Social Security Issues and the Tax Reform Debate

Social Security Issues and the Tax Reform Debate

124 Economic Analysis and Policy Vol. 15 No. 02, September 1985 SOCIAL SECURITY ISSUES AND THE TAX REFORM DEBATE Daryl Dixon , Ch~is Foster and Ph...

3MB Sizes 0 Downloads 36 Views

124

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

SOCIAL SECURITY ISSUES AND THE TAX REFORM DEBATE

Daryl Dixon , Ch~is Foster and Phil Gallagher ' Policy Co_ordination Unit Canberra , A.C.T.

1.

INTRODUCTIO N

The tax reform debate of the past year has dwelt on deficiencies i n the personal income tax base, such as the non-taxation of capital gains . These deficiencies are believed to have been exacerbated by high marginal tax rates and the net reault has been a loss of confidence in the tax system and reductions in both efficiency and eQuity. A major strategy which has emerged in response to this situation and the commitments made under the so_called Trilogy' in 1984 is to make a switch from direct to indirect taxes . A ke y consideration in making a switch is that the poor should not be made worse-off. There is no doubt that a redistribution of income can be made to those at the bottom of the income distribution which would, in broad terIM, compensate for a s witch . However , there is a range of issues which would not be addressed in a pure substitution strategy and which have a major effect on equity as well as efriciency and simplicity. These include : (a)

overlap of the tax and social security systems which can create unintended effects such as so-called poverty traps; and

(b)

tax expenditures such as concessions for occupational superannuation which tend to have regressive effects on the distribution o f i ncome, and the dependent spouse rebate which, while being progressUve oversll, is of little bene f it to some married couples whose low income means that they pay insuffic i ent tax to obtain the !\Ill benefit of the r e bate.

Sec tion 2 of this paper examines the 8ub8ti tution strategy from a social security perspecti ve, concentrating in partioular on issues relating to compensation . In doi ng this, previou8ly unreleased Australi an Bureau of , Daryl Dixon is Head of the Government ' s Policy Co-ordination Unit i n the Ministry of Community Services. Chris Foster and Phil Gallagher are both Policy Adv isers in the Unit, which was formerly known S8 the Social Welfare Policy Secretariat . Any view8 expressed in the paper are t hose of the authors and are not necessar ily those of the Unit or of the Governlllent . The Trilogy refe rs to the policy commitment gi ven during the 1984 Election campaign by the Prime Minister that Government expenditures, taxation , and the deficit should not increase as a proportion of GDP .

125

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

stCltistics ( ABS) data from the Household Expenditure Survey (HES) are analysed to identify the characteristics of the groups who may fare worst in any switch from direct to indirect taxes. 2 Section 3 examines issues relating to overlap of tax and social security, including the so-called poverty trap and proposals to eliminate such problems . Section II examines variOUS tax expenditures , including the dependent spouse rebate . Previously unavailable data are analysed producing resul ts whi ch are qui te surprising compared t o conclusions often reached by others. The paper concludes .... ith the obse rvation that compensation f or a s .... itch from direct to indirect taxes is possible but that there are broader welfare and tax issues whi ch should not be ignored .

2.

INDIRECT TAXES AND COMPENSATION

It is generally concedecl that a switch from direct to indirect taxes would increase the overall burden .") f taxation on those with lower incomes . This is because the main items currently excluded from the wholesale sales tax are such things as food, clothing, and various services . Expenditure on food in particular represents a high proportion of the incomes of lower income groups than the higher i ncome groups. This can be seen in Table 1. In order that the overall burden of taxation is not shifted towards the bottom of the income distribution it would be necessary to make adjustments to either or both the personal income tax and SOCial security payments. The existing income tax system cannot adequately fulfil the compensation role because social security recipients, who comprise the bulk of the poor , tend to be below the tax threshold. The social ~ecurity system is better suited to directing assistance to those most in need because it is based on the "nuclear" family unit and its definitions of inCOme tend to be somewhat stricter and more comprehensive than those of the tax system. Some important examples of instances of stricter t reatment include the treatment of maintenance, annuities, and income accruing in Approved Deposit Funds . In addition to this the social security system also takes into account some assets in determining pension (but not other) entitlements. In looking at the compensation issue there are at least three main Questions that have to be cons1deretJ. They are: (a)

Who are the low income units?

(b)

How do the consumption patterns of these units differ from those of other income units?

(c)

Can existing arrangements compensate low income units?

2

This paper was finalised prior to the release of the White Paper so the issues raised in it are based on the authors' guesses as to the direction it would take .

