1 X k=1 λk ck ). 14
  • 31. APPLICATION 1: ATKINSON-STIGLITZ Atkinson-Stiglitz model: - θ: privately observed productivity - c ∈ RK−1 + : vector of consumption of goods - produced with const marginal costs λ - L: labor in efficiency units - consumption is weakly separable: U(v(c), L, θ) → consumption goods c are incentive-separable - Planner’s objective is (α is the marginal value of public funds): W (Ux , x) = V (Ux ) + α(L − K−1 X k=1 λk ck ). By Theorem 1, it is welfare-improving to let agents trade goods c freely given prices proportional to marginal costs and type-dependent budgets Theorem 1 implies the Atkinson-Stiglitz theorem: - prices proportional to marginal costs → no distortionary taxes on c - type-dependent budgets → implemented with non-linear income tax 14
  • 32. APPLICATION 1: ATKINSON-STIGLITZ Theorem 1 is much more general than the Atkinson-Stiglitz theorem: 1. We allow for more complex settings and incentives - only a subset of commodities may be incentive separable - there may be moral hazard (or other incentive problems) - combine redistributive (Mirrlees ’71) and social-insurance (Varian ’80) strands of optimal income tax literature 2. We rule out any distortionary mechanisms - e.g. quotas, public provision,... 3. What is essential is the ability to implement budgets for S goods, rather than the availability of nonlinear income tax - Theorem 1 applies even if earnings are unobserved (e.g. due to tax evasion) but total expenditure on IS goods is observed 15
  • 33. APPLICATION 2: FOOD VOUCHERS Let E ⊆ {1, ..., K} denote consumption of various food items, −E : other commodities and decisions Jensen & Miller (2010): idiosyncratic food tastes more pronounced when food consumption is high. 16
  • 34. APPLICATION 2: FOOD VOUCHERS Let E ⊆ {1, ..., K} denote consumption of various food items, −E : other commodities and decisions Jensen & Miller (2010): idiosyncratic food tastes more pronounced when food consumption is high. • Let v(xE ) denote the nutritional value of food (common to all agents) • Utility function U(x, θ) =    UL(v(xE ), x−E , θ) if v(xE ) ≤ v (only nutrition matters) UH (x, θ) if v(xE ) > v (nutrition and tastes matter) - v: nutritional threshold above which idiosync. food tastes become active 16
  • 35. APPLICATION 2: FOOD VOUCHERS Let E ⊆ {1, ..., K} denote consumption of various food items, −E : other commodities and decisions Jensen & Miller (2010): idiosyncratic food tastes more pronounced when food consumption is high. • Let v(xE ) denote the nutritional value of food (common to all agents) • Utility function U(x, θ) =    UL(v(xE ), x−E , θ) if v(xE ) ≤ v (only nutrition matters) UH (x, θ) if v(xE ) > v (nutrition and tastes matter) - v: nutritional threshold above which idiosync. food tastes become active • θ is privately observed → food items are IS for types θ s.t. v(xE (θ)) ≤ v • Food items produced with const marginal cost 16
  • 36. APPLICATION 2: FOOD VOUCHERS Conflicting motives for taxes on food: - food choices of the poor are IS → suboptimal to distort - the rich exhibit heterogeneous tastes for food → optimal to distort their food choices (Saez, 2002) 17
  • 37. APPLICATION 2: FOOD VOUCHERS Conflicting motives for taxes on food: - food choices of the poor are IS → suboptimal to distort - the rich exhibit heterogeneous tastes for food → optimal to distort their food choices (Saez, 2002) Solution? A food vouchers program: Corollary 1 Consider a feasible allocation x0. The planner’s objective can be improved by assigning budgets to all agents with θ such that v(xE (θ)) ≤ v, and letting them spend these budgets on food items E (but no other goods) priced at marginal cost. - means-tested: allocated only to ”poor” types {θ : v(xE (θ)) ≤ v} - allows to buy food at producer prices: isolates from tax distortions 17
  • 38. APPLICATION 2: FOOD VOUCHERS Conflicting motives for taxes on food: - food choices of the poor are IS → suboptimal to distort - the rich exhibit heterogeneous tastes for food → optimal to distort their food choices (Saez, 2002) Solution? A food vouchers program: Corollary 1 Consider a feasible allocation x0. The planner’s objective can be improved by assigning budgets to all agents with θ such that v(xE (θ)) ≤ v, and letting them spend these budgets on food items E (but no other goods) priced at marginal cost. - means-tested: allocated only to ”poor” types {θ : v(xE (θ)) ≤ v} - allows to buy food at producer prices: isolates from tax distortions Compare to the U.S. food stamps program (SNAP) - means-tested, purchases exempt from commodity taxes (state and local) - however, restricted eligibility of unemployed—suboptimal in our setting 17
  • 39. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES x: allocation of consumption goods and other households decisions 18
