Tax Incidence in Thailand

Report
Jonathan Haughton
Suffolk University, Boston
[email protected]
For the World Bank.
May 11, 2011
1.
2.
3.
4.
Who bears the burden of taxes?
Who benefits from government spending?
What are the net effects?
Who would gain/lose under different
possible tax reform packages, and by how
much?
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 2

Central government tax revenue
 11.6% of GDP in 2004
 Buoyant since 2002
 71% of revenue from indirect tax
▪ Vat: 19% rate; but yields just 4.9% of GDP
 Income tax: 3.4% of GDP
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 3
30.0
Revenue + Grants as % of GDP
25.0
20.0
15.0
10.0
5.0
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008 (e)
All other taxes,
0.91
Grants, 0.00
Fees, charges, and
non-tax revenues,
2.36
Corporate income
tax, 7.96
Taxes on trade,
1.93
Natural resources
tax, 1.44
Excise taxes, 1.46
Individual income
tax, 0.60
Value-added tax,
7.11

Revenue/GDP: 20% to 2000, now 25%
 VAT: 7% of GDP (at 10% rate!)
 Trade: from 4% to 2% of GDP, despite explosion of
imports
 All income tax: Peaked at 10% GDP in 2006, but
dependent on SOE sector
 NB: PIT raises 2% of revenue; surprisingly, not
rising
Revenue Results of the Simulation
Budget revenues, million baht
2006
2007
2008
2009
1,502,423
1,690,494
1,623,052
1,557,898
1,536,615
of which:
Revenue Department
969,966
Personal income taxes
153,514
Corporation taxes
339,325
Petroleum income taxes
41,019
Value added taxes
400,967
Specific business taxes
27,887
Stamp duties
6,991
Others
264
Excise Department
394,760
Oil taxes
107,641
Tobacco taxes
52,207
Liguor taxes
41,777
Beer taxes
62,317
Car taxes
85,412
Beverage taxes
14,363
Electronic appliance taxes
5,290
Motorcycle taxes
2,688
Battery taxes
1,095
Telecommunication taxes
19,634
Others taxes
1,715
Miscellaneous
621
Customs Department
107,662
Import duties
104,095
Export duties
278
Miscellaneous
3,289
Other
200,977
Other government agents
85,626
Treasury Department
5,254
State Enterprise Reforms
1,484
State Enterprises
108,613
Total (Gross)
1,673,365
Deduct
Tax rebates of Revenue Department
150,626
VAT
128,675
Other taxes
21,951
VAT allocation for PAO (Provincial Administration Organizations)
7,889
Export duties compensation
12,428
49,825
Total (Net)
1,502,423
1,142,926
188,697
394,024
57,442
459,822
34,882
7,803
257
418,039
106,316
53,021
47,622
69,561
88,919
16,675
5,200
2,947
1,769
23,278
1,770
961
100,526
98,151
362
2,014
239,362
105,490
4,372
647
128,853
1,900,853
175,356
142,314
33,042
9,399
11,766
1,690,494
1,212,640
211,043
408,934
66,569
480,497
37,418
7,920
259
283,987
76,969
42,372
32,008
52,817
56,290
11,596
3,757
1,580
1,562
3,564
1,191
281
99,434
96,676
390
2,368
208,292
67,720
3,661
10,659
89,301
1,836,939
196,258
169,023
27,235
10,879
11,420
1,623,052
1,271,371
204,000
462,134
75,379
501,743
20,220
7,647
249
264,118
54,081
41,844
36,779
52,316
57,944
12,394
3,743
1,789
1,627
111
1,169
321
98,961
96,250
474
2,237
167,412
81,887
5,386
80,138
1,801,862
220,660
186,480
34,179
11,471
11,833
33,229
1,557,898
1,141,817
199,090
392,339
87,799
437,038
17,578
7,747
225
334,723
123,445
46,348
39,477
51,974
53,200
12,747
2,749
1,602
1,580
1,068
532
81,527
77,923
340
3,263
182,787
83,718
3,574
95,495
1,740,853
183,567
141,128
42,439
9,183
11,488
45,181
1,536,615
Total Net Revenue after Deductions
2005

Step 1. Make assumptions about incidence
 Statutory incidence ≠ Effective incidence
 See Table
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 8





VAT: on consumers
Excises: on consumers
PIT: on earners
Business profits: on earners
Other taxes:
▪ Property transfer; local fees and contributions: on
payers.
▪ In this study, CIT, trade taxes, natural resource taxes, not
included. Incidence covers half of revenue.
S(1+t)
price
supply
Pd2
P1
Ps2
demand
Q2
Q1
quantity
price
What elasticity?
Pd2
S(1+t)
P1=Ps
2
supply
demand
Q2
Q1
quantity
Table 4. Incidence Assumptions
Tax
Personal income
Value Added Tax
Incidence Assumption
Borne by earners (because their take-home pay
would rise if there were no tax).
Shifted onto consumers.
Fuel excise
Shifted onto consumers because of horizontal
supply curve.
Alcohol excise
Shifted onto consumers because of horizontal
supply curve.
Shifted onto consumers because of horizontal
supply curve.
Shifted onto consumers.
Incidence Assumption
Benefits accrue to households whose children are
currently enrolled in publicly-supported schools.
Benefits accrue to households whose sick
members sought help at a publicly-supported
hospital or clinic.
Benefits accrue to households to the full extent of
the subsidies.
Tobacco excise
Car excise
Subsidy/spending
Education
Health
Social subsidies
May 11, 2011
Comments on operationalization
Uses amounts paid as reported in the ENNIV-2000 survey, and
grosses these up to get an appropriate total yield.
Assumed to be 0% on exempt items – which understates the tax –
using 57 expenditure lines. No adjustment was possible for the source
of purchases (e.g. supermarket, street market, etc.), which overstates
tax, especially on poorer households.
Allocated in proportion to reported household spending on energy.
[This is an overstatement, as some effects work via transport,
electricity, and individual use, where the incidence pattern is different.]
Reported spending on alcohol is typically highly understated.
Allocated in proportion to reported household spending on tobacco;
also typically highly understated.
Allocated in proportion to car value.
Comments on operationalization
Applied average budgetary cost per enrolled student, differentiated by
main level of education (primary, secondary, tertiary).
Survey provides information on costs in public and private facilities;
subsidy was taken to be the differential between the amount actually
paid and an appropriate private-sector norm.
This overstates the benefits, given that labor supply is not completely
inelastic, but probably not by much.
JH: Tax & Expenditure Incidence in Peru
Page 12