126

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 1 EXPE NDITURE PATTERNS Of SELECTED(a) LOW I NCOME HOUSEHOLDS COMPARED TO ALL HOUSEHOLDS ( pe r cer.t of expenditure) Expenditure Group

Household Head's Income Unit Type Married Married Single Sole Pensioner Couple Person Parent CouPle(d) .... ith one on on Benefit(b) penSion(c) child on Bene f it(e)

CUrren t Hous i ng Costs fue l and Po .... er food Alcohol To bacco Clothing & foot .... ear Medical & Health Othe r Expenditure (f)

TOTAL EXPENDITURE $

33 . 7'

3. 1' 17 . 6

10 . 2 4. 9 26 . 3

21 . 9'

4.0' 26 . 4 2 . 9'

12 . 9 2.7 20 . 1 3. 2 1. 6

5. 8'

0. 6

2 . 2' 4. l '

2. l '

2· 3 2.8

6 . 4'

6. 1

3. 6

1. 1

29 · 3

32.6

2. 9 44 . 5

34 . 6

4.0 48 . 9

100 . 0

100 . 0

100 . 0

100 . 0

100 . 0

121 . 47

159 . 64

165 . 09

196 . 52

355 . 13

66 . 09

130 . 62

146 . 87

161.63

AVERAGE WEEKLY I NCOME $ Notes:

26 . 2' 4. 7 26 · 3

All House holds

4.5' 3. I'

2. 6

6. 6

Estimate has a relative sta nda r d error bet .... een 25 and 40 per cent. (a) The i ncome ranges for these households have been seleoted so that there is a high probability that only one income unit (nuclear family) constitutes the househO l d . The income ranges and other infof"lllation on the groups is given in footnotes (b) to (e) . (b) Pr incipal s ource o f household i noome was unemploymen t o r sickne s s benefi t and a nnual household i n co~e .... as $4,000-$5 , 999 . (c) Pr i nc i pa l sourc e o f household i ncome ....as Supporting Parents Benefitll o r Widows Pen Ilion a nd a nnua l household income .... as '

$6 , 000-$7 , 999 . (d) (e) (f)

~:

Principal source of household income .... as a pension othe l' than those f or sole parents (e . g . Age, Invalid, Service , Disability pension) and annual household income was $6,000- $7 , 999 . Principal source of household income was unemployment or sickness benefit and annual household income was $6 , 000-$6 , 999 . for transport , fU rn iture and equipment , household services and operation , r ecreation , personal care and miscellaneous goods and services .

Un published pr elimi na r y 1964 Household Expenditure Survey data .

Economic Analysis and Policy

127

Vol. 15 No. 02, September 1985

Anybody who is familiar with social and economic theory and existing data sources will appreciate that answers to these Questions cannot be unequivocaL Some of the issues surrounding each of these Questions are considered below. For a more complete discussion see Cox and Foster (1985) and Harding and Whiteford (1985) . 2.1

Problems with Defining Low Income

There are difficulties in defining the Circumstances in which persons can be said to have the same capacity to pay taxes or the same need for social security payments. These issues also apply to the question of those who should be compensated. Income as defined by the Assessment Act is not a reliable indicator of a person's we i t is not comprehen::live . Some per::lons with low taxable incomes may enjoy a much higher ::Itandard of living than one would expect on the basis of their reported income. A higher standard can be aChieved by a person who receives income in a non_taxable form or is able, as is a self-employed person, to make use of various deductions, ::Iplit incomes, or use business resources for domestic purposes. Non_disclosure of income to the Tax Office and the Department of Social Security is also a problem with using administrative income statistics, although the same problems undoubtedly arise in the CensUs and population surveys. Apart from problems with the administrative definlt ion of income there are a number of more general issues whlch influence the design of tax and social security systems. For e xample , the incomes of persons are not fixed al!! their circumstances can chan[;e over the course of a year. This is the case with school leavers, migrants , and persons who become unemployed or change their marital statu3. The 30cial security system is responsive to these changes because it is based essentially on current income but the tax system works on an annual ba3is so that a person who has only worked for three months would receive the ::lame tax threshold as someone who worked the full year. A person '3 Circumstances also change in a life cycle context where their earnings tend to increase with training and experience. There can be interruptions to thi3, for example, when a woman leave::l the workforce to raise a family. The appropriate tax or social security treatment of women in this case is a matter of considerable debate as the debate on the dependent spouse rebate illustrates. So far the compensation debate has proceeded on the basis that increased indirect tax payments should be offset by increases in disposable income. That this is desirable fol' low income units, especially pen::lioner::l and beneficiaries, is widely accepted. However, there are several problems with such an approach apal't fl'om defining what is a low income and how much each pel'son or income unit pays in indirect taxes. First, it assumes that the existing distribution of income is desirable. It also assumes that low income i::l a reliable basis for making compensation payments. In genel'al it is, but other factors, such as the presence of children in an income unit, may be a bettel' baSis for allocating compen3ation. This can be seen in the work on equivalence scales and poverty lines of the Social Welfare Policy Secretariat (1981) and Dixon and Gallagher (1985) which demonstrates tha t those in the community with the greatest needs relative to income::l include sole pal'ents and the unemployed.

