seigs, gpis, and bcas: playing to different stakeholders dennis coates and scott farrow department...

31
SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1 www.umbc.ed u

Upload: maurice-thornton

Post on 11-Jan-2016

220 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

1

SEIGs, GPIs, and BCAs:

Playing to Different Stakeholders

Dennis Coates and Scott FarrowDepartment of Economics

UMBC

www.umbc.edu

Page 2: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

2

Outline

• SEIG• Regional Issues• BCA Issues• Conclusion

Page 3: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

3

SEIG is Problematic

• Criticisms not new – Walker (2008)• Impact themes overlap

– Double counting– Dubious categorization of costs and benefits

• Impact theme relative importance unclear– Politically determined– Not transparent

Page 4: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

4

Examples

• Personal entertainment– Health and Well-Being – Economic and Financial – Recreation and Tourism

• Contribution to economic growth– Economic and Financial: Contribution to GDP– Economic and Financial: Changes in investment, etc– Employment and Education: Direct and indirect job

creation– Employment and Education: Annual and hourly wages

Page 5: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

5

SEIG is Problematic 2

• Who are the stakeholders/decision-makers?– General Public– Gambling “Interests”– Policy makers and politicians

• Unstructured information may be better than none at all– Unbiased/disinterested research– Value judgments left to stakeholders

Page 6: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

6

SEIG is ProblematicSummary

• Flexibility or All things to all people• Adding Lemons and Cherries?• Conflating Positive with Normative

Page 7: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

7

Focus on two topics:

• Regional analysis– Unit of Analysis– Spillovers– Economic Development– Interjurisdictional competition

• BCA– Impacts or markets – Surplus or risk loving uncertainty– Problem or pathological gamblers– Equity

Page 8: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

8

Issues in Regional Analysis

• Unit of Analysis• Interjurisdictional Competition• Spillovers• Economic Development

Page 9: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

9

Unit of Analysis

• Anielski and Bratten SEIG– Four possibilities

• Individual, Household• Community – neighborhood, town or city; clubs or

interest groups; may be geographic but need not be• Regions – “larger geographic areas than communities”• Province

– Two Unanswered Questions• What is the political jurisdiction for the analysis?• Given the political jurisdiction, how wide-spread is access

to gambling?

Page 10: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

10

Spillovers/ExternalitiesWhat is the jurisdiction of the analysis?

• Vaillancourt and Roy identify geographical dimension as one of two significant methodological issues – rationale for Canada as unit of analysis

• Smaller geographic area, the better for extracting benefits from those other jurisdictions/populations

• Smaller geographic area, the more likely to export costs onto other jurisdictions/populations

Page 11: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

11

Spillovers

• Fiscal Federalism– Multiple jurisdictions– Boundaries set so costs and benefits are captured

• Costs– Gambling– Travel, etc

• Benefits– Gambling– Other

Page 12: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

12

Economic Development

• Job creation– Gambling– Induced

• Income growth– Gambling– Induced

• Tax revenues• Neighborhood revitalization?

Page 13: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

13

Interjurisdictional Competition

• Gambler mobility– Slot machines in MD– More venues in BC , AU

• On the border– Attract gamblers from neighbors– Retain home grown gamblers– Spillovers (MA and CT)

• Race to the Bottom

Page 14: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

14

Selected Issues in GPIG & BCAG

1. GPIG

2. BCAG

– Logic models and Surplus

– Uncertainty

– Problem or pathological gamblers

– Equity

Page 15: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

15

GPIG

• Within SEIG: Several valuation methods mentioned, including GPIG and BCA.