Step 2. Quantify the effects
 Trace effect of a tax, spending change on every
household in a survey
 ENNIV 2000. 3,997 households, 19,957 people,
LSMS template
 NB. Assumes equal sharing within household
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 13

Operationalize the analysis
 Excel spreadsheet
 Stata dataset and programs.
Change spreadsheet; it invokes Stata, returns the
results.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 14
 9,189 households, including 4,298 from 2004
round.
 Two visits per household; 93% of interviews in
June, Sept or Oct.
 Some explicit tax information (e.g. business
taxes); otherwise has to be inferred (e.g. VAT, PIT).

Socio-economic survey, 2009
 47-page data dictionary
 139,590 individuals
 From 43,844 households
▪ Aside: US 1% IRS file plus non-filers: c. 150,000 filers.




19% rate. Fairly stable since 1992.
Exports zero rated
Exemptions include clothing, rice, milk, fish,
vegetables, ed. fees, home consumption,
housing
In 2000:
 42% of central gov. tax revenue
 Collected 7.3% of household expenditure.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 17
Table 11. Reconciling Survey with Budget Data
Tax revenue, based on VHLSS-2006
Based on budgeted
Per capita, VND ‘000 Nationally, VND trn
national receipts, VND trn
VAT
0.522
43.9
54.8
Excise taxes
0.085
7.1
17.1
Personal income tax
0.006
0.5
5.2
Business profits tax
0.121
10.2
100.8
Fees
0.212
17.8
3.9
Source: VHLSS-2006 data from Table 7; budgetary information from Table 1.
Categories
Tax Rates
Wide base of goods, 2004
Wide base of goods, 2000
Exports
Most services, some goods
Revenue
Tax revenue, 2004, m soles
Tax revenue, 2000, m soles, net of refunds
Tax revenue as % of total tax revenue, 2000
Actual tax revenue / estimated revenue
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (>0 = progressive)
A
19.0%
18.0%
0.0%
Exempt
D
D
D
D
D
D
D
D
D
D
13,206.9
9,550.6
41.5
85%
Expend/cap
0.470
0.455
-0.00116
-0.00125
-0.015
Inc/cap
0.535
0.358
-0.01145
-0.01207
-0.176
L
A
R
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the Reynol
Source: Based on ENNIV-2000. More complete results are shown in Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 19
Regressivity probably overstated, if poor are more
likely to buy in informal sector


By expenditure/cap: slightly regressive
By income/cap: highly regressive
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 20

Income:
 Easy to measure
 Widely used in developed countries
 Overstates regressivity: “lifetime income”

Expenditure
 Less underreporting than income
 Closer to “permanent income”
 May understate regressivity
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 21

In 2000, summarized:
 Alcohol, especially beer (t=27.8% of 84% of rec.




