Personal Direct Tax Reform Proposal

Excerpts from the
Annual State Revenue Report for 1998

Chapter XVIII

PERSONAL DIRECT TAX REFORM PROPOSAL - 1998 (continued)

5. Funding the Reform

The expected cost of lowering the tax rates in the manner stipulated above, with no harm to disadvantaged population groups, came to billions of sheqalim, partly offset by cutting tax benefits. The technique chosen was to eliminate all existing benefits and was to create a new and minimal list of justifiable tax benefits. The benefits slated for elimination included tax exemptions for transfer payments, including National Insurance contributions and health tax (i.e., they would be taxed at regular rates), most exemptions on investment income, and benefits for shift labor and residence in development areas.
In creating the new list of benefits, the decision-makers' preferences were taken into account alongside purely professional considerations.
Below is the short list of tax benefits that were discussed when work on the reform began. It should be stressed this is an initial alternative that was eventually rejected, for reasons described later on.

a. Retirement Contributions

Today, depositors with provident funds are given tax benefits at various rates and by various methods, depending on the deposit plan (severance pay/compensation payments) and the source of income (wage included in the pension base or other work income). The tax system that pertains to social contributions is rather intricate and contributes much to complicating the tax system. The intricacies create a waste of money; preferential benefits for one plan over other plans result not in an extra yield for savers and an increase in national saving but in extra profits for those who offer the plan, along with management and marketing costs that do nothing to help attain the goals of the benefit.
The proposed reform will abolish the existing benefits for provident funds (including advanced-training funds) and replace them with a deduction of up to 17 percent (employee and employer combined) of income up to NIS 11,000 per month (gross). Any withdrawal of funds originating in an allowable contribution would be fully taxed, except for pensions, which would be given a 35 percent deduction up to NIS 6,000 per month, as is allowed today.
The advantage of this scheme is that payment of tax is deferred and may be charged at a lower marginal tax rate, e.g., if the worker has stopped working at the point of withdrawal. This benefit is meant to encourage saving for a "rainy day," and this is also the logic of giving double benefits to those who choose the pension track. The government does waive some of the tax but makes sure that workers who reach retirement age do not become a burden on its budget.

b. Limited Tax Rate on Investment Income

We considered the possibility of imposing a limited-rate tax on investment income, in such a way that the tax rate on gains from long-term saving would be lower than the tax rate on short-term saving. According to the principle of horizontal equality, these incomes should be taxed at the standard rate, i.e., up to 35 percent-the highest marginal tax rate, including National Insurance contributions and health tax. However, these incomes are highly mobile by nature, and it was feared that capital would flee the country if ordinary taxation were applied. Furthermore, the reduced tax rate was meant to stimulate saving and investments. Notably, taxation at a limited tax rate would have the side-effect of making it easier to collect the tax by withholding at source.

c. Conversion of Credit Points for Working Mothers into a Deduction for Child-Care Expenses

It was proposed to abolish the credit points that working mothers receive today and to replace them with a deduction, against receipts, of expenses for care of preschool children (up to NIS 1,000 per month).¹¹

d. Charitable Donations

In cognizance of the importance of the activity of public institutions, it was decided to retain the charitable-donation credit but to reduce its rate from 35 percent to 20 percent due to the planned decrease in tax rates.

e. Rent and Gambling Income

It was decided to tax these incomes at a limited rate of 10 percent as an alternative to a deduction for expenses-except for rent expenses, which would be offset against rent income in order to avoid detriment to residents' mobility.
After the list of benefits to be abolished or reduced was determined, it became possible to quantify the extent of the reduction in tax rates. We estimated that the elimination of all exemptions on the old list would expand collections by NIS 12 billion (assuming a decrease of 20 percent in the weighted marginal tax rate under 20 percent). From this sum, NIS 3 billion would be subtracted-the extent of tax benefits given in the new list. Thus, the cutback of benefits would permit a NIS 9 billion reduction in tax rates.¹²

¹¹One cannot give a deduction for child-care expenses and continue to give credit points, because this would result in double benefits.

