taxation in kind with special attention to military manpower conscription by casey b. mulligan

33
Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Upload: ashton-sartell

Post on 14-Dec-2015

215 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Taxation in Kind

with special attention to

Military Manpower Conscription

by Casey B. Mulligan

Page 2: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Taxation in Kind• examples

– conscription of manpower by the military or civil service– conscription of materials and factories by the military– eminent domain (land, intellectual property)– jury duty– human organs– nationalization of industry– regulatory taking?

• how do these taxes affect the economy?• what is the optimal use of in-kind taxes?• what is the political economy of in-kind taxes?• despite work on cash vs. in-kind transfers, practically no

work on in-kind taxes– exception: Vietnam era literature on volunteer vs drafted army and

more recent law & econ work on eminent domain

Page 3: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Efficient Outcomes with Taxation In Kind: An Example

• military manpower recruitment model– Becker (1957), Sjaastad & Hansen (1970), Warner & Asch

(2001)– m soldiers needed– young men differ in terms of their opportunity cost r, which is

distributed G– r is private information– efficient allocation recruits only those with r < G-1(m)

• two recruitment systems– use monetary tax revenue to fund volunteer army with market

clearing wage G-1(m)– draft all young men, and levy a fine in the amount G-1(m) on

those who do not report for duty– both obtain efficient recruitment (in the model)– the volunteer system spends revenue, the draft obtains it

Page 4: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Two approaches• prohibit sorting mechanisms (in the theory)

– also assume that conscripts are unpaid– volunteer system is (by assumption) more selective– in-kind taxation (by assumption) spends less revenue– Warner, Ashe, Johnson, Barro, Oi, Friedman, Shavell

• my approach– in-kind taxation is selective, but must spend revenue in order

to maintain incentives.– without incentives, taxpayers take actions to avoid the in-

kind tax private property rights create public sector value (see also Innes 2000 and Stroup 1997 on public land use)

• Vietnam era draft dodging– join a protected occupation, enroll in college, have children– move to Canada– gain weight, embellish medical conditions– obtain political favors

Page 5: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Warner-Ashe-Johnson-Barro-Oi-Friedman-etc argument: selectivity vs. monetary tax dwc

• volunteer army offers selection and longevity, but incurs the dwcs of tax collection– volunteer salaries are a transfer from civilians to military,

with a deadweight cost in the process• dwc of salaries is on top of dwcs of other government spending• if there are conscripts, don’t pay them (just adds to dwc)• dwcs are convex –> most important for large armies

– volunteer army selects the people most willing to serve– volunteer army turns over less, which is important due to the

expense of military training

• this argument also appears in the legal literature on eminent domain (e.g., Shavell)

Page 6: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Selectivity vs dwc Argument

m = personnel per capita

10

S(m)

0

1( )

mS x dx

m

( ( ))S m

(convex) deadweight cost of military payroll

(approx. linear) gain from market selection

avg social cost of random selection

Page 7: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Figure 3. The Policy Possibility Frontier

Revenue

The Figure displays two policy possibility frontiers, each corresponding to different model parameters. The circle indicates the combination of revenue and allocative inefficiency available under all-volunteer recruitment. The square indicates the combination available with a random in-kind tax.

0

allocative inefficiency

RV

I

IR

I

IR

random in-kind tax

government indifference

curve

Page 8: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

In-Kind vs Monetary Taxation as Alternative Means of Achieving Efficient Sorting

• in-kind taxation can achieve efficient sorting– regulation– bribes– permits and licenses– buyout/commutation and replacement options– Mulligan and Shleifer find such mechanisms used around the world for

military conscription. Commutation also used for U.S. recruitment during the Revolutionary War, and in the Civil War (both sides)

• Vietnam era sorting, despite the draft– selected young, males– potentially strong sorting. among young males: 75+% of draft eligible did

not serve, 10% of ineligible did serve– drafted military arguable more white, less schooled