128

Economic Analysis and Policy

2.2

Vol. 15 No. 02, September 1985

Expenditure Patterns

Households with the lowest incolDos spend 8. greater proportion of their i ncome on food, current housing (rent , lIIortgage paYlllents , ra tes and insurance) and fuel and power than households in general (See Table 1) . Howe ve r, there Bre cons iderable differences within the low incollle group . This is to be expected when one conside r s that the r e are SOllie 3. 1 million pensioners and beneficiaries. Many of these are retired but there are large numbers of unemployed and sale parents and these tend to have quite different expend iture patterns. A further issue which ,e merges in a consideration of the 19811 preliminary HES data ( s e e Table 1) is that the expenditures of sOllie households exceed their repo r ted incoroes . This is the phenomenon known as dissa ving . Dissaving is to be expected when one considers that both incoroe and needs vary over one's life cycle . The extent of dissaving identified in the HES, however , appears to be considerably exaggerated . There are a variety of factors which may contribute to this inc lud in g the cons umption of accumulated savings , borrowings , under _repo rting of income , a nd the way in which HES data is co l loeted and tabulated. Thc relati vo iIIIportcmco o f caeh of t hcoc :io unknown . Strictly speaking , the tena dissavins is only an appropriate description i n the case of consumption out of accumulated savings . This would include bank deposits, lump sum payments , and the sale of a house or othe r assets . Furth er roor e, sOllie dissaving would be planned and not i ndicati ve of low incomes . This would be the case, for e xample, with an aged :income unit which is consuming savings i t set a s ide for retirement . Other dissaving, however, would be unanticipated s uch as where a person becolllos unemployed . 2.3

Adequacy of EXisting Arrangelllents

It is possible to identify many i nstances which illustrate the difficulties in providing exact compensation . Several of t hese have al r eady been mentioned , na.roely, diffiCulties in defining those w1th low incomes a nd differences in e xpenditure patterns . In the current debate pensioners and beneficiaries are being given the highest priority for compensation . In this case indexation by the Consumer Price I ndex (C pO will give reasonable protection for those covered by indexation but problems could arise in the context of : (a)

lags between price rises and pension inc r eases ;

(b)

marked differences in beneficiaries and the determi ning the CPI; and

(c)

payments which are not indexed .

the consumption patterns of pensioners and average o f Au a.tra l ian households used in

Payment s 1n respect of chi ldren, s upp lementary (rent) aSSista.nce , the junior rate of unemployment and sickness benefi t , and the fixed r ate of pension for persons aged 70 and over are the major paymenta. which are not i ndexed. Pensioner and beneficiary househOlds in pr ivate rental accommodat:ion and with children would therefore be the major loser s unless such payments were also increased.

129

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

Ileplacing the existing multi-rate wholesale sales tax with a single-rate broad _based retail sales tax would have differential effects on prices. Some existing items would increase in price whereas others could either fall in price or remain the same . Pensioners and beneficiaries could be disadvantaged by CPI indexation if the price of the goods and services they consume increase compared tc> some "luxury" goods which could fall in price. A further issue is that current expenditure patterns undoubtedly differ from the weights in the CPI which were established in 1975-76 . Pension and benefit indexation ar rangements operate on a siX-monthly basis so there could be a delay between price increases and compensation. The seriousness o f this would depend upon the speed and magnitude of price increases. It could be necessary to make some form of prospective adjustment to social security P8Yl11ents to cover the period up to normal indexation. A major issue which has emerged in the compensation debate is that there are per"on" who IIIll y not be adequately compensated by either t.hfl ta x or SOCial aecurity systems. These are most likely to be single income unita and married units without children not in receipt of income support payments where taxable incomes are too low for them to benefit fUlly from income tax cuts. So far the debate has proceeded in the absence of detailed quantitative evidence. Tables 2 to 5 , derived from the unit record tape of the 1981_82 ABS Income and Housing Survey, are a significant addition to available information. The numbers in the tables are only indicative because they are derived from the 1981_82 data and bec.!luse the definition of low taxable income will depend on the magnitude of the tax cuts proposed for the various groups in the White Paper. Nevertheless, several important points emerge from the anslYBis, i ncluding: Single per Bon income ur.its comprise the bulk of income units not fUlly in either system (about 66 per cent). Nearly 70 per cent of these Single people are under 25 years of age (see Table 4) and only 37 per cent were in the workrorce for the fUll year (see Table 5). Most had either just entered the workforce part way thro ugh the year or had periods of unemployment during the year (about 23 per cent received a pension or benefit at some stage during the year). (b)

Most young single person income units share accomod.!ltion with others (about 95 per cent), gcnerally their parents (see Table 4). This means that their income needs are probably lower than i f they were living alone o r away from home. I t should be noted that hOUsing rents are unlikely to be included in an indirect tax.