¹²At the existing tax rates, the benefits in personal income tax came to NIS 15 billion (not including land-betterment tax on dwellings, income from abroad, the benefits on National Insurance contributions and health tax). Since the reform was supposed to lower the tax rates, by 25% in average, the value of the tax benefits would decrease commensurably.

This framework allowed us to compose various alternative tax tables. The example below was elaborated and discussed at the beginning of work on the reform:

Table XVIII-2 Direct-Tax Rates, after the Reform-Alternative 1

Brackets (NIS/month)	Direct-tax rates (percent)
	Income tax	National Insurance⁽¹⁾	Total
0-2,865	15	4	19
2,866-30,000	25	10	35
30,000+	35	0	35

Source: Economic Research and State Revenue Administration

Note to the Table:
(1) National Insurance contributions include health tax but without the employer's share.

Table XVIII-2 shows that the proposed tax brackets are rather flat: the number of brackets declined from seven under current law (see Table XVIII-10) to only two. The 19 percent rate on income is up to half the national average wage; above that, the rate climbs to 35 percent. In view of the two constraints that were adopted ab initio-a maximum marginal rate of 35 percent and revenue neutrality-the tax rates could not be made more progressive than that. The new table eliminates the anomaly of declining tax rates above the ceiling of National Insurance contributions and health tax. The National Insurance and health-tax ceiling is raised to NIS 30,000 but is not abolished.

Table XVIII-3 Effect of the Reform on Net Income, by Net-Income Deciles per Standard Adult- Alternative 1

(Average per household, NIS per month, 1998)

	Before reform					After reform					Change, net
Decile	Income tax	social security (1)	Total direct taxes	Gross income	Net income	Income tax	social security (1)	Total direct taxes	Gross income	Net income	(NIS)	(Pct.)
Avg	1,996	707	2,672	12,512	9,840	1,911	768	2,678	15,512	9,834	-6	-0.1
1	41	79	120	3,426	3,307	225	170	395	3,426	3,032	-275	-8.3
2	235	209	444	4,932	4,448	479	279	758	4,932	4,175	-313	-7.0
3	383	301	684	6,418	5,734	646	359	1,005	6,418	5,413	-321	-5.6
4	567	384	951	7,304	6,353	785	422	1,207	7,304	6,097	-256	-4.0
5	812	487	1,300	8,674	7,374	996	514	1,510	8,674	7,163	-211	-2.9
6	1,162	649	1,811	10,577	8,767	1,311	643	1,954	10,577	8,623	-144	-1.6
7	1,609	800	2,410	12,598	10,188	1,698	791	2,489	12,598	10,109	-79	-0.8
8	2,722	1,096	3,818	16,025	12,207	2,491	1,904	3,585	16,025	12,440	223	1.9
9	4,419	1,408	5,827	20,832	15,005	3,734	1,495	5,230	20,832	15,602	597	4.0
10	7,697	1,650	9,348	34,291	24,943	6,732	1,906	8,639	34,291	25,653	709	2.8

	Before reform	After reform
Cost of reform (NIS billions per year)		1.3
Poverty line (50 percent of median net income per standard adult in NIS per month)	1,445	1,462
Households under poverty line (percent)	13.2	13.3 ⁽²⁾
Households under poverty line (percent)	14.6	14.4 ⁽²⁾
Ratio of uppermost quintile to lowest quintile	5.13	5.03
Gini Index	0.327	0.322

a.	The proposal increases the tax burden of people in the lowest decile. Consequently, disposable income in this decile declines by 8 percent. Disposable income also decreases in deciles 3-7, but to a smaller extent than in the first decile.
b.	In contrast, disposable income in deciles 8-10 increases by 2-4 percent.
c.	Overall, the reform transfers NIS 3.3 billion from the seven lowest deciles to the three highest deciles.
d.	The percent of poor households increases by 10 percent. Since these households are larger in size than the average, the percent of poor individuals rises even more. Since the poverty line also falls, the income of the poor declines.
e.	The ratio of disposable income in the uppermost quintile to that in the lowest quintile, and the Gini index of disposable income, also point to a substantial worsening in distribution of disposable income.