• modeling the dwcs of in-kind taxation– some resources are un-fit for public use. e.g., obese soldiers, land that is not

a desirable habitat for endanger species– resources can be converted from unfit to fit (at cost ) or vice versa (at cost

). Initial distribution is fit and (1-) unfit– private value r distributed G– m units needed for public use

Page 9: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Efficient Sorting

• public sector selects low private value resources– among those selected, unfit are converted to fit– fit threshold is higher by the amount of the conversion cost

(i.e., more fit are selected for public use)– no fit are converted to unfit

( ) (1 ) ( )

( , , , ( ))f f

f

m G r G r

r S m G

Page 10: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Figure 1. Efficient Allocation to Public Use

cum. density

The Figure illustrates the efficient allocation of fit and unfit resources to public use. G(r) is the distribution of private values, which is the same for both fit and unfit. is the cost of converting unfit resources to fit.

0 r = private valueru = rf - rf

G(r)

Area = fraction of unfit resourcesconverted to fit

and used publicly

Combined area =fraction of unfit resources

retained for private use

Area =fraction of fit resourcesretained for private useCombined area =

fraction of fit resourcesused publicly

Page 11: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Figure 2. Aggregate Marginal Opportunity Cost Curve

r

The Figure illustrates the aggregate opportunity cost S(m) of public use, assuming efficient sorting into the public sector. G(r) is the distribution of private values, which is the same for both fit and unfit. and are the costs of converting unfit resources to fit and vice-versa, respectively.

0m = public use

G-1(m/)

G()

1

S(m)

G()

least incentives

needed

prevent destruction encourage conversion

Page 12: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

In-Kind vs Monetary Taxation as Alternative Means of Achieving Efficient Sorting

• “all-volunteer” system: offer a wage to all fit resources that volunteer• in-kind tax system: take all resources. Let any owner “buy out” at bf and

bu. Compensate those that remain, at different rates for fit and unfit• in-kind taxation obtains revenue RI. All-volunteer system obtains revenue

RV, where is a per-interaction administrative cost

• both systems obtain the same sortingconscripts must be compensated enough and the buyout rate be sufficiently

low to obtain efficient supply[see Table 1]

• in-kind obtains more (spends less) revenue when:– m is large or I is small or are large; is small– private values are large positive theory of taxation in kind based on factor supply conditions

• transaction/administrative costs are boring, but critical when considering the efficient assignment of property rights

( , , , ( ))V VR S m G m

Page 13: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Table 1: In-Kind Tax Parameters that Induce Efficient Sorting

m < G() m [G(),G()] m > G()

conscript wage, unfit 0 0 0

conscript wage, fit = w 0 S(m) -

buyout price, unfit = bu 0 0 S(m) -

buyout price, fit = bf S(m) S(m) -

revenue = RI(m) (-m)S(m) - I - mS(m) - I (1-m)S(m) - - I

revenue advantage = RI(m) - RV(m)

S(m) + Vm - I + Vm - I S(m) - + Vm - I

Page 14: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Table 1: In-Kind Tax Parameters that Induce Efficient Sorting

Unfit: need incentive to return to pvt sector, or incentive to become fit

m < G() m [G(),G()] m > G()

conscript wage, unfit 0 0 0

conscript wage, fit = w 0 S(m) -

buyout price, unfit= bu 0 0 S(m) -

buyout price, fit = bf S(m) S(m) -

revenue = RI(m) (-m)S(m) - I - mS(m) - I (1-m)S(m) - - I

revenue advantage = RI(m) - RV(m)

S(m) + Vm - I + Vm - I S(m) - + Vm - I

Page 15: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

m < G() m [G(),G()] m > G()

conscript wage, unfit 0 0 0

conscript wage, fit = w 0 S(m) -

buyout price, unfit = bu 0 0 S(m) -

buyout price, fit = bf S(m) S(m) -

revenue = RI(m) (-m)S(m) - I - mS(m) - I (1-m)S(m) - - I

revenue advantage = RI(m) - RV(m)

S(m) + Vm - I + Vm - I S(m) - + Vm - I

Table 1: In-Kind Tax Parameters that Induce Efficient Sorting

Fit: need incentive to return to remain fit, given that unfit are let free

Page 16: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Figure 3. The Policy Possibility Frontier

Revenue

The Figure displays two policy possibility frontiers, each corresponding to different model parameters. The circle indicates the combination of revenue and allocative inefficiency available under all-volunteer recruitment.