prodr. price)
Soft drinks (t=17% ex factory)
Cigarettes (t=37.2%)
New vehicles (t=10%)
Motor fuel
▪ Gasoline (S/.2.90 per gallon)
▪ Diesel S/.2.29 per gallon)
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 22
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 23
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 24
Expend/cap:
Kakwani (>0
Income/cap:
Kakwani (>0
= progressive)
Revenue as
% of all tax
revenue
Actual
revenue as %
of estimated
Alcohol
0.057
-0.110
3.6
213
Soft drinks
-0.105
-0.283
0.6
127
Tobacco
-0.003
0.5
170
JH: Tax & Expenditure Incidence in Peru
= progressive)
Note under-reporting of expenditures
May 11, 2011
Page 25
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 26
Expend/cap:
Kakwani (>0
Income/cap:
Kakwani (>0
= progressive)
Revenue as
% of all tax
revenue
Actual
revenue as %
of estimated
Motor fuel
0.378
0.246
9.3
732
Vehicles
0.412
0.288
1.0
225
JH: Tax & Expenditure Incidence in Peru
= progressive)
Notes:
Fuel numbers overstate progressivity, because of the “bus and truck
problem”
Vehicle numbers assume purchases in proportion to ownership, which is
awkward
Ideal is to run this through an input-output table
May 11, 2011
Page 27
Distributional
Results of the simulation
Expenditure/cap
Est. tax/capita
Est. tax/expenditure
2009
2009
2009
Expenditure per capita deciles
1 (poorest)
2
3
4
5
6
7
8
9
10 (richest)
14,536
20,212
24,473
29,099
34,391
41,082
49,595
61,650
82,146
163,833
809
1,221
1,564
1,944
2,400
2,987
3,897
5,100
7,173
15,511
5.6%
6.0%
6.4%
6.7%
7.0%
7.3%
7.9%
8.3%
8.7%
9.5%
102,720
60,462
41,138
37,126
53,224
8,020
5,163
3,289
2,897
4,715
7.8%
8.5%
8.0%
7.8%
8.9%
76,862
46,010
6,334
3,328
8.2%
7.2%
52,102
4,261
8.2%
Regions
1:
2:
3:
4:
5:
Bangkok Metropolis
Central (excluding Bangkok)
North
Northeast
South
Areas
Urban
Rural
Thailand overall
Average
Categories
Tax Rates
0 – 27 UIT (i.e. 0 – S/.86,400 p.a.)
27 UIT – 54 UIT (i.e. S/.86,400 – 172,800 p.a.)
> 54 UIT (i.e. > S/.172,800 p.a.)
Wages & salaries, # of UITs deductible
15%
21%
30%
7
Revenue
Tax revenue, 2003, m soles
Tax revenue, 2000, m soles, net of refunds
Tax revenue as % of total tax revenue, 2000
Actual tax revenue as % of estimated revenue
2,669
2,117
9.2
467
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (>0 = progressive)
Expend/cap
0.470
0.627
0.00257
0.00217
0.157
Inc/cap
0.535
0.582
0.00065
0.00031
0.047
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the Reyno
Source: Based on ENNIV-2000. More complete results are shown in Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 30
Categories
Tax Rates
Includes:
Personal income tax
Excises on alcohol, tobacco, beer, cars, soda
Fuel excise tax
Revenue
Tax revenue, 2003, m soles
Tax revenue, 2000, m soles, net of refunds
Tax revenue as % of total tax revenue, 2000
Actual tax revenue as % of estimated revenue
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (>0 = progressive)
Expend/cap
0.470
0.545
0.00971
0.00831
0.075
15116
65.7
120
Inc/cap
0.535
0.460
-0.00801
-0.01087
-0.075
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the Reyn
Source:
Based on ENNIV-2000.
Progressive
or not? More complete results are shown in Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 31
In cid en ce of M ain Taxes
0
.1
.2
.3
P e ru, 20 00
0
10 000
20 000
30 000
Expenditure p er cap ita, S/.
Fitted values
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
40 000
50 000
ta xburd
Page 32
Tax as a % of expenditure, by decile, 2006
VAT as a % of expenditure, by decile, 2006
10.0
Taxes paid as % of expenditure
Taxes paid as % of expenditure
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Expenditure per capita decile (poor = 1, rich = 10)
Overall
5
6
7
8
9
10
9
10
VAT
Excise Tax as a % of expenditure, by decile, 2006
Educational Fees as a % of expenditure, by decile, 2006
1.6
2.5
Taxes paid as % of expenditure
Taxes paid as % of expenditure
4
Expenditure per capita decile (poor = 1, rich = 10)
1.4
1.2
1.0
0.8
0.6
0.4
0.2
-
2.0
1.5
1.0
0.5
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Expenditure per capita decile (poor = 1, rich = 10)
Excises
4
5
6
7
Expenditure per capita decile (poor = 1, rich = 10)
Educational fees
8
Agricultural Fees as a % of expenditure, by decile, 2006
Other Fees as a % of expenditure, by decile, 2006
1.4
Taxes paid as % of expenditure
Taxes paid as % of expenditure
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-
1.2
1.0
0.8
0.6
0.4
0.2
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Expenditure per capita decile (poor = 1, rich = 10)
Agricultural fees
5
6
7
8
9
10
9
10
Other fees
Tax on business income as a % of expenditure, by decile, 2006
Personal Income Tax as a % of expenditure, by decile, 2006
3.5
0.3
Taxes paid as % of expenditure
Taxes paid as % of expenditure
4
Expenditure per capita decile (poor = 1, rich = 10)
3.0
2.5
2.0
1.5
1.0
0.5
-
0.3
0.2
0.2
0.1
0.1
-
1
2
3
4
5
6
7
Expenditure per capita decile (poor = 1, rich = 10)
Business income
8
9
10
1
2
3
4
5
6
7
Expenditure per capita decile (poor = 1, rich = 10)
Personal Income Tax
8
Tax as a % of income, by decile, 2006
VAT as a % of income, by decile, 2006
7.0
Taxes paid as % of income
Taxes paid as % of income
14.0
12.0
10.0
8.0
6.0
4.0
2.0
-
6.0
5.0
4.0
3.0
2.0
1.0
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Income per capita decile (poor = 1, rich = 10)
Overall
5
6
7
8
9
10
9
10
VAT
Excise tax as a % of income, by decile, 2006
Educational Fees as a % of income, by decile, 2006
1.2
1.8
Taxes paid as % of income
Taxes paid as % of income
4
Income per capita decile (poor = 1, rich = 10)
1.0
0.8
0.6
0.4
0.2
-
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Income per capita decile (poor = 1, rich = 10)
Excises
4
5
6
7
Income per capita decile (poor = 1, rich = 10)
Educational fees
8
Other fees as a % of income, by decile, 2006
0.8
1.0
0.7
0.9
Taxes paid as % of income
Taxes paid as % of income
Agricultural fees as a % of income, by decile, 2006
0.6
0.5
0.4
0.3
0.2
0.1
-
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-
1
2
3
4
5
6
7
8
9
10
1
2
3
Income per capita decile (poor = 1, rich = 10)
Agricultural fees
5
6
7
8
9
10
Other fees
Personal income tax as a % of income, by decile, 2006
Tax on business income as a % of income, by decile, 2006
0.3
Taxes paid as % of income
3.0
Taxes paid as % of income
4
Income per capita decile (poor = 1, rich = 10)
2.5
2.0
1.5
1.0
0.5
0.2
0.2
0.1
0.1
-
1
2
3
4
5
6
7
Income per capita decile (poor = 1, rich = 10)
Business income
8
9
10
1
2
3
4
5
6
7
Income per capita decile (poor = 1, rich = 10)
Personal Income Tax
8
9
10

In Vietnam, incidence pattern by expenditure/cap
decile similar to that by income/cap
 Contrast with Peru

These taxes are progressive overall
 Heavier, more progressive, than 1998
 VAT is generally progressive – mainly due to home
production (1/3 at bottom, 1% at top); not the case in all
countries
 Excise taxes are progressive – also a bit surprising.
Table 10. Measures of Tax Progressivity
Expenditure per capita
Income per capita
All allocable taxes
Personal income tax
Business income tax
Excise taxes
Fees: other
VAT
Fees: educational
Fees: agricultural
Gini coefficient (GX)
0.390
0.408
Concentration Ratio (CT,X)
0.461
0.986
0.686
0.504
0.459
0.443
0.354
-0.018
Property tax
0.632
Property tax: own residence
0.619
Property tax: other than residence
0.751
Property tax with VND100m threshold
0.776
Source: Based on Vietnam Household Living Standards Survey 2006.
Kakwani measure (CT,X – GX)
0.071
0.597
0.296
0.114
0.069
0.054
-0.036
-0.408
0.242
0.229
0.361
0.386