These findings made it clear that another alternative, one that would cause no detriment to income distribution, should be sought. As the work continued, many alternatives were examined until the last one, shown below, was formulated.

6. Changes Made as the Reform Was Being Prepared

The reform had a rather revolutionary point of departure in terms of the extent of reductions in tax rates and benefits. As it was being formulated, it underwent a series of changes due to the need to adjust the theoretical proposals to Israel's political and social realities.
The first retreat from the alternative described above was one that everyone accepted; it stemmed from the undertaking to refrain from worsening income distribution.
Furthermore, and irrespective of concern for the progressivity of the tax system, the reform planners began to doubt the possibility of modifying the set of tax benefits in any significant way. Whenever a benefit was slated for abolition, skeptics responded with the expression "It won't fly," i.e., the Knesset will not accept the change. Whenever a tax benefit is created, so is a coalition that will fight to perpetuate it-directly, on the part of taxpayers who benefit from it, and indirectly, on the part of those by whose mediation the benefit is given. Reducing a tax benefit inflicts losses on powerful coalitions that, naturally, will attempt-successfully, in most cases-to torpedo the change and preserve their achievements.
With this awareness in the background, we were left with two possibilities:

1........... To disregard the difficulty of passing the change in benefits and to ask the Knesset to
............. approve the alternative shown above, with corrections warranted by the wish to avoid
............. harm to income distribution, and to hope that the beneficiaries of the reduced tax rates
............. would persuade the Knesset to go along. Accordingly, if the Knesset refused to
............. downscale the exemptions, the government would withdraw the entire reform proposal.

2........... To present the Knesset with a more modest reform proposal-to retain some of the hard-
..............to-abolish tax benefits and to offer, commensurably, a smaller cutback in tax rates.
............. The budget constraint would be maintained in any case.

Eventually, a third path was chosen: one that lowered tax rates to a greater extent than NIS 9 billion-to make the tax system more progressive-and reduced the change in the map of benefits to inflict the least harm on most of the population. Consequently, the revenue-neutral, reform turned into a reform that, according to the model, carried a NIS 5 billion deficit.
Had the process of legislating the reform continued, the principles set forth in the first stages of work would probably have eroded further as the government and the Knesset deliberated the plan. The final result would have been a deficit that would have made the reform unviable.
As stated, the changes in the tax laws were examined by means of a model that overestimated the cost of the reform because of two drawbacks. On the one hand, the model was static and did not take account of the rebound effect of the reform on growth and on compliance with the tax laws. On the other hand, because of limitations in the data, the model was incomplete in that it left income derived from abroad and betterment of dwellings out of the tax base.
Would the extra tax collection resulting from accelerated growth, less tax evasion, and taxation of income from abroad and from betterment of dwellings have offset the NIS 5 billion deficit? Taxation of income derived from abroad and income from betterment of dwellings probably would not have augmented collections by any significant sum. First, Israel's post-reform tax rates would resemble those abroad, and second, we intended to defer the betterment tax in cases where one dwelling is swapped for another.
As for the effects of the reform on growth and compliance with tax laws, there is no doubt that collections would have benefited. To be cautious, however, it seemed proper to estimate this increment-over several years-at less than NIS 5 billion and to cope with the possibility that the tax reform would eventually cause the budget deficit to increase.