0

allocative inefficiency

RV

I

IR

I

IR

Page 17: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Public Project Size Encourages In-Kind Taxation• volunteer expenditure increases with m • in-kind expenditure is not monotonic, but necessarily increases

less with m than it does for the volunteer system– administrative cost is marginal for volunteer, fixed for in-kind– supply price is a revenue inflow for the in-kind tax, outflow for the

volunteer system

• in-kind taxation likely inferior at m = 0, superior at m = 1, and the two systems are equivalent at some critical m between 0 and 1

• other reasons for size effects– political/legal scale economies (Mulligan-Shleifer)– insurance scale economies– alleviate deadweight costs of monopsony (later)– relative composition of volunteer and drafted resources (1970s literature)

Page 18: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Efficient Project Size

• U(m) is the social benefit of the project is the extra weight the government puts on revenue

– set = 0 to calculate first-best efficient– set = marginal dwc of monetary taxes for second-best efficient– other ’s can be interpreted politically

• given m, choose the system with the most revenue, regardless of the magnitude of monetary dwcs actually encourage the volunteer system for small mmonetary dwcs encourage in-kind taxation for large m the proper empirical specification interacts with m

( ) ( ) ( )

( ) ( ) ( )

V V V V V

I I I I

R m U m S m

R m U m S m

Page 19: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

The Scope and Conduct of Taxation in Kind• hold out

– relative to labor and materials, land is more efficiently taken because labor and materials are not subject to hold out: they are usually supplied in a competitive resale market, or can be produced from raw materials

• gender, region, and occupation– local projects will have strictly limits on the buyout rate and

conscript compensation in order to achieve efficient sorting national projects more likely to tax in kind

• unskilled occupations more likely to be subject to taxation in-kind

• women less likely to be drafted• cross-country data on military conscription

Page 20: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Observations on Conscription Use and the Deadweight Costs of Monetary Taxes

• yes, the size of the force (per capita) predicts the use of conscription– across countries– USSR fall followed by European shifts to voluntary army– but, the force effects come out of many theories

• proxies for dwc of taxes do not seem to be correlated with conscription use– amount of nonmilitary spending– nonmilitary “demand” shifters (e.g., elderly)– GDP per capita (i.e., development leads to more efficient

taxes, but not less conscription)– why are conscripts paid?– why aren’t more occupations drafted?

• my approach answers these questions, and replaces the simple correlation prediction with an interaction term

Page 21: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Regulatory Complementarity and Increasing Returns• regulations have some costs that are fixed in the sense

that they grow less-than-proportionally with the population to be regulated– adoption: testing the law, reaching a political consensus– administration, enforcement

• some regulatory resources are general (i.e., apply to many regulatory areas)– mobilized interest groups– networks of compliance officials

• costs of conscription– “The costs to government of implementing and using replacement decreased

in some ways over time, as central government enlarged its bureaucratic and coercive reach.”

– “The habit of regular and thorough communication between different levels of administration.”

legal origin and population should have similar “effects”

Page 22: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Types of Military Manpower Systems

• all-volunteer (since 1970, about 1/3 of countries)– hereafter, “volunteer”– 0 countries in our sample are all-conscript– lowest %s volunteer are Senegal (5%), Switzerland

(9%), and Turkey (15%)

• universal or random conscription (20%)– entire population of able young men– no exemptions for college, special occupation, etc.