Regional effects
Personal Income Tax
Property Tax
Expenditure incidence
Marginal full incidence
Policy reforms
Higher tax rates in south
Highest rate in South-Central Coast (sampling
error?)
 Urban/rural gap in burden is surprisingly small


Table 12. Structure of Current and Proposed Personal Income Taxes
Coverage
Rates & Base
Current tax
Wages, salaries, bonuses, housing
allowances.
For Vietnamese
Income p.a., VND m
Tax rate, %
New tax
Wages, salaries, bonuses, housing
allowances, business income, interest,
dividends, capital gains, prizes.
Taxable income p.a.,
VND m
0 – 60
60 – 120
120 – 216
216 – 384
384 – 624
624 – 960
>960
Tax rate, %
0 – 60
0
5
60 – 180
10
10
180 – 300
20
15
300 – 480
30
20
>480
40
25
For foreigners
30
0 – 96
0
35
96 – 240
10
240 – 600
20
Personal deduction: VND 48 m
600 – 960
30
Deduction/dependent: VND 19.2 m
> 960
40
Sources: IMF (2007); Shukla (2006); Law on Personal Income Tax /2007/QH12.
1 (poor)
2
Expenditure per capita deciles
4
5
6
7
3
8
9
10 (rich)
All
% hh
paying
thousands of VND p.a.
Household expenditure per capita
2,000
2,938
3,640
4,324
5,077
5,981
7,149
8,811
11,593
22,681
7,417
Total tax paid
of which:
taxes on household enterprises
personal income tax
Total new tax paid
on income
156
252
372
458
557
709
984
1,151
1,679
3,146
946
100.0
35
163
-
116
-
184
1
651
61
121
6
15.4
0.2
1
13
190
21
1.0
Total tax paid
of which:
taxes on household enterprises
personal income tax
Total new tax paid
on income
7.8
8.6
10.2
10.6
11.0
11.9
13.8
13.1
14.5
13.9
13.4
0.1
-
0.1
-
0.3
-
0.3
-
0.6
-
0.6
-
2.3
-
1.3
-
1.6
0.0
2.9
0.3
1.6
0.1
-
-
-
-
-
-
0.1
0.0
0.1
0.8
0.3
3
-
3
-
12
-
12
-
30
-
-
8
as percentage of expenditure
Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS
survey of 2006. All figures are in prices of 2006.

Method: strip out business income tax and
PIT, and apply rules of new tax
 Revenue: from VND127k to VND21k per person.
 New tax highly progressive.
 But: excludes foreigners; very small sample size
for PIT and for large household enterprises.

Done by General Dept. of Taxation in 2005,
using tax rolls
 Survey, asking about income, spending
 15,500: 3,200 foreign, 7,200 PIT, 5,100 business
income tax
 11,535 responses (74%); low in HCMC, among
foreigners. Results re-weighted.

Anything equivalent in Thailand?
Vietnamese
Taxpayer income
of which:
Wages and salaries
Enterprise
Business
Finance and capital
Transfers
Other
Memo items:
Total household income
Household current spending
Household asset purchases
Personal income tax/taxpayer
Estimate, 2005
Estimate, 2009, old rules
Memo: % growth
Estimate, 2009, new rules
Personal income tax revenue
Estimate, 2005
Memo: Actual PIT revenue, 2005
Estimate, 2009, old rules
Estimate, 2009, new rules
Foreigners
Household
enterprises
thousands of VND per year
Full sample
195,003
631,835
244,158
302,514
71,263
498
37,474
79,732
98
5,938
620,939
810
1,548
8,190
5
342
43,753
327
191,095
7,022
227
1,734
177,172
507
80,459
40,869
121
3,386
370,619
79,273
255,995
671,127
116,266
38,100
266,619
39,741
59,914
399,276
74,014
145,971
239,751
527,604
120%
392,175
74,315
136,219
83%
162,330
77,792
175,319
125%
153,426
4,504
15,907
253%
39,057
628
2,217
5,443
14,849
4,200
32,677
24,290
% foreign
7,336
22,812
65%
13,446
16,024
48,340
45,756
68%
53%

Based on 2009
 NB. PIT very elastic
 Total revenue: -5%
▪ Foreigners: down by a quarter
▪ Vietnamese: tax payments up

Reconcile with VHLSS data
 High-income individuals would pay more
 Not-so-high income individuals would pay less, and not be in the tax net as
much
 Difficult to merge the two; in progress.

Keep: taxes foreigners, adds equity. Enforcement could be better.
Perhaps limit top rate to top CIT. Adjust brackets for inflation (as in US).

VHLSS has limited data on property:
 Residence; non-agric. real estate
 Assume tax would not apply to agricultural
property, or to movable property.

1% of capital value; arbitrary
1 (poor)
2
Expenditure per capita deciles
4
5
6
7
3
8
9
10 (rich)
All
% hh
paying
thousands of VND p.a.
Household expenditure per capita
2,000
2,938
3,640
4,324
5,077
5,981
7,149
8,811
11,593
22,681
7,417
Total tax paid
of which:
taxes on residence
taxes on other property
Total tax paid
with VND100m threshhold per household
50
82
119
168
210
277
384
545
1,028
2,944
580
97.9
49
1
80
2
113
5
157
10
200
10
257
20
357
26
500
46
922
106
2,587
358
522
58
97.5
8.9
1
4
12
44
335
798
2,677
424
35.8
Total tax paid
of which:
taxes on residence
taxes on other property
Total tax paid
with VND100m threshhold per household
2.5
2.8
3.3
3.9
4.1
4.6
5.4
6.2
8.9
13.0
7.8
2.4
0.1
2.7
0.1
3.1
0.1
3.6
0.2
3.9
0.2
4.3
0.3
5.0
0.4
5.7
0.5
8.0
0.9
11.4
1.6
7.0
0.8
0.1
0.1
0.3
1.0
1.2
1.8
2.8
3.8
6.9
11.8
5.7
62
109
197
as percentage of expenditure
Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS
survey of 2006. All figures are in prices of 2006.
1 (poor)
Income per capita
2,370
2
3,511
3
4,387
4
Income per capita deciles
5
6
7
5,279
6,268
7,508
9,076
8
9
11,118
14,660
10 (rich) All
29,343
9,351
% hh
paying