7. Characteristics of the Reform Proposal in Its Final Form

As noted above, the reform proposals underwent various changes as they were being prepared, and the final result, crafted shortly before the reform was mothballed, deviated from the original plan in several main respects:

a. Maximum Average Tax Rate

According to the original plan, the highest marginal tax rate would not exceed 35 percent. In the end, it was decided to set the maximum average rate of direct taxes at that level. This constraint eases the burden on the tax authorities but assures taxpayers that, come what may, they will pay no more than 35 percent of their gross income in income tax, National Insurance contributions, and health tax.
This means that all Israelis, employees and self-employed, those with work income and those with investment income in Israel, will always know that nearly two-thirds of their income will remain theirs after all direct taxes, as enumerated above, are paid.
To calculate the average tax rate, it was proposed to take account of all income derived in Israel and elsewhere, including special income to which preferential tax rates apply and including exempt income, with the special and exempt income taken into account first.
The tax payments factored into the computation include all remittances on account of income tax, National Insurance, and health tax, in Israel. National Insurance contributions by employers and by self-employed taxpayers, which correspond to employers' contributions, were excluded from the computation of the average tax rate.
Setting the average tax ceiling had the result of gradually lowering today's credits and exemptions at high income levels, and a total phaseout at especially high income levels.
This method, on the one hand, assures taxpayers that their total tax bill will never exceed 35 percent of their income and, on the other hand, eliminates tax benefits for high income earners, as reflected in the tax threshold, credits, social deductions, exemptions, and special tax rates.

b. Reduction of National Insurance Contributions

While reducing the total tax rate for high-income earners, the plan also lowered National Insurance contributions for low-income taxpayers from 5.76 percent to 3 percent, in order to improve the net income of disadvantaged groups.

c. Protection of Disadvantaged Groups

From the outset, the aim was to make sure that the disposable income of the lowest deciles would not be harmed, although this protection was not extended to the level of individual households. Later on, the constraint was toughened: the proposals were tested to make sure that the reform would not harm a large minority of households in the lowest deciles.

d. Protection of Specific Population Groups Not Enumerated among the Disadvantaged

As the reform plan was being elaborated, political constraints proved to have a substantial effect on the formulation of its details. For example, the planners often expressed doubts about the viability of tampering with tax arrangements applying to certain groups, especially when such measures might ignite a public struggle. In this state of affairs, we had to ask whether the added revenue obtained would be worth the damage caused by public confrontation with a politically powerful group. For these reasons, the benefits for donors, persons with disabilities, shift workers, and residents of development areas were retained.

8. The Reform Proposal in Its Final Version

The original reform proposal abolished all existing tax benefits and created a new and minimal list of tax-relief measures. After the fact, it was found that the elimination of several benefits would be harmful to poor population groups. (For example, a sweeping elimination of the tax exemption for child allowances would hurt low-decile households) and to those with strong lobbies.
In the proposed solution, the tax benefits were sorted into three groups:
......... *...... Benefits to be retained, such as those for the disabled, charitable donations,
.................. and shift work;
......... *...... Benefits to be phased out for persons earning more than NIS 15,000 per month;
......... *...... Benefits to be eliminated but replaced with limited tax rates.
Inaddition to these, the special model of benefits for provident funds was retained, as noted below.

a. Benefits Retained

The final version of the proposal left a large share of the existing benefits unchanged (such as the credit for charitable donations and the benefit for shift labor).¹³

b. Social Benefits Reduced

It was proposed to replace the current exemption on some social allowances with a deduction set at exactly the level of the allowance, with a phaseout provision for persons earning NIS 15,000 per month or more. This would make it possible to downscale the existing tax benefits without harming the lower and middle classes.
A similar reduction was proposed for tax benefits that are given in the form of credits. The full credit would be retained for people in low and middle income categories and gradually reduced (by 10 percent) for those earning NIS 15,000 per month or more.
Since most of the population earns less than NIS 15,000 per month, the level of these tax benefits (exemptions or credits) would not change. The added revenue generated by downsizing these benefits for high-income earners was estimated at only NIS 200 million per year. However, the phaseout technique complicates the tax laws and raises the effective tax rate to 58 percent. Both of these results clash with the spirit of the reform. Despite these drawbacks and the scanty added tax revenue, it was decided to adopt this method because of the implicit message it would send about increasing the burden on high-income earners for reasons of social justice.