(c)

~~~reiniS t:e si~nbi:;~anftorn:emb~rur~fngSi~~~~_~~r~~~ i:~:m~n unr~~~i~~o government pension or benefit (see Table 5). Sollie of these would seem to be those receiving a superannuation pension or similar fixed income. However, "ome report no i ncome for the yoar which i. dlffioult to understand except if they are living at home and choosing not to obtain income support.

(d)

The re are some sole parents who are not receiving a pension or benefit and who are not fully in the tax system (see Table 3) but compensation

:r:

130

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

to this group can be directed through payments for children as well as increases in the sole parent re bate or othe r tax cuts. (e)

Married income units outside the social security system and marginally within the tax system have totally different c haracteristics to single income units in the same position (see Table 2). They are mostly employed rull year and full time and are spread fairly evenly across all age groups . Whereas single person income units in this situation are pre dominately wage t~d salary earners (55 per cent), married income units are predominately selfMemployed (66 per cent). These tend to be farmers, tradeamen and those in the transport industry.

(f)

The majority of married income units in this aituation have dependent children (68 per cent). Thoae with children wo uld be less adversely affected than those without children (at the same equivalent income ) because the social security system makes seve ral paymenta in r espect of children which are available to persons in the workforce (including the selfMemployed). These are family allowances which are paid on a non-income tested basis and family income supplelllent (nS) which is paid on an income tested basia .

Several major poliey issues tl. ri88 from t.hA ft')rl!in1na: . The first is the extent to which it i8 considered necessary to compensate persons who are at an early stage of the lifecycle and those who have apparently chosen not to olaim unemployment benefit between jobs. Those receiving low youth wages include app r entices, farlll labourers, and 3hop assistants. This is to Borne extent a transitional phase although low wages will continue for those working in casual or partMtime jobs. Also a low paid low skilled wage and salary earner could marry and not receive the full value of the dependent spouse rebate when they decide to have children. Increasing the rate of unemployment benefit to compensate for a change to indirect taxation could preaumably increase take_up by those with recurrent spells of unemployment or marginal attachment to the labour force. A second issue is whether the appropriate benchmark for COmpensation is current or annual inCOme. As Table 5 shows, many single person income units are only in the workforce for part of the year and both their current and period (annual) income may not adequate ly reflect their true standard of living. A third issue is whether low taxable income is a aatisfactory basis for providing COmpensation given the dominance of the selfMemployed amongst married income units outside the tax and 80cial security systems. It would be in 80me cases because there i s no doubt that the selfMemployed are also 8u8ceptible to poverty. I n any event most of the married income units have ohildren and SOme tax liabilities which mea n that they are able to be compensated at leaat partly through existing meClhani:5ms such as the family income supplement or the tax syatelll.

131

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 2 NUMBER AND PROPORTION OF MARRIED COUPLE INCOME UNITS NOT FULLY IN TAX OR SOCIAL SECURITY SYSTEMS(l)

('000 in 1981-82) Number of Children 1 2

Head'3 Current Employment Status

None

Wage & Salary Earners

11.4

6.2

11.4

7.6

36.6

Self- Employed , Own Business

37.8

31. 0

39.1

35.2

143.1

• •

3.3

12.4

".3

25.9

50.5

218.0

".9



Not in Labour Force

16.7

3.2

Total

70.7

43.2

Unemployed

Proportion of all Married Couple Income Units

• (1)

Source:

301"

Total

more

7.31

53.6

6.5J

9.4S

6.4S

Relative standard errol" over 40 per cent • Impu ted tax liability under $ 1,000 and less than $1,000 of income from government pensions, benefits and TEAS in 1981-82. Only units with a measured 1981-82 annual income are included in the Tables. The tax imputation was made on the basis of a taxable income which excluded tax exempt Government cash payments but included 5 per cent of the value of lUmp sum superannuation payments received that year. The Sole Parent Rebate and the Dependent Spouse Rebate are used in the imputation along with the 1981_82 tax scales.

AnalysiS of ABS unit record tape from 1981_82 Income and Housing Survey.

132

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 3 NUMBER AND PROPORTION OF SOLE PARENT INCOME UNITS NOT FULLY IN TAX OR SOCIAL SECURITY SYSTEMS(1)

( ' ODD in 1981-82)

Head's Current Employment Status

Wage and Salary Earners Self Employed/Own Business Unemployed /Not i n Labour Force

Total

Number of Children Ono 2 or More

Total

6. 8

5.5





3·9

5.6

'.3

9. 9

14.7

11.'1

26.1

12.'1

Proportion of All Sole Parent Income Units

• ( 1)

Source:

Relative standard error over 40 per cent • Impu ted tax liability under $1 , 000 and less than $1,000 of income from government penSions, benefits and TEAS in 1981-82.

Analysis of ABS unIt record tape from 1981-82 Income and Housing Survey

Economic Analysis and Policy

133

Vol. 15 No. 02, September 1985

TABLE II NUMBER AND PROPORTION OF SINGLE PERSON INCOME UNITS NOT FULLY IN TAX OR SOCIAL SECURITY SYSTEMS, BY AGE , SEX AND LIVING ARRANGEHENTS(l)