• GPIG: suggestion to use modified National Income and product accounts (e.g. GDP)– History of concern in GDP for adjustments

(household labor, environment (net dep)…– GDP: C+I+G+net exports or income (flow, not

stock/balance sheet)

Page 16: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

16

GPIG: doesn’t resolve issues• GPIG comments (e.g. Anielski, GPINovaScota)

– Not implementing national income accounting : only identify candidates as subset of SEIG

– Double counting: GDP of gambling + personal expenditures– Should personal expenditures be wagers net of payouts as suggested? Aren’t

expenditures the wagers and the payouts income?– Inequality adjustment: more later, how to investigate value judgment?– Usual co-morbidity issues– Property value (NIPA is a flow net of depreciation, property would need

separate balance sheet)– Cost of bankruptcy, lost productivity: these are mediated in part through

markets; expected to be built into wages and risk based interest rates,– Addiction: change in surplus mentioned but actually change in expenditure

(no surplus measures in GDP), more later

Page 17: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

17

Logic Model: SEIG, GPIG, and BCARegion or Province is central actor? What is causing what?

Page 18: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

18

Gambling marketConsumers: CSProducers: PS(WTP: Risk Loving)Government

Revenues and expenditures

Input Markets

Labor MarketGambler’s incomeRegional employment

Financial MarketsBorrowing/lending

External EffectsHouseholdCrimeCommunity

Partial Equilibrium (can extend to general equilibrium) Standing: Whose impacts count…e.g. Provincial

BCA: Links among markets, Gov’t and Externalities

Legal conditions and multiple market interactions

Page 19: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

19

Is Gambling Consumer Well Analyzed?

• Fundamental issue of consumer behavior Gamblers must be risk loving in this activity since the payout is less than “fair”

• Complex links between surplus and Willingness to Pay with uncertainty (latter preferred with uncertainty): – Broader economic literature not resolved (expected and non-

expected utility) people such as two nobel prize winners (Freidman, Markovitz)

• Current candidates: • Consumer “expenditures”, but that is usually that is what is

deducted from total benefits to get consumer surplus• Standard surplus (Grinols…at margin no effect)--• Distance surplus, but no link to change in price (Grinols)• What happened to valuation under uncertainty?

Page 20: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

20

Expenditures and Surplus contrastedFundamentally different measures

Price

P

Demand

Q Quantity

Expenditures (P*Q)

Consumer surplus (Total WTP less expenditures)

Page 21: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

21

Risk Loving, and Paying to Gamble at point of indifference

(Contrasts: Insurance--paying not to gamble)

Gamblers: Z= ex-ante WTP for access to gamble (including travel and house take); if cost is less, a “surplus” but not the standard surplus.

Utility

Z E(U(G))=U(I)

IncomeIE(I)

Page 22: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

22

WTP to gamble and change in location(related to “distance surplus”)

• ZG Gambler: WTP ex-ante for gamble can include– Fixed fee (travel)– House/state take– “surplus” if doesn’t have to pay the full amount– If gambling location gets closer, same WTP but more may go for

gambling and/or surplus, gamble more

• ZNG Non-gambler: may have WTP in a smaller amount that is not sufficient to travel so no observed “gamble”; as locations get closer, then ZNG may be sufficient to gamble; new entry of gambling.

• Price of gamble is extracted from WTP• If offered better gamble (e.g. “looser slots”) at least willing

to pay earlier amount and may get added utility surplus.

Page 23: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

23

Problem or Pathological Gamblers

• Consumer sovereignty or not?– Standard model: people responsible and

understand own actions– Non-standard model (some aspects of Behavioral

economics)• Lack of control (some with WTP to be “normal”), then a

modest surplus loss ( Australian report or Vining and Weimer (JBCA, 2011) based on gambling to P&P gamblers that returns less than expected.

• Other behavioral: could be issues in lack of understanding of probabilities

Page 24: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

24

Surplus Loss from Addiction• (Vining/Weimer/Thomas) smoking, ~25% loss in Total Surplus

– Preliminary oddity: if only use distance surplus as consumer benefit, then addiction adjustment appears small.