c. Exemptions Nullified and Replaced with Limited Tax Rates

Some income that had been exempt thus far was to become taxable at a limited rate, as in the original proposal.

d. Provident Funds

The original proposal attempted to put some order into the maze of regulations and benefits that apply to savers with provident funds and to shift the greater part of benefits to retirement age. Ultimately, it was decided to adopt a seriously scaled-back alternative that improves the current state of affairs to some extent but creates no fundamental change.

¹³ Notably, since July 1990, the shift-labor credit has not been offered for income that coupled with the regular income - exceeds NIS 7,960 per month (as of 1999). Furthemore, holding the maximum average tax rate to 35 percent results in the effective elimination of the tax benefit for donors and didabled persons who ear NIS 30,000 per month or more.

Table XVIII-4 Cost of Changes in Direct Personal Taxation and Sources for Their Funding
(NIS billions, 1998 prices)

Costs				Sources of funding
		Long-term				Long-term
	First Year	Incl. surtax	Excl. surtax		First year	Incl. surtax	Excl. surtax
Reduction of income-tax rates	-10.3	-11.4	-11.4	15 percent surtax on income tax	4.0	4.5	0.0
Reduction of National Insurance contribution rates	-1.3	-1.5	-1.5	15 percent surtax on National Insurance contributions and health tax	1.8	2.0	0.0
Limiting the average tax rate to 35 percent	-0.8	-0.6	-0.6	Taxing investment income(including transition surtax)	1.7	5.5	4.8
				Taxing allowances and limiting credits	0.2	0.2	0.2
Total gross cost	-12.4	13.5	-13.5	Total sources of funding	7.7	12.2	5.0
Net cost	-4.7	-1.3	-8.5

Source:Household Expenditure Survey, 1992/93, Central Bureau of Statistics, processed by Economic Research and State Revenue Administration.

The cutback in tax benefits proposed in this alternative is estimated at NIS 5 billion (Table XVIII-4)-NIS 4.8 billion from taxing investment income and NIS 0.2 billion from eliminating the exemption on social allowances and credits for persons earning more than NIS 15,000 per month.
As the reform was being elaborated, it was assumed that in the first few years there would be only NIS 2 billion in added tax revenue while the reduction of tax rates would immediately depress collections by NIS 12 billion. To make the reform less expensive in its initial years-until the tax base would expand-it was decided to impose a transition surtax of 15 percent of tax liability. According to the plan, this surtax would be phased out within three years, and the pace of the phaseout would be commensurate with collections after the reform went into force-all of which to prevent a significant increase in the state-budget deficit.

Table XVIII-5 Tax Rates After the Reform-Final Alternative

a. Without 15 percent transition surtax

Brackets (NIS per month)	Marginal tax rate (percent)
	Income tax	National Insurance⁽¹⁾	Total
0-2,865	15		18
2,866-3,800	5	10	25
3,801-9,500	25	10	35
9,501-22,700	30	10	40
22,701-30,000	40	0	40
30,001+	45	0	45

b. Including 15 percent transition surtax

Brackets (NIS per month)	Marginal tax rate (percent)
	Income tax	National Insurance⁽¹⁾	Total
0-2,865	17.25	3.45	20.70
2,866-3,800	17.25	11.50	28.75
3,801-9,500	28.75	11.50	40.25
9,501-22,700	34.50	11.50	46.00
22,701-30,000	46.00		46
30,001+	51.75		51.75

Source: Economic Research and State Revenue Administration
Note to the Table:
(1) National Insurance contributions include health tax but without the employer's share.