('000 in 1981-82) Under 20

20_211

25-311

35- 114

115-511

Over 55

Total

Ha'e Living Alone With Others



96 . 8

5. 8

8.7

80.6

27 . 8

,.5. ,,

".8 8.7

38.6 221l . 6



5.7 7.3

21. 2 15.1

110.2 180.8

7.1

".1 12.2

33 · 0 23 . 8

78.8 1105.11

56.8

1l81l.2

'"

5.,

Fema1e Living Alone With Others



'.3

'.0

50.5

15.8

6. 1

182 . 7

10.0 131.1

12.6 43 . 6

12. 0

Total

187 . 6

11l1.1

56 . 2

19 . 1

Proportion of All Single Income Units

1l7.6%(2} 18.1l%

Total Living Alone With Others

Notes:

Source :



85.9

,.,

10.1%

11.8%

16 . 3%

Relative standard error over 110 per cent •

(1)

Imputed tax liability under $1,000 and less than $1,000 of income from government pensions , benefits, and TEAS i n 1981 _82. Only units with a measured 1981-82 annual income are included in the Tables.

(2)

Excludes dependent children and students.

Analysis of ASS unit record tape from 1981 _82 Income and Hou5ing Sur vey

134

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABI..E 5 NUMBER OF SINGI..E PERSON INCOME UNITS NOT FUI..I..Y IN TAX OR SOCIAl.. SECURITY SYSTEMS, BY AGE AND I..ABOUR FORCE STATUS(l) ('000 in 1981-82) Ase of Income Unit Head Period Labour Force Status

Under 20

Full Year Full Tim.

Total

20~211

25- 311

35- 44

45- 511

88.11

26.0

11. 5

6. 6

6. 6

8. 7

147.9

Part Time Full Tim.

74.5

69 . 2

28.5

•••



5. 3

183 . 8

Full Yea r Part Tim.

5. ,

6.6

3. ,





18.8

Part Year Part Tim.

9.5

16 . 2

5. 5

• •



3.6

39.5

3. 9

5.'









11. 5

3. 6

'.3









13.2



13.7

' .9

,.,

8.5

35.9

69 . 5

111 1 • 1

56.2

19.1

23 . 3

56 . 8

11811.2

Full Year Full Part Time

,,'

Full and Part Year

Unemployed Full Year Not 1n

Labour Force

Total

• (1)

~:

OVer 55

Relative standard error over 40 per cent • Imputed tax liability under $1,000 and less than $1 , 000 of income from government pensions, benefits , and TEAS in 1981-82 . Only units with a measured 1981-82 annual income are 1ncluded 1n the Table .

Analysis of ASS un1t record tape froD! 1981_82 Income and Hous1ns Survey

Economic Analysis and Policy

135

Vol. 15 No. 02, September 1985

One way of compen!';lating income units outside the social security system lo/hich has been ::suggested is to pay a refundable tax credit. However, it would be difficult to direct assistance to those in this situation because of uncertaintie::s surrounding the definition of income amI the duration of employment. In particular it would be difficult to deal with people whose circumstances change throughout the course of the year in any system where compensation did not significantly lag behind increased expenditure. It would also be difficult to deal with those not in the tax system. There are other avenues which can be explored, however, including specific labour market initiatives directed at those in low skilled or low paid occupations.

3.

OVERLAP AND POVERTY TRAPS

The tax and social security systems overlap because most pensions and benefit3 arc included in taxable income. Payments that are excluded include invalid pension, pensions paid to the blind, additional payments for children, family allowances, and supplementary (rent) assistance. Pensioners and beneficiaries also receive an income_tested tax rebate. This pattern of exclusions, inclusions, and rebates produces some high effective maJ"ginal tax ratE'S in combination with social security income tests. These can be seen in Table 6. It should be noted that problems also arise in the context of the supplementary (rent) income test, the sudden-death cut-out for fringe benefits and the overlap with income tests administered by other Commonwealth, State and local organisations. In the tax system the highest rates are faced by the higher income earners, yet high rates are faced by social security recipients with compaJ"atively low incomes . Nor do the rates rise smoothly; rather they rise and fall as indicated in Table 6 . Some pensioners and benefiCiaries may be in income ranges where the choices they face can mean that employment may not be an attractive option because it could result in a drop in disposable income. This is the so_called poverty trap. Some sole parents in particular could find themselves in this situation when the costs of child care are taken into account. I t is unlikely that the numbers involved are very large although this may reflect non-disclosure of income rather than the absence of income . The solutions to the problem are likely to be complex and require a package of lower tax rates, child care subsidies and labour market training. High effective marginal tax rates are not the only cause of non-disclosure of income for tax and social security purposes. Another is that the operation of the tax and SOCial security systems is difficult to understand. This could change in the future but at the moment it is not unusual for aged women who have had few dealings with the tax and social security systems to become widowed and find themselves having to deal with such problems for the first time in their lives. tlnderstanding how the pensioner tax rebate works or remembering to include the pension, which is not subject to PAYE withholding, in taxable income may not always be easy.

136

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 6 EFFECTIVE MARGINAL TAX RATES FOR PENSIONERS AS AT NOVEMBER 1984 , EXCLUDING MEDICARE LEVY

Single Pensioner

Pr ivate Income ($pw)

Effective Tax Rate

W

0 - 15.70

0

Married Pensioner Couples

All Income bl One Person Private Effective I ncome Tax Rate ($pw) m o - 30 . 95

0

Income Egualll SElit Private Income ($pw) o - 50.00