Highly progressive
1% is steep; 7.8% of expenditure, so
politically infeasible at this level.
 Cash flow concerns
 Based on bubble prices?
 Introduction of tax would reduce base

Excludes corporate ownership
Tax paid
Expenditure
per capita
Tax as % of
expenditure
Tax paid with
threshhold
'000 VND
Region
1
2
3
4
5
6
7
8
Urban
Rural
Overall
Tax as % of
expenditure
# of households
sampled
'000 VND
791
8,241
9.6
599.6
7.3
305
5,772
5.3
172.6
3.0
244
4,393
5.5
140.8
3.2
248
5,053
4.9
119.5
2.4
387
6,766
5.7
238.1
3.5
312
5,955
5.2
190.4
3.2
Note: The1,422
regions are: 1: Red River
Delta,
2:
Northeast,
3:
Northwest,
4:
North
Central
Coast,
5:
South
Central
Coast,
6:
11,852
12.0
1,224.0
10.3
Central Highlands, 7: Southeast, 8: Mekong River Delta.
277
6,698
4.1
135.7
2.0
1,944
1,317
429
1,014
852
582
1,188
1,863
1,425
272
12,394
5,602
11.5
4.9
1,211.6
136.2
9.8
2.4
2,307
6,882
580
7,417
7.8
423.6
5.7
9,189




Applied to education, health, targeted social
programs
Step 1: Measure unit subsidies
Step 2: Identify coverage
Step 3: Present results
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 51

In 2000
▪ 98.8% of children 7-10 were at school
▪ 94% of 14-year-olds were at school
▪ 13% of pupils/students were at private schools

Costs
▪
▪
▪
▪

Pre-K:
S/. 583 per child per year
K & primary: S/. 386
Secondary: S/. 624
Tertiary:
S/.2,506
Assigned based on attendance
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 52
Cost per pupil,
S/.
583
386
624
2,506
Categories
Pre-kindergarten
Kindergarten & primary
Secondary
Tertiary
Spending
Spending, 2003, m soles
Spending, 2000, m soles
Spending as % of total tax revenue, 2000
Actual spending as % of estimated spending
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (<0 = progressive)
Expend/cap
0.470
0.102
0.01282
0.01234
-0.369
A
D
D
D
D
D
D
D
D
D
D
5,921
4,740
20.6
69
Inc/cap
0.535
0.090
0.01305
0.01271
-0.444
L
A
R
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the Reyno
Source: Based on ENNIV-2000. More complete results are shown in Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 53
00
Sorted by expenditure/capita
per pupil,
S/.
583
386
624
506
21
40
20.6
69
Inc/cap
0.535
0.090
0.01305
0.01271
-0.444
All Peru (2000)
Spending/
Expend, %
3.6
Decile 1 (poorest)
Decile 2
Decile 3
Decile 4
Decile 5
Decile 6
Decile 7
Decile 8
Decile 9
Decile 10 (richest)
15.6
9.8
7.6
7.0
5.4
4.8
4.0
3.5
2.9
1.3
Lima/Callao
Amazon (4 depts)
Rest of Peru
2.2
4.5
4.8
Spending/
Income, %
3.0
% of
spending
100.0
Sorted by
inc/cap
Spending/
Income, %
3.0
11.4
6.0
5.8
5.0
4.2
3.7
3.4
2.8
2.7
1.2
7.9
8.1
8.3
9.5
9.0
9.8
10.1
11.0
12.7
13.7
34.7
13.5
8.3
6.2
4.7
4.4
3.6
3.0
2.2
1.0
n/a
n/a
n/a
27.2
15.7
57.1
1.8
3.9
4.0
and RS2 is the Reynolds-Smolensky measures of redistributive capacity.
“progressive but not well-targeted”
Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 54

Cost of procedures
 Used ENNIV-2000 reported data on cost at private
facilities
 Adjusted downwards substantially to be
consistent with government spending
▪ Cost/consultation
S/. 0.75/minute
▪ Cost of hospitalization S/. 75/night
▪ Cost of an analysis
S/. 6.2/item
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 56
Consultation, S/. per minute
Hospitalization, S/. per day
Analysis, S/. per person
Cost per pupil,
S/.
1.5
150
12.5
Spending
Spending, 2003, m soles
Spending, 2000, m soles, net of refunds
Spending as % of total tax revenue, 2000
Actual spending as % of estimated spending
1,922
8.4
48
Categories
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (<0 = progressive)
Expend/cap
0.470
0.137
0.00480
0.00371
-0.334
Inc/cap
0.535
0.006
0.00641
0.00546
-0.529
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the R
Source: Based on ENNIV-2000.
More complete results are shown in Appendix
1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 57
Sorted by expenditure/capita
r pupil,
/.
All Peru (2000)
4
c/cap
535
006
00641
00546
529
Spending/
expend
1.5
Decile 1 (poorest)
Decile 2
Decile 3
Decile 4
Decile 5
Decile 6
Decile 7
Decile 8
Decile 9
Decile 10 (richest)
5.6
3.4
3.5
2.1
2.5
1.5
1.9
1.6
1.3
0.6
Lima/Callao
Amazon (4 depts)
Rest of Peru
0.8
2.4
1.9
Sorted by
inc/cap
Spending/
income
1.2
Spending/
income
1.2
% of
spending
100.0
4.1
2.1
2.7
1.5
1.9
1.2
1.6
1.3
1.2
0.5
6.9
6.9
9.4
6.9
10.1
7.6
11.7
12.6
14.0
13.9
17.9
7.3
3.1
3.2
1.7
1.9
1.0
1.2
1.0
0.3
23.3
21.2
55.5
0.6
2.1
1.6
n/a
n/a
n/a
RS2 is the Reynolds-Smolensky measures of redistributive capacity.
pendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 58
Table 17. “Social subsidies,” 2000
School meals (“desayuno escolar”)
Milk (“vaso de leche”)
Food support (“comedor popular”)
Mother’s groups (“club de madres”)
Family food basket (“canasta familiar (panfar)”)
Food for work (“alimento por trabajo”)
Direct food support (“donación directa de alimentos”)
Child nutrition (“papilla u otro alimento para menores”)
School uniforms
School books and materials
Source: Based on ENNIV-2000.