  • 40. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES x: allocation of consumption goods and other households decisions Production sector consists of J firms. For a firm j ∈ {1, ..., J}: • zj ∈ RK is the production vector (negative entries represent inputs) • Zj is the set of possible production vectors z = (z1, ..., zJ ) is a production plan, z = PJ j=1 zj is aggregate production vector Z is the Minkowski sum of Zj over j 18
  • 41. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES x: allocation of consumption goods and other households decisions Production sector consists of J firms. For a firm j ∈ {1, ..., J}: • zj ∈ RK is the production vector (negative entries represent inputs) • Zj is the set of possible production vectors z = (z1, ..., zJ ) is a production plan, z = PJ j=1 zj is aggregate production vector Z is the Minkowski sum of Zj over j Allocation (x, z) is feasible if x ∈ I, z ∈ Z and z ≥ x The planner maximizes W (Ux , z − x), non-decreasing in z − x 18
  • 42. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES x: allocation of consumption goods and other households decisions Production sector consists of J firms. For a firm j ∈ {1, ..., J}: • zj ∈ RK is the production vector (negative entries represent inputs) • Zj is the set of possible production vectors z = (z1, ..., zJ ) is a production plan, z = PJ j=1 zj is aggregate production vector Z is the Minkowski sum of Zj over j Allocation (x, z) is feasible if x ∈ I, z ∈ Z and z ≥ x The planner maximizes W (Ux , z − x), non-decreasing in z − x A feasible production plan z0 is efficient if there does not exist z ∈ Z such that z ≥ z0 and z ̸= z0 18
  • 43. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES Corollary 2 For any feasible allocation, the planner’s objective can be (weakly) improved by choosing an S-undistorted allocation of incentive-separable goods and an efficient production plan. Key insight: production decisions are (trivially) incentive-separable. 19
  • 44. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES Corollary 2 For any feasible allocation, the planner’s objective can be (weakly) improved by choosing an S-undistorted allocation of incentive-separable goods and an efficient production plan. Key insight: production decisions are (trivially) incentive-separable. To ensure positive decentralizing prices, we need additional assumptions on feasible production plans: Assumption The aggregate production set Z is closed, bounded from above and convex. Furthermore, for any z0 ∈ Z and any nonempty proper subset A ⊂ {1, ..., K}, there exists z ∈ Z such that zA ≤ zA 0 , z−A ≥ z−A 0 and z−A ̸= z−A 0 . 19
  • 45. APPLICATION 3: ATKINSON-STIGLITZ MEET DIAMOND-MIRRLEES Theorem 2 For any feasible allocation, there exists a price vector λ ∈ RK ++ such that the planner’s objective can be (weakly) improved by simultaneously: (i) allowing agents to purchase incentive-separable goods at prices λ subject to type-dependent budgets; (ii) allowing firms to maximize profits and trade all goods taking prices λ as given, and taxing their profits lump-sum. 1. Beneficial to jointly remove distortions in production and IS-consumption 2. Achieved by allowing hhs and firms to trade goods at common prices However, harmonizing consumption taxes w/out addressing production distortions could reduce welfare: • with a production distortions, the producer prices are distorted 20
  • 46. CONCLUSIONS • Incentive separability: a useful notion to study optimality and decentralization in complex incentive problems • We focused on applications within Public Finance - Atkinson-Stiglitz theorem - food stamps - Atkinson-Stiglitz meet Diamond-Mirrlees • Potential applications in other contexts - employee compensation / principal-agent problems within organizations - monopoly pricing with multiple goods / attributes - ... 21
  • 48. Examples of incentive-separable decisions IS decisions: {(xS , x−S 0 ) : U(xS (θ), x−S 0 (θ), θ) = U(x0(θ), θ), ∀θ ∈ Θ} ⊆ I. 1. Voluntary participation: I = { x : U(x(θ), θ) ≥ U(θ), ∀θ ∈ Θ } ; All decisions are incentive-separable. 2. Private information: I = x : U(x(θ), θ) ≥ max θ′∈Θ U(x(θ′ ), θ), ∀θ ∈ Θ ; Weak separability of decisions S, U(x(θ), θ) = e U (v(xS (θ)), x−S (θ), θ), ∀θ ∈ Θ, implies incentive separability of S. 22
  • 49. Examples of incentive-separable decisions 3. Moral hazard: Hidden action a ∈ A affects the distribution of observed state ω: U((a, y), (τ, ω)) = X ω u(y; ω)P(ω|a, τ) − c(a; τ), y ∈ RL +; Set I: Agent with type τ reports τ truthfully, takes recommended action a(τ), and consumes y(τ, ω). With θ = (τ, ω) and x = (a, y), decisions y are incentive-separable. 4. Easy to extend to combinations of incentive constraints, dynamic private information, verifiable information, labels, ... 23