Effective Tax Rate

m 0

15.71 - 30.00

37 . 50

30 . 96- 50.00

37 . 50

50.01-73.78

50.00

30.01-78.32

68 . 75

50 . 01-15 . 88

18 . 125

73.79- 221.64

68 . 75

78.33- 213.8

62.50

75.89-79 . 52

68 . 15

227.65-356.60

62 . 50

79 . ~3- 201 . 611

62.50

201.65-3311 . 92

66 . 25

3311 . 93-356.60

72.50

Source:

Department of Social Security .

Wh ile being the unavoidable outcome of pursuing selective or needs based socal security policies it seems that there is scope for the simplification and reduction of the marginal tax rates . The chOices for r educing the rates fall between two approaches . The first is the approch recommended by the Commission of Inquiry into Poverty (1915) which is to integrate tax and social security . This would involve replacing e xisting social security payments with a minimum income guarantee and maki ng adjuBtmentB fo r differing circumstances such as family size and whether the person iB living alone and renting . There would be two rates of payment, a low rate fcr general taxpayers and a higher rate for persons in the SOCial Becurity system. Under this app r oach all !.ncome except the minimum income guarantee (which would act as a tax threshold) would be taxsble at a proportional rate . The overall system would be made progressive by the minimum income guarantee .

137

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

Achieving a uniform marginal rate of tax for both income support recipients (negative taxpayers) and general taxpayers (positive taxpayers) would require an increase in the average rate of tax. At the time of the Poverty Report the tax rate which would have been required was ilO per cent . Since then there have been some significant developments which would have increased the proportion of the population receiving the higher rate of income guarantee . These are the dramatiC increase in unemployment, growing numbers of the aged, and the increased incidence of sole parenthood. The result is that the 40 per cent rate would be inadequate today. It would aleO have been inadequate in 1975 unless a comprehensive income tax base could have been introduced. In other words, all tax expenditures would have to have been eliminated and capital gains and fringe benefits would have to be taxed. The ra te 0 f tax 1:5 not the only or indeed the major difficulty with this approach. Other major problemll relate to system dellign, and in particular definitionll of income and the income unit . Integrated lIystems have to solve the problem of overpayments to those with fluctuating incomes and to those who are able to understate income to receive a higher payment. comprehensive definitions of income and generalised withholding are prerequisites, but to date no country has introduced a fully integr ated syetem. t'le other approach to Simplifying and lowering effective marginal tax rates has been the subject of earlier papers by Dixon and Foster (1983, 1985) . It recognises that the design problems of an integrated syetem are virtually intractable . The approach they develop b to remove overlap of the two systems by exempting pensioners and beneficiaries from paying tax , to have special arrangements for allOwing persons to enter and leave either system during the year, and to have a generalised withholding tax to improve the administration of the tax system. The bssis of the present approach is that the social security system operates like s tax system for those entitled to a cash payment. Currently the level of the entitlement is reduced as private income rise II above a threshold known as the free area. The rate of withdrawal (or abatement) is 50 per cent for pensioners and up to 100 per cent for beneficiaries. The separate tax administration then subjectll the total income of the 1ndividual including the above cash payments, if appropriate, to tax. The income test can be structured to produce roughly the same resul t as the combined tax and income tests. A lowering of rates, however, requires 1ncreased outlays. In order to direct the benefits to those who current l y receive little or no advantage from either the free area or the tax threshold it would be necessary to cash out the free area. This can be done by paying a cash grant or refundable tax credit. The value of this payment would decline with income . At the same time it would be possible and highly desirable to relltructure the income test. The two income test rates ourrently applying to pensioners (0 and 50 per cent) could be replaced by a 25 and a 60 per cent rate. These measures would increase the level of · the pension for those with little or no private income but reduce the tax advantages from lower effective marginal tax rates for higher income pensioners.

Economic Analysis and Policy

138

~.

Vol. 15 No. 02, September 1985

TAX EXPENDITURES

Apart from compensation for indirect tax increases and poverty traps the other major aspect of the income tax which can effect the social security system is tax expenditures. Defining a tax expenditure is difficult , but for convenience a tax expenditure can be considered to be any part of the tax system which could be replaced by a direct expenditure. Under this definition the tax threshold could be included even though it is generally recognised as an essential feature of any tax system. Its inclusion, however , is not inconsistent given that the tax threshold could be provided as a cash transfer (i . e., refundable tax credit) as would be the case in the Henderson Guaranteed Minimum Inoome Scheme (Commission of Inq u iry into Poverty , 1975). Indeed, most who would obtain a net benefit from a refundable tax credit in Australia are already in receipt of a pension or benefit . Tax expenditures are of concern because they tend to be regressive in their incidence and are rarely .subjected to scrutiny in the Budget context. They include concessions given to ocoupational superannuation , a host of provisions benefitting agriculture and business, and the dependent. spouse rebate . COllllllents on two of these, occupational superannuation conoessions and the dependent spouse rebate, follow. ~.1