• conscription with exemptions or deferrals (30%)• conscription with buyouts or substitutes (8%)• [impressment]

Page 23: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Countries with Buyouts (c. 1996)

• Albania, Argentina, Bolivia, China, Ecuador, Egypt, Iran, Iraq, Jordan, Paraguay, Syria, Tunisia, Turkey have (probably) legal buyouts

• e.g., Iran– “PhDs and men with BAs who left Iran before March 1990

may get exempted on paying a fee of USD 16,600.... Men who left Iran after 1990 may get exempted on paying a fee of USD 1,000-3,000.”

• e.g., Iraq– “It is possible to get exempted from service by paying a sum

of money. In 1998 this was believed to be 10,000 USD.”

• e.g., China– “Given the economic prosperity of [some] areas, young

people seem quite willing to pay the fines imposed for non-compliance with the conscription law.”

Page 24: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

manpower cost envelope

Page 25: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Data Sources

• Military Balance (1985, 90, 95) from the International Institute for Strategic Studies in Sweden– whether any draftees– number of personnel, number drafted– [also military equipment]– cross-checked with

• United Nations Commission on Human Rights• War Resisters International• 1upinfo.com• child-soldiers.org

• War Resisters International: indicates type of system`

Page 26: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Table 2. Conscription and the Size of Military across Countries, 1985-96(1) (2) (3) (4) (5) (6)

regressor0.13 0.16 0.52 0.15 0.15 0.38

(1.48) (2.52) (1.86) (0.94) (3.59) (2.33)

-12.77 0.84 3.77 -6.74 0.93 5.03(1.93) (0.48) (2.08) (2.19) (0.90) (3.50)

British Legal Origin -0.32 -0.49 -0.03 -0.55 -0.46 -0.04(3.12) (4.79) (0.23) (4.37) (5.63) (0.17)

Log(Population)/10 0.70 1.04 0.59 0.33 0.73 0.67(2.35) (3.00) (1.73) (1.28) (3.48) (1.17)-0.09 -0.07 -0.24 0.00 -0.08 -0.33(1.40) (1.02) (2.20) (0.01) (1.59) (4.60)

Constant 1.23 1.14 3.26(1.01) (2.25) (3.28)

Conscription Source, Years WRI '96 MB 1985-95Sample Demobil'd Full Mobil'd Demobil'd Full Mobil'dCountries 41 129 31 41 130 32Pseudo- or Adj. R-squared 0.52 0.28 0.56 0.37 0.42 0.42

NOTESDemobilized sample has armed forces per male aged 15-24 < 0.03Mobilized sample has armed forces per male aged 15-24 > 0.10Absolute value of z statistics (probits) or robust t statistics (linear regressions) in parentheses

Fraction of Population Aged 65+, 1975-95

Real GDP per capita 1975-79, log

Armed Forces Per Male Aged 15-24, 1985-95, log

effects on the fraction of years with conscription

marginal effects from conscription probit

Page 27: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Table 2b. Conscription and the Size of Military across Countries, 1985-96(1) (2) (3) (4)

level regressors marginal effects from probit0.27 0.26 0.17 0.15

(2.98) (3.12) (4.01) (3.90)-6.96 -6.50 -2.90 -2.18(2.10) (2.06) (1.90) (1.69)0.13 0.04 -0.02 -0.02

(1.30) (0.44) (0.28) (0.30)(0.99) 0.16(1.28) (0.38)

British Legal Origin -0.56 -0.57 -0.49 -0.47(4.50) (4.97) (6.06) (6.01)

Log(Population)/10 0.80 1.19 0.61 0.64(1.86) (3.00) (2.54) (3.23)

regressors interacted with log (Armed Forces per Male/.05)9.79 10.75 4.59 4.50

(2.32) (2.90) (2.49) (3.54)0.01 0.03 -0.02 -0.04

(0.10) (0.51) (0.42) (1.01)1.40 0.04

(1.54) (0.11)