Table XVIII-5 shows that the tax rates are much more progressive now than in the original proposal; there are six tax brackets here, as against seven in the existing situation and only two in the original proposal. If we take account of the transition surtax, the highest tax rate looks even higher than it is today-51.75 percent as against 50 percent today and 35 percent in the original proposal.
However, the average tax-rate limitation makes the rates less progressive. For a taxpayer with 2.75 credit points and income that is taxed only at regular rates, the average tax rate rises gradually and bumps against the 35 percent average-rate ceiling (40.25 percent including the transition surtax) at NIS 30,000 per month. Above this income, as a result of the average-rate limitation, the effective marginal tax rate declines from 40 percent to 35 percent. This taxpayer does not have a 45 percent tax bracket, of course. For other taxpayers, who have additional credits or deductions that were not previously abolished-such as those for donors or the disabled¹⁴ -the average-tax limitation comes into effect at a higher income, thereby, practically speaking, abolishing the remaining tax benefits. Those with income that is taxed at limited rates face a similar phenomenon: they will benefit from the average tax-rate limitation at higher income levels.
Formally, their regular income would be taxed at 45 percent, but effectively they will cease to benefit from limited tax rates and their capital income will be taxed at a rate no higher than 35 percent.
Hence, the anomaly of declining tax rates at high-income levels has not been abolished. However, it derives not from the ceiling of National Insurance contributions and health tax, as it does today, but from the average tax-rate limitation and the gradual elimination of social tax benefits for persons whose monthly income exceeds NIS 15,000.
Although Table XVIII-5 shows steadily rising tax rates, the effective marginal tax rate climbs from 40 percent to 50 percent in the range where social benefits are eliminated, returns to 40 percent afterwards, and declines to 35 percent at the point where the average tax-rate limitation is encountered.

¹⁴ It is worth bearing in mind that most social benefits were abolished at monthly gross income exceedind NIS 15,000.

9. Impact of the Reform

Graph XVIII-1 shows that during the transition period, the marginal tax rate declines steeply, by 10 percent on average, at income levels between NIS 9,500 and NIS 30,000. At lower income levels, the changes are smaller-for taxpayers with 2.75 credit points, the income-tax threshold rises from NIS 2,600 to NIS 2,800 and the rate of National Insurance contributions and health tax declines from 5.76 percent to 3.45 percent up to half of the national average wage.

Graph XVIII-1
Income-Tax, National Insurance, and Health-Tax Brackets-Employee
(including transition surtax after reform)

Graph XVIII-2
Income-Tax, National Insurance, and Health-Tax Brackets-Employee
(Not including transition surtax after the reform)

Table XVIII-6
Effect of the Reform on Net Income by Net Income Deciles per Standard Adult

(Average per household, NIS per month, 1998)

	Before reform					After reform					Change, net
Decile	Income tax	National Insurance (1)	Total direct taxes	Gross income	Net income	Income tax	National Insurance (1)	Total direct taxes	Gross income	Net income	(NIS)	(Pct.)
Avg	1,996	707	2,672	12,512	9,840	1,911	768	2,678	15,512	9,834	-6	-0.1
1	41	79	120	3,426	3,307	225	170	395	3,426	3,032	-275	-8.3
2	235	209	444	4,932	4,448	479	279	758	4,932	4,175	-313	-7.0
3	383	301	684	6,418	5,734	646	359	1,005	6,418	5,413	-321	-5.6
4	567	384	951	7,304	6,353	785	422	1,207	7,304	6,097	-256	-4.0
5	812	487	1,300	8,674	7,374	996	514	1,510	8,674	7,163	-211	-2.9
6	1,162	649	1,811	10,577	8,767	1,311	643	1,954	10,577	8,623	-144	-1.6
7	1,609	800	2,410	12,598	10,188	1,698	791	2,489	12,598	10,109	-79	-0.8
8	2,722	1,096	3,818	16,025	12,207	2,491	1,904	3,585	16,025	12,440	223	1.9
9	4,419	1,408	5,827	20,832	15,005	3,734	1,495	5,230	20,832	15,602	597	4.0
10	7,697	1,650	9,348	34,291	24,943	6,732	1,906	8,639	34,291	25,653	709	2.8