Occupational Superannuation

Much has been written on the tax expenditure.s f o r occupational superannuation (Social Welfare Policy Secretariat , 1982) as it is quite a controversial area. The central issues can be summarised as: (a)

valuable Commonwealth concessions are given to occupational superannuation, notwithstanding the increased lump sum taxes in 1983,

(b)

not everybody has acceSf. to an employer_provided scheme ; and

(c)

benefits are generally paid prior to normal age ret i rement .

If the aim of occupational superannuation is the provision of retirement incomes it would follow that the concessions could be replaced by higher universal age pension payments. However, it is not as simple as this . First, the estimated value of the revenue forgone in the Budget Papers of $2.6 billion for 1984- 85 (see COllllllonwealth Treasury, 198~, p.319) would nct be available for transfer. PreSUmably, some would be diverted into other lightly taxed areas. Second, occupational superannuation fulrills (imper[ectly it should be noted) an income maintenance function and over two million workers are members o[ a superannuation scheme . Third, it could be argued that the preferential savings status of occupational superannuation is a recognition of the disincentives to long term sa ving created by a progressive income tax in a period of inflation where the tax base and tax brackets are not indexed. Although this indicates that abolishing the tax concessions would be difficult, i t is increasingly being recognis ed by the pri va te sec tor, unions, government and private individuals that improvements are essential .

139

~.2

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

Dependent Spouse Rebate (DSR)

The dependent spouse rebate has traditionally been viewed as providing recognition of the reduced capacity to pay tax o f a taxpayer with a depende nt spouse compared to a two-income couple. Two-income households have the advantage of two tax thresholds whereas a single_income married household ha" one . The DSR does not f\llly take the place of the tax threshold but this is presumably a recognition of the higher costs involved in wo rking compared to staying at home. In practical terms the value of the rebate is much lower for higher income househol
3

Since the 198~-85 Budget the dependent spouse rebate has been made available to couplea in a de facto relationship.

140

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE '7 PERCENTAGE OF MARRIED INCOME UNITS RECE I VI NG NO, PART OR FULL DEPENDENT SPOUSE REBATE, BY COMB INED INCOME CLASS , 1981-82

Oroaa Inoome Husband a nd Wife

Married Income Units ( , 000)

($pa)

Under

In Receipt of a Rebate ( 1) (%) None

Part

Full

Total

,5,000

102 . 3

95 . 2

'-8

0

100 . 0

5,000- 9,999

567.4

70.2

18.9

10 . 9

100.0

10,000-111 , 999

487 .2

32.6

21. 5

115 . 9

100.0

15,000-1 9 , 999

629 .5

31 . 7

25.2

43 . 2

100 . 0

20,000-24 , 999

537 . 0

51.0

20 . 2

28 . 8

100.0

25,000-2 9, 999

1119.8

71.4

13.0

15 . 6

100.0

30,000- 311, 999

264.5

77 .4

12.1

10 . 5

100 . 0

35 , 000 and over

388 . 0

83.8

9.1

7.1

100 . 0

3395.7

57 . 6

17.8

24 . 5

100.0

TOTAL

~:

~!.:

( 1)

Rebate imputed from annual taxable incolDe data for husband and wife .

Analysis of ABS unit record tape frOID 1981 -82 Income and Housing Survey

141

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 8 PERCENTAGE OF MARRIED InCOME UNITS WITH DEPENDENT SPOUSE , BY AGE OF WIFE AND AGE OF YOUNGEST CHILD

Age of youngest Child

Age of Wife Under 20

20- 39

110-59



2.7

18.1

0-' Years

0. 2

32.0

, .2

5- 9 Years



13.6

3. 9

10_ 111 Years

' .9

7.'

15-20 Years

0.7

5.6

No Children

TOTAL

Notes : ~:

53.9

OVer 60

Total

9.3

30.1 33.5

• • •

17.6 12.3 6.'

100.0

• Relative standar d error o ver 110 per cent. Analysis o f ABS unit record tape f r om 1981 _82 InCOme and Housing Survey .

142

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

TABLE 9 VALUE OF DEPENDENT SPOUSE REBATE BY INCOME CLASS AND AS A PROPORTION OF COMBINED INCOMES

AD!ount of Rebate(1)

Gross Incolll9 of Husband & Wife ($pa)(2)

$Million

J of Total

Rebate a s a Proportion of Gross Incollle o f Married InCOlDe Uni ts

m

Under 5 ,UUU(3)

O. ,

0. 1

1.6

5,000- 9,999

89 . 2

10 . 0

•••

'0,000-'4 , 999

2 17.8

24 . 3

5.2

15,000-19,999

282 .3

31.5

3.8

20,000-24,999

165 . 0

18 . 11

2. 8

25 , 000-29,999

71 . 9

8.0

2.2

30,000-311,999

35 .3

3.,

1.8

Over 35 , 000

35 . 1

3. ,

1. 2

897 . 1

100 .0

3.3

TOTAL

~:

~I

( 1)

Rebate imputed from annual taxable income data obtained from 198 1-82 Income and Housing Survey.

(2)

The mean income (froID all source~!l of married couple incOlDc units in 1961-82 was $21,220.

(3)

Included in this table is a larger n\1Jllber o f low i ncome units than show up in the tax statistics . this could be due to non-filing of a return where no tax would be payable.

A.n.alysh Survey.

or

ASS unit. record tape from 1981-82 InC9IDe a nd Housing

143

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