Conscription Source, Years WRI '96 MB 1985-95Sample Full Full Full FullCountries 112 129 112 130Pseudo- or Adj. R-squared 0.37 0.37 0.46 0.45

NOTESconstant also included, but not shown in the TableAbsolute value of z statistics (probits) or robust t statistics (linear regressions) in parentheses

Armed Forces Per Male Aged 15-24, 1985-95, log

effects on the fraction of years with conscription

Fraction of Population Aged 65+ minus 0.05, 1975-95Real GDP per capita 1975-79, log minus 7.8

Non-Military Public Spending/GDP, 1975-90

Non-Military Public Spending/GDP minus 0.20, 1975-90

Fraction of Population Aged 65+, 1975-95

Real GDP per capita 1975-79, log

Page 28: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Monopoly and Monopsony• In-Kind vs. Monetary taxation is partly a question of property rights• changing property rights does not eliminate market power on either side of

the market, but it does change the effects of that power– e.g., as the sole employer of military personnel, the government would consider

either the effects of m on the volunteer wage or on the commutation fee– however, commutation fee effects are offsetting

• first effect is unique: increase m in order to increase the size of the lump sum tax• second effect is same as volunteer: reduce m to hold down the volunteer wage• the first effect dominates: even without admin costs, the volunteer system recruits less

– two recruitment systems compared on the basis of the dwcs of monopoly/monopsony

• volunteer m is too small, in-kind m is too large• in-kind S(m) is closer to efficient• larger m increases this case for in-kind taxation• the lump sum tax effect is limited by possibilities for in-kind tax avoidance

• hold out: government sometimes has to purchase from a factor market monopolist– how can hold out be distinguished from heterogeneous private values? … the

case of Mrs. Miller– hold out has two efficiency effects: dwcs of monetary taxes (by reducing

Treasury revenue) and reducing public project scope (via imperfect information)

( ) ( ) ( ) ( )U m S m S m mS m

Page 29: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

In-Kind Taxation for Private Use• Historical examples

– WWII BMW factor workers– urban renewal– utility company easements

• should the public sector auction off the power of eminent domain?

• reasons for a higher standard for private use– market power: private beneficiaries are usually not factor

market monopsonists they face a more elastic factor market supply

– efficient sorting: private beneficiaries have even less reason than the public sector to consider aggregate efficiency (i.e., would choose the wrong point on the policy possibility frontier)

– holdout: private beneficiaries can assemble land in secrecy• other comparative statics are the same

Page 30: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

ordered probit model

Page 31: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Three Main Empirical results

• military manpower systems are ordered– out of 12 possible, (v,u,x,r) and (v,u,r,x) fit the best– “fairness” ordering (u,x,v,r) fits worst

• population and civil law have similar effects– on various manpower margins– on other regulations (business entry procedures,

labor law index)

• “deadweight tax cost” variables matter only when the armed force is large

Page 32: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Why Nixon Drafted, but Bush didn’t

• U.S. has common-law origin• is noncommunist and democratic• 1970: population was 205 million

– about 18 million men aged 15-24– armed forces were 3.1 million– our model predicts 58% draft likelihood

• 2003: population was 290 million– about 21 million men aged 15-24– armed forces were 1.4 million– our model predicts 43% draft likelihood– if armed forces doubled, only 53% likely

Page 33: Taxation in Kind with special attention to Military Manpower Conscription by Casey B. Mulligan

Further Readings on this Approach

• Mulligan and Shliefer– “Conscription as Regulation.” American Law and

Economics Review. 7(1), Spring 2005: 85-111.– “The Extent of the Market and the Supply of

Regulation.” Quarterly Journal of Economics. 120(4), November 2005: 1445-73.

• Glaeser and Shliefer– “Legal Origins.” Quarterly Journal of Economics.

117(4), November 2002: 1193-1229.