Mean amount received per
household, S/. p.a.
23.58
34.31
23.14
1.48
1.05
1.67
0.90
1.94
4.25
14.13
Allocated based on reported usage
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 59
Cost per pupil,
S/.
Categories
Major items:
School meals
Milk
Food support
23.58
34.31
23.14
Spending
Spending, 2003, m soles
Spending, 2000, m soles, net of refunds
Spending as % of total tax revenue, 2000
Actual spending as % of estimated spending
Distributional effects, 2000
Gini coefficient
Quasi-Gini coefficient, tax/capita
RS1 measure (>0 = progressive)
RS2 measure (>0 = progressive)
Kakwani measure (<0 = progressive)
Expend/cap
0.470
-0.217
0.01854
0.01504
-0.688
3,646
15.9
256
Inc/cap
0.535
-0.191
0.01651
0.01438
-0.725
Note: RS1 is the Reynolds-Smolensky measure of disproportionality and RS2 is the R
Source: Based on ENNIV-2000. More complete results are shown in Appendix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 60
Sorted by expenditure/capita
pupil,
All Peru (2000)
/cap
535
191
01651
01438
725
Spending/
expend
2.8
Decile 1 (poorest)
Decile 2
Decile 3
Decile 4
Decile 5
Decile 6
Decile 7
Decile 8
Decile 9
Decile 10 (richest)
20.5
14.0
9.1
7.2
5.2
4.4
2.8
1.7
0.7
0.2
Lima/Callao
Amazon (4 depts)
Rest of Peru
1.9
3.7
3.3
Spending/
Income
2.3
% of
spending
100.0
Sorted by
inc/cap
Spending/
income
2.3
15.0
8.7
7.0
5.1
4.1
3.4
2.4
1.4
0.6
0.2
13.5
15.1
12.9
12.6
11.3
11.6
9.2
7.1
3.8
2.8
44.6
19.0
9.3
5.8
4.6
3.4
2.3
1.6
1.0
0.1
n/a
n/a
n/a
31.0
16.9
52.1
1.6
3.2
2.8
RS2 is the Reynolds-Smolensky measures of redistributive capacity.
endix 1.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 61

System seems progressive overall
 But
▪ Some taxes not included (e.g. CIT)
▪ Not all spending can be included
▪ Rely on our incidence assumptions
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 62
All Peru (2000)
Tax/
expend
3.7
Tax/
income
3.1
% of tax
revenue
100.0
Sorted by
inc/cap
Tax/
income
3.1
Decile 1 (poorest)
Decile 2
Decile 3
Decile 4
Decile 5
Decile 6
Decile 7
Decile 8
Decile 9
Decile 10 (richest)
-34.3
-19.1
-11.1
-6.5
-3.4
-0.5
1.7
3.7
6.4
11.5
-25.1
-11.9
-8.5
-4.6
-2.6
-0.4
1.4
3.0
6.0
10.5
-17.1
-15.7
-11.9
-8.6
-5.5
-0.9
4.1
11.6
27.6
116.4
-60.9
-23.8
-8.1
-4.1
-0.5
0.0
1.9
3.2
4.5
9.1
Lima/Callao
Amazon (4 depts)
Rest of Peru
6.8
0.0
1.5
82.5
0.0
17.5
5.6
0.0
1.3
Sorted by expenditure/capita
es
ap
5
8
053
413
4
n/a
n/a
n/a
S2 is the Reynolds-Smolensky measures of redistributive capacity.
JH: Tax & Expenditure Incidence in Peru
dix 1. May 11, 2011
Page 63
Table 22. Revenue and Progressivity Measures for Taxes and Spending, 2003
Reynolds-Smolensky measures
Social subsidies
Educational spending
Gasoline excise tax
Health spending
Personal income tax
Vehicle excise tax
Alcohol excise tax
Cigarette excise tax
Soda excise tax
VAT
Memo:
All taxes
Social spending
Tax net of spending
May 11, 2011
Revenue or
spending, S/.
millions
3,646
4,740
2,145
1,922
2,117
240
817
106
141
9,551
15,116
10,308
4,808
Disproportionalit
y (RS1)
0.01854
0.01282
0.00626
0.00480
0.00257
0.00075
0.00036
0.00000
-0.00011
-0.00116
Redistributive
capacity (RS2)
0.01504
0.01234
0.00582
0.00371
0.00217
0.00075
0.00032
0.00000
-0.00011
-0.00125
0.00971
0.03451
0.04755
JH: Tax & Expenditure Incidence in Peru
0.00831
0.02975
0.04036
Kakwani measure of progressivity
-0.688
-0.369
0.378
-0.334
0.157
0.412
0.057
-0.003
-0.105
-0.015
V. progressive
Progressive
Progressive
Progressive
Progressive
0.075
-0.475
1.254
Page 64

Addresses question:
 If a tax were raised, and spending increased as a
result, what would effect be on incidence?