Page 25: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

25

Equity and BCA

Increasing attention with new Administration in U.S. and regulatory analysis

• Standard approach in textbooks is to consider distributional weighting using marginal utility of income for different income classes, but lack of agreement on weights– Sensitivity: default in BCA is equal MU(income) which is an

Atkinson weight of 0.– As an assumption, easily varied, perhaps to find cross-over point– Some guidance; UK Greenbook for regulatory analysis specifies

Atkinson weight of 1; US Census reports .25, .5, .75– Example from Analysis of VLTs in Maryland (Farrow and Shinogle)

Page 26: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

26

Distributional weights based on values of inequality aversion, e, used by the U.S. Census Bureau and the UK Treasury

Population Quintile, Median, %

Mean US HH Income by Quintile: 2007

Default: e=0 e=.25 e=.5 e=.75 e=1

0-20 $11,551 1 1.4 2.1 3.0 4.320-40 $29,442 1 1.1 1.3 1.5 1.7Median $50,233 1 1.0 1.0 1.0 1.040-60 $49,968 1 1.0 1.0 1.0 1.060-80 $79,111 1 0.9 0.8 0.7 0.680-100 $167,971 1 0.7 0.5 0.4 0.3 D

Data source: US Census, 2008b; author’s calculations

Page 27: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

27

Short example: VLT’s in MarylandBase case: positive (but uncertain) Net Benefits

Video Lottery Terminals Basic Model I: direct effects only

Basic Model I Mil 2008 $ Mil 2008 $

Benefits MIPAR

Delta CS: Consumer distance $25 Delta CS: Consumer distance $25

Delta Gov't Revenue $913 Delta Gross Gov't Revenue $913Delta G: Annaul fee for Prob. Gamb $6 $6

Delta PS: MD Profits $36 Delta PS: MD Profits $36

New sales tax $2.5

Unemployment effects $0

Welfare benefits $980 Modified Benefits $982

Costs

Delta Gov Rev (2% Admin) $27 Delta Gov Rev (2% Admin) $27

Delta Gov Rev: other cost $48 Delta Gov Rev: other cost $48

External costs $428 External costs $428

Loss in lottery sales 57

Loss in other taxes 34

Change other CS or PS 0

Welfare costs $503 Modified Costs $594

Annual Net Benefits $477 Modified Net Ben $389

Specific Secondary effects

Page 28: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

28

Distributional Weights: Can change sign

e=.5 e=.25

~4:1 2:1 weight

w/Distributional impact:gambler $43 $31

$913 $913

$6 $6

$36 $36

$2 $2

$0 $0

$1,000 $989

$27 $27

$48 $48

$428 $428

$57 $57

$34 $34

w/Distributional impact:gambler 971 345

$1,565 $938

-$565 $50

Extended model with Distributional effects

Page 29: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

29

Summary What can be learned from gambling analyses

• SEIGs and GPIGs: potential impacts of interest to various stakeholders

• Regional Issues: spillovers, race to the bottom• BCA issues: framework exists but lacks central

WTP measure based on gambling; can include addiction and equity.

Page 30: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

30

Extra slides

Page 31: SEIGs, GPIs, and BCAs: Playing to Different Stakeholders Dennis Coates and Scott Farrow Department of Economics UMBC 1

31

Simulation results(using @Risk with Excel)

5.0% 90.0% 5.0%

75 521

-200-1000100200300400500600700

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

Modified Net Ben / Extended model with Parameter Uncertainty

Modified Net Ben / Extended model with Parameter Uncertainty

Minimum -153.0188Maximum 688.3253Mean 308.9715Std Dev 135.0578Values 10000

@RISK for ExcelPalisade Corporation

0.76

-0.60

0.10

0.10

0.09

-0.8-0.6-0.4-0.20.00.20.40.60.8

Coefficient Value

75 to 20 miles-Grinols

proportion of unemploy hired / Number of term

Profit percent / Number of term

External costs / Extended model with Parameter Uncertainty

Delta Gross Gov't Revenue / Extended model with Parameter Uncertainty

Modified Net Ben / Extended model with Parameter UncertaintyCorrelation Coefficients (Spearman Rank)

@RISK for ExcelPalisade Corporation