	Before reform	After reform
Cost of reform (NIS billions per year)		1.3
Poverty line (50 percent of median net income per standard adult in NIS per month)	1,445	1,462
Households under poverty line (percent)	13.2	13.3 ⁽²⁾
Households under poverty line (percent)	14.6	14.4 ⁽²⁾
Ratio of uppermost quintile to lowest quintile	5.13	5.03
Gini Index	0.327	0.322

Source:Household Expenditure Survey, 1992/93, Central Bureau of Statistics, processed by Economic Research and State Revenue Administration.

Note to the Table:

(1)Including health tax.

(2)After the reform, 15.8 percent of households and 18.2 percent of individuals will be under the pre-reform poverty line (NIS 1,445 per month).

Table XVIII-6 summarizes the effect of the reform on the distribution of the tax burden. The table assumes that the contraction of tax benefits is fully reflected and that the tax rates include the transition surtax.
As it turns out, the reform is progressive by any yardstick. The tax burden in the first nine deciles decreases by 7 percent-from NIS 1,929 per month on average to NIS 1,794. Consequently, disposable income in these deciles rises by 1-2 percent. The uppermost decile is harmed, even though its tax rates decline. Its share in tax payment climbs from 35 percent before the reform to 38 percent after it, and its disposable income decreases by 2 percent. According to our estimate, the uppermost decile receives 65 percent of capital income and, therefore, is the first to be hurt by the reduction in tax benefits.
In the long term, the reform costs NIS 1.3 billion- NIS 2.5 billion because of the reduced tax burden in the first nine deciles, offset by NIS 1.2 billion in added collections from the uppermost decile.
The percent of poor households does not change substantively. However, the reform raises the poverty line, i.e., it boosts the net median income. When we measure the incidence of poverty after the reform by means of the pre-reform poverty line, we find that the proportion of poor households declines by 3 percent and settles at 12.8 percent of all households.
The ratio of disposable income between the uppermost quintile and the lowest quintile and the Gini index of disposable income show the same thing-the proposed tax reform in its final version somewhat reduces inequality in the distribution of disposable income.
As stated, the reform was also examined by means of two additional tools-a simulation of payslips of 40,000 civil servants and computations based on various representative of taxpayers. An example of the use of these tools follows.
Processing of civil servants' payslips shows that, in view of data on the payslip only (i.e., not including income from social allowances and capital), the reform betters the circumstances of 99 percent of employees. About 500 employees (1 percent of the total) are harmed by the reform. Some 400 of the 500 earned NIS 8,000 per month, loosing NIS 30-NIS 70 per month. The employees were categorized by various characteristics-sex, place of work, place of residence, occupational rank-in order to identify and respond to problematic provisions of the reform.
Table XVIII-7 gives an example of how representative taxpayers analysis can be used. The representative taxpayer shown in Table XVIII-1 above is made up of male employees with working wives and four children, who receive an investment income in addition to their wage. The investment income was set at the equivalent of 20 percent of wage exceeding NIS 10,000 per month. Table XVIII-7 shows that the reform increases these taxpayers' disposable income up to a wage level of NIS 15,000 per month. Above this level, the inclusion of child allowances and investment income in the tax base reduces disposable income. The table also shows how the reform modifies the average tax rate. In the pre-reform situation, the maximum average tax rate verges on 50 percent of taxable income. For example, a taxpayer who earns NIS 100,000 per month pays NIS 49,000 in direct taxes. However, since the same taxpayer has NIS 27,000 in additional exempt income, his effective average tax rate is only 39 percent. After the reform, the same taxpayer pays a maximum of 40 percent (including the transition surtax) of his total income. The table shows clearly how the elimination of exemptions improves both horizontal equality and the progressivity of the tax system, despite the considerable decline in tax rates.