Reasons for the differences in the results obtained from the 1981-82 Income and Hous ing Survey and statements to the effect that the rebate is very regressive (Harding , 198~ , P.37) included the fact that the current analysis is based on gross income unit income and not individual taxable income. It is also based solely on married income units in order to minimise problems with looking at redistributions taking place at different points in the life cycle . A single person or a sole parent would tend t o have a lower current mean income than a married couple but at some stage the majority of persons marry or form households and also receive increases in their incomes. Some of these life cycle issues are illustrated quite clearly in Table 8. For eXBlIlple, the largest single category of married income units with a dependent spouse are those where the wife is aged 20-39 and ha s a child under II years of age . Those without dependent ch ildren tend to be over 110 years of age. None of the preceding is nece ssarily an argument in favour of retaining the It could be con verted to a cash payment in the same way that rebates for children were converted to family allowances in 1976 . One of the reasons for doing so at the time was to give low income households the full value of the rebates . The same case could be advanced for the DSR as many low income households do not receive the roll benefit of the rebate.

DSR .

From a welfare perspecti ve there wou ld have to be compensatory cash payments available to low income married households without children if the DSR were abolished in order to aVOid increasing the incidence of poverty. Not all married women , for example, have the potential to enter or return to the labour force. Some as indicated have young child ren and some would have had little or no labour force experience. All this suggests that any equitable replacement scheme for the dependent spouse rebate will necessarily have to involve a comprehensive package of lDeasures , including the provision of child care , employment training, alternative social security payments, and changed tax incentives.

5.

CONCLUSION

This paper has examined several of the major issues which have been raised in the current tax reform debate. The major issue is , of course, whether it is possible to compensate low inCOme households for a switch to ind irect taxes. What this debate has ~hown is that the existing social security system is capable , in broad terms, of prOViding compensa tion to pensioners and bene ficiaries . It is also capable of providing some compensation to married couple income units with child r en who are in the workforce through payments such as family allowances and family income supplement . The debate has also drawn at.tention to some weaknesses in current tax and soc ial security arrangements which lDean that it is difficult to provide compensation to low income single person income units and married couples without c hildren. Analysis of 19~1-82 Income and 1I0using Survey data , however , suggests that many income units in this category were I:Iharing accommodation with others and either in the workforce ror part of the year or self-employed. For the8e groups there are significant difficulties in defining the appropriate level o f income and the income unit to be used in provid i ng COmpensation . This would also be true e ven if a spec ific

144

Economic Analysis and Policy

Vol. 15 No. 02, September 1985

refundable tax credit were i nt roduced to direct assistance to the "working poor". The issue of the "working poor" and overlap of tax and social security income tests raises the more difficult general problem of integration o f tax and social secu rity. One proposal (Dixon and Foster, 1985) would incorporate the non_taxation of pensioners and beneficiaries, increased rates of pension and benefit (by "cashing_out" the income test free area in the same way as a refundable tax credit works), a part_year tax system, and a generalised withholding tax. This aspect of the debate has been largely ignored as a r esult of the concentration on the compensation issue. The adequacy o f existing payments has also tended to be ignored. These two issues, however, are central to ongoing social security concerns.

REFERENCES Commission of Inquiry into Poverty , Australia, Canberra : AGPS. Commonwealth Treasury Cox ,

('98~),

(Henderson ,

R.)

(1975) ,

Poverty

in

Budget Paper No 1, canberra: AGPS.

J. and Foster, C. (1985) , "Tax Changes and Social Welfare ", paper presented at Conference on Changing the Tax Mix, Monash University, February.

Dixon, D. and Fo:!!ter , C. (1983), An Alternative Path to Integration of Social Security and Personal I ncome Tax Arrangem ent:!! . Sydney: Australian Tax Research Foundation, Occasional Paper No .1 . Dixon, D. and Foster, C. (1985), "Simplification of Social Security and Tax" , paper presented at 1~th COnference of EconOmists, Sydney , Hay . Dixon, D. and Gallagher, P. (1985) , "A Welfare Perspective on Labour Force Measurement", paper presented at Conference on Unemployed i n Australia - Perspectives on Measurement, ANU, Canberra, April. Har ding, A. (19811) , Who Benefits?: The Australian Welfare State and Redistribution, Social Welfare Research Centre, Reports and Proceedings No. 1l5 . Harding, A. and Whiteford, P. ( 1985 ), "EqUity , Tax Reform and Redistribution", Department of Social Secu rity Submission to Economic Planning Advisory Council, Research Paper No . 28, Canberra. Social Welfare Policy Secretariat (1981), Canberr a : AGPS.

Report on Poverty Measurement,

Social Welfare Policy Secretariat ( 1982), Alternative Strategies to Meet the Income Needs of the Age", Canberra: AGPS.