Step 1: Estimate effect of more tax revenue
on spending
Step 2: Simulate effect of higher VAT in Peru
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 65

Estimated using regression of form
yit     . xit   i   it


i.e. spending/GDP (y) is a function of tax
revenue/GDP (x)
Panel Data: 16 Latin American countries,
1980-2002. “Fixed effects” estimates.
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 66
Table 25. Average and marginal effects of incremental tax revenue
% of GDP
Tax revenue
Government expenditure
Of which:
Education
Health
Social security
Law and order
Interest payments
Other
16.0
21.7
% of tax
revenue
100.0
135.4
Sample
size
283
283
Marginal
effect
100.0
109.4
3.0
2.0
4.9
2.7
2.8
5.5
18.8
12.6
30.8
16.7
17.5
34.4
265
263
251
169
273
163
7.6
7.3
36.2
26.0
-0.2
35.1
p-value
0.00
0.00
0.00
0.00
0.00
0.79
0.00
Note:
•Low marginal effects for education, health
•No marginal effect for interest payments
•High marginal effects for social spending
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 67
 2000: assume VAT from 18% to 19%
 Of extra revenue:
▪ Education takes 7.6%
▪ Health takes 7.3%
▪ Social subsidies take 28.4%
▪ Remaining spending not allocable
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 68
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 69

A higher VAT need not hurt the poor
 Result is robust to whether we use
expenditure/cap or income/cap to sort households
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 70
 Greater disaggregation of spending
 Sensitivity to assumptions about incidence
 Source of purchases/evasion
 Better treatment of transport
 Use income tax data
 Add social security pensions
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 71
 Confidence intervals (bootstrapping?)
 Create integrated “calculators”; and do it in-house
 Need better access to primary data
 Adjust for quality/value of benefits
 Theory: why income/cap results differ so much
from expend/cap results
▪ Measurement error (Deaton)/permanent income
May 11, 2011
JH: Tax & Expenditure Incidence in Peru
Page 72
7. Nine Current Issues of Tax Policy
Having set out the key features of each main tax, we now turn to an analysis of several current issues of
tax policy, one at a time. We are then ready, in the final section, to discuss a number of complete tax
reform strategies.
Issue 1: How High a VAT?
Given the government’s need for more revenue, it is widely accepted that the VAT rate will need
to rise. The issue is: by how much?
A good place to start is with an analysis of an increase in the VAT from 10% to 12%, which is the
smallest increase under serious consideration and is generally accepted to be quite feasible. The impact is
set out below: overall tax revenue would rise by 3.9%, equivalent to 0.65% of GDP; and the increase in
the tax burden would be relatively larger for more affluent households. We have assumed that as the
VAT rate rises, people purchase fewer goods that are subject to the VAT; in part this is because they face
a budget constraint, so a rise in the VAT rate from 10% to 12% would, if spending is held constant, raise
revenue by 17.9% and not by 20%. In addition, faced with a higher VAT rate, buyers would tend to
substitute items that are not subject to the VAT. We use a “revenue loss parameter” of 0.8, which means
that a 20% increase in the VAT rate would be accompanied by a 16% (i.e. 0.8 times 20%) rise in revenue
(and we allow for an additional response for motor fuel and tobacco). Even this may be too low; Agha
and Haughton (1996) estimate, based on data from OECD countries, that every percentage point increase
in the VAT rate is associated with a 2.7% fall in compliance, which would imply a revenue loss parameter
of about 0.7.
An increase in the VAT from 10% to 12% would raise little more than a fifth of the revenue that
is considered necessary if Lebanon is to escape from its current vicious cycle of debt. This prompts one
to ask what would be the effect of an increase in the VAT rate from 10% to 16%. The result would be a
LBP510bn rise in revenue, equivalent to 2% of GDP. Again, the burden would fall slightly more on rich
than on poor households, although even the latter would find their real expenditure cut by 2.3%. Here too
we have taken into account that a 60% rise in the nominal VAT rate would lead to an increase in revenue
of substantially less than this.
An important question is whether the VAT, introduced just over two years ago, is robust enough
to withstand a substantial increase in the rate. This is a clear case where “Tax administration is
Exercise 1: Raise Value Added Tax from 10% to 12%
Categories
Total tax revenue, LBP bn
Of which:
VAT
Fuel excises
Tobacco excises
Memo: Gini coefficient for all taxes
Beirut
Suburbs of Beirut
Rest of Mt. Lebanon
North Lebanon
South Lebanon
Nabatieh
Beka'a Valley
Old
4,502
1,361
819
185
0.474
Change in tax/expenditure
LBP/cap p.a.
% change
-71
-0.95%
-59
-0.98%
-70
-1.01%
-40
-1.03%
-37
-1.02%
-40
-1.02%
-36
-0.94%
New
4,679
1,555
804
182
0.472
All Lebanon
Poorest quintile
Poor-mid quintile
Middle quintile
Mid-upper quintile
Upper quintile
% change
+3.9
+14.3
-1.8
-1.8
Revenue Loss
Parameter
0.8
Change in tax/expenditure
LBP/cap p.a.
% change
-0.99%
-51
-14
-24
-35
-52
-130
-0.81%
-0.85%
-0.88%
-0.94%
-1.11%
Exercise 2: Raise Value Added Tax from 10% to 16%
Categories
Total tax revenue, LBP bn
Of which:
VAT
Fuel excises
Tobacco excises
Memo: Gini coefficient for all taxes
Beirut
Suburbs of Beirut
Rest of Mt. Lebanon
North Lebanon
South Lebanon
Nabatieh
Beka'a Valley
Old
4,502
1,361
819
185
0.474
Change in tax/expenditure
LBP/cap p.a.
% change
-205
-2.74%
-172
-2.84%
-204
-2.92%
-116
-2.97%
-106
-2.94%
-115
-2.94%
-105
-2.73%
New
5,012
1,923
777
175
0.472
All Lebanon
Poorest quintile
Poor-mid quintile
Middle quintile
Mid-upper quintile
Upper quintile
% change
+11.3
+41.3
-5.2
-5.2
Revenue Loss
Parameter
0.8
Change in tax/expenditure
LBP/cap p.a.
% change
-2.86%
-148
-42
-70
-100
-152
-375
-2.34%
-2.46%
-2.56%
-2.71%
-3.19%
from the capacity and evolution of tax administration.
Of the 15,400 registered VAT payers, roughly 1,000 are audited every quarter, almost double the
proportion of most other jurisdictions, and a reflection of the newness of the tax. On the other hand the
VAT department is well trained and adequately staffed, and could probably make a VAT of 16% work
without too much leakage, particularly as three fifths of the revenues are collected on imports.
A broader question is how far one can push a VAT. Agha and Haughton have constructed an
interesting graph, reproduced here, that shows that once the VAT rate exceeds about 15%, the pressure to
shrink the base becomes very strong. The implication is that the revenue-maximizing point is then
reached quite quickly, and appears to be in the region of 7-8% of GDP; all the countries in the sample
whose VAT raised more than 7% of GDP are ranged along the frontier in this diagram.. Our simulations,
detailed above, show that a VAT of 16% would raise about 7.0% of GDP in revenue, which is as far as
one can reasonably expect to go.
2. A Revenue-Neutral Reform, Emphasizing Equity
An alternative revenue-neutral reform would keep the VAT at 10%, but improve the efficiency of taxes
on the income side. The wage and salary tax, and the tax on proprietors’ income, would be unified with a
10% basic rate, rising to 20% on income above LBP60 million; and with a personal tax deduction of 1.5
million. The corporation income tax would rise to 20%, while the capital gains and inheritance taxes
would be abolished. A property tax levied at 0.5% of the value of property above LBP120m would
replace the current graduated built property tax. Total revenue would rise fractionally, but the tax system
would become more progressive, with the tax Gini rising from 0.474 to 0.480. Further details are
provided in the table.
3. The Prime Minister’s Proposal of 2003
Although it is no longer actively under consideration, the Prime Minister proposed, in 2003, a
radical set of tax changes that would shift the weight of taxation away from income and onto expenditure.
Specifically he suggested raising the VAT from 10% to 16% and the corporation income tax from 15% to
20%; at the same time several taxes would be abolished, including those on wages and salaries, built
property, and inheritance, as well as some minor excises, and educational fees.
We estimate that the plan would have raised tax revenue by 5.8% in 2003, an increase of
LBP263bn. Although the plan would have considerably simplified tax administration, it had two
drawbacks. First, it did not raise enough additional revenue to be convincing to the Paris II donors, who
wanted (and still want) to see a more substantial effort to achieve fiscal balance, and would not have
favored the elimination, even temporarily, of taxes on personal income. Second, the proposal would have
fallen most heavily on the poor, who spend on consumption (and so would be hit by the higher VAT) but
are less likely to pay taxes on wages or capital income (and so would not benefit from the reduction in
those taxes.
Exercise 2: A Revenue-neutral equity-enhancing tax reform
Categories
Total tax revenue, LBP bn
Of which:
Wage and salary tax
Sole proprietorships
Corporation profits
Dividends & capital gains
Inheritance tax
Built property tax
Property registration fees
Other taxes and stamp duties
Memo: Gini coefficient for all taxes
Old
4,502
173
119
295
51
37
82
203
305
0.474
Change in real expenditure
LBP/cap p.a.
% change
-44
-0.59%
-40
-0.66%
-49
-0.70%
-11
-0.29%
-9
-0.25%
-9
-0.24%
-17
-0.43%
New
4,534
184
107
393
0
0
201
122
289
0.480
Change, LBP bn
32
11
-12
98
-51
-37
119
-81
-16
% change
+0.7%
+6.3%
-9.7%
+33.3%
-100.0%
-100.0%
+146.0%
-40.0%
-5.2%
Revenue Loss
Parameter
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Change in real expenditure
LBP/cap p.a.
% change
-0.52%
-27
All Lebanon
Beirut
Suburbs of Beirut
Rest of Mt. Lebanon
Poorest quintile
-1
-0.06%
North Lebanon
Poor-mid quintile
-3
-0.10%
South Lebanon
Middle quintile
-4
-0.09%
Nabatieh
Mid-upper quintile
-15
-0.26%
Beka'a Valley
Upper quintile
-113
-0.96%
Components of tax proposal:

Simplify tax on wages and salaries, and on the self-employed: 10% basic rate, rising to 20% on income above LBP60 million.

Remove: tax on dividends and capital gains; inheritance tax; minor excises (alcoholic & non-alcoholic beverages, entertainment tax, 5% sales
tax on small hotels); fees related to education (examination fees, registration fees for schools and University of Lebanon).

Raise corporation income tax to 20%.

Raise Built property tax to flat 0.5% of value, but cut property registration fees from 5% to 3%.
Exercise 3: Prime Minister’s Tax Reform Package floated in 2003
Categories
Total tax revenue, LBP bn
Of which:
Wage and salary tax
Sole proprietorships
Corporation profits
Inheritance
Built property tax
VAT
Fuel excises
Tobacco excises
Other taxes and stamp duties
Memo: Gini coefficient for all taxes
Old
4,502
173
119
295
37
82
1,361
819
185
305
0.474
Change in real expenditure
LBP/cap p.a.
% change
-73
-0.97%
-90
-1.48%
-107
-1.53%
-80
-2.04%
-77
-2.14%
-87
-2.23%
-70
-1.81%
New
4,765
0
93
393
0
0
1,923
777
175
278
0.452
Change, LBP bn
263
-173
-26
98
-37
-82
562
-41
-10
-26
% change
+5.8%
-100.0%
-22.0%
+33.3%
-100.0%
-100.0%
+41.3%
-5.2%
-5.2%
-8.7%
Revenue Loss
Parameter
1.0
1.0
1.0
1.0
1.0
0.8
Elasticity = -0.2
Elasticity = -0.2
1.0
Change in real expenditure
LBP/cap p.a.
% change
-1.63%
-84
All Lebanon
Beirut
Suburbs of Beirut
Rest of Mt. Lebanon
Poorest quintile
-39
-2.18%
North Lebanon
Poor-mid quintile
-61
-2.15%
South Lebanon
Middle quintile
-82
-2.10%
Nabatieh
Mid-upper quintile
-100
-1.79%
Beka'a Valley
Upper quintile
-136
-1.16%
Components of tax proposal:

Raise VAT rate from 10% to 16%

Raise corporate tax rate from 15% to 20%

Remove: tax on wages and salaries; built property tax; inheritance tax; professional tax; minor excises (alcoholic & non-alcoholic beverages,
entertainment tax, 5% sales tax on small hotels); fees related to education (examination fees, registration fees for schools and University of
Lebanon).
[email protected]

similar documents