State tax revenues* in 1998, according to the forecast, will total NIS 122.5 billion**, up 3% in real terms from 1997. This will be the result of an anticipated 3% GDP increase and of an estimated NIS 0.2 billion tax supplement deriving from legislative amendments.

The tax divisions were also required to increase collection efforts in order to obtain a NIS 1.09 billion increase in revenues (a billion shekels in 1997 prices). If these efforts are successful, the State tax revenues in 1998 will be NIS 123.6.

Tax collection as a percentage of GDP is expected to grow from 32.0% in 1997 to 32.3% in 1998, constituting a 0.3% increase of the burden. In 1996 the government tax burden was 31.4% of GDP. If the collection forecast is realized, the increase will derive principally from increased collection efforts and not from higher tax rates.

The revenue forecast for 1998 was prepared in September 1997 on the basis of actual revenues in the first 8 months of the year, and was estimated at NIS 109.1 billion (in current prices, see Table 1).

In 1998 the real growth in tax revenues will be NIS 4.85 billion, so that the collection is expected to reach NIS 123.6 billion. This derives from three factors:

Growth: a forecast NIS 3.57 billion increase as a result of a 3% increase in activity.

Improved tax collection: it was decided to increase collection in 1998 by NIS 1.1 billion.

Legislative amendments: These will add NIS 0.2 billion to tax revenues, as specified below.

---

*This section deals with State tax revenues, including those received in foreign currency, collected by the Income Tax and Property Tax Division, the Customs and VAT Division and fees charged by Government Ministries. The revenues in this chapter do not include VAT on defense imports. This section does not deal with the income of other public sector entities that collect taxes and compulsory payments: National Insurance, local authorities and authorities working by special law, such as the Ports Authority.

**All the data are in 1998 budget prices, unless otherwise stated. According to the budget assumptions, the consumer price index in 1998 will stand at an average of 163.8 points.

---

As stated, legislative amendments in 1998 will add little to tax collection (NIS 0.2 billion), subsequent to a 0.7% increase in GDP in 1997. However, legislative amendments in 1993-1996 led to a cumulative 2.2% decrease of GDP in tax revenues - 0.8% of GDP in the Income Tax Division and 1.4% in the Customs Division (see Table 2). In these years many taxes were reduced (personal and corporate income tax, VAT, customs and purchase tax) or even abolished (employers' tax on the business sector, travel tax, general import tax and tax on service imports), in order to promote economic growth and to increase Israel's integration in the world economy.

Legislative amendments added 0.5% of GDP to the Income Tax revenues only in 1993. Here the methods implemented should be explained: increased tax is registered in the year in which the legislative amendment affects the revenues and not at the time of the legislation. Thus there may well be a substantial time lag between the year in which the legislative amendment is enacted and the year in which it is actually manifest in the revenue. For example, the increasing of depreciation rates to 100% in 1991 reduced the revenues in 1991 and 1992, and explains the tax increase in the Income Tax Division in 1993. A legislative amendment received in the middle of a specific year will partially affect revenues in the current year, while the rest of the influence will be felt in the following year.

A lower than planned deficit contributed to considerable tax decreases in 1993 and 1994. In early 1995, when it seemed that tax reductions could be continued, the government decided on a further personal income tax reduction. This decision came into force in September but its main influence on tax collection was manifest in 1996. In the second half of 1995 it was already evident that the deficit would be higher than planned. Since then, alongside steps to reduce government expenditure, the trend of legislative amendment has been reversed, and tax rates have been increased. In late 1995 and early 1996 indirect tax (on cigarettes and benzene) was increased, and when this was insufficient to reduce the deficit, direct tax was also increased the following year, through postponing of updating of credit points and tax brackets.

Anticipated tax increases in 1998, as a result of legislative amendments, are presented below by Divisions (see Table 3).

In 1998 a net NIS 70 million tax increase is anticipated from legislation enacted in the past. No new legislation is planned in the framework of the 1998 budget:

Study fund for the self-employed: Until 1995 only wage-earners received a tax benefit on account of payments into study funds. In 1996, this benefit was also extended to self-employed workers. From 1996, a self-employed worker who pays 2.3% of his income (up to a ceiling) into a study fund is entitled to a tax deduction of up to 1.5%. From 1997 these rates were raised to 4.7% and 3% respectively, and in 1998 they will be raised to 7% and 4.5% respectively. The cost of this benefit is estimated at NIS 330 million over three years, starting from 1997.

Increased depreciation rates for computers: The depreciation rate for personal computers rose from 20% to 25% in 1996, and to 33% in 1997. The depreciation rate on other computers rose to 25% in 1997. The depreciation rate on other computerized equipment remained at 15%.

Increased credit for northern border residents: Northern border residents are entitled to credit on 10% of their income without ceiling. In 1996-1998 the credit was raised to 20% for Kiryat Shemona residents and to 15% for other northern border residents, all up to an income ceiling of NIS 153,240 per annum. in pursuance of the existing legislation. In 1999 the rate will be restored to its original level.

No updating of tax brackets for individuals: Until 1996 tax brackets, the value of credit points and other tax functions were updated three times yearly (on January 1 and with cost of living increments in February and in August) according to the full consumer price index (CPI) increase.

This principle was determined when the inflation rate was 15% - 20% per annum. Since 1992, when the annual inflation rate fell to 10%, a once yearly updating suffices, without causing large differences in real terms in the course of the year. It was therefore decided that from 1997 personal tax function indexes will be updated according to the full CPI increase on January 1, only. A once-yearly updating with a 7%-8% inflation rate is the equivalent of updating three times yearly with a 15% - 20% inflation rate.

As a temporary provision and in order to reduce the budget deficit, it was determined that in January 1997, updating of the tax brackets and the credit points, which had already been updated in August 1996, would be postponed until January 1998, according to the index increase during 1997 (in practice the CPI increase from November 1996 until November 1997).

The fact that tax function indexes were not updated increased tax revenues by NIS 1.2 billion. Most of this amount was collected in 1997, but NIS 310 million are expected to be received in 1998 (mainly from the self-employed workers filing their 1997 tax returns).

Details of the tax brackets and credit points are presented below:

Increased deduction at source on income from a second job and royalties: In February 1997, deduction at source on income from a second job and royalties was increased to 50%, bringing it to the maximum tax rate and increasing tax revenues in 1997. However, part of the increase will be refunded in 1998 to those who do not reach the top tax bracket.

Increased value of use of company car: In April the value of use of a company car by a wage earner was increased from NIS 310 monthly to NIS 750 monthly, increasing tax revenues from wage-earners by NIS 100 million in 1997 and another NIS 60 million in 1998. On the other hand, the tax on the employer was reduced by NIS 100 million, all in 1998, since the employer is entitled to deduction to the amount of the value of the benefit.

Decrease in customs: the customs revenue will fall by NIS 50 million as a result of implementation of a further stage of exposure of the Israeli economy to imports from third countries, under the exposure program which commenced in 1991.

Ban on input tax deduction on jeeps and mini-vans: Formerly VAT regulations allowed input tax deduction on purchase of vans, jeeps and mini-vans, but not on purchase of private cars. In light of the increased private use of jeeps and mini-vans, it was decided that from April 1997 input tax deduction would not be allowed on these vehicles. This led to a tax revenue increase of NIS 40 million in 1997 and a further NIS 30 million in 1998.

Taxi cab licenses: Due to the abolition of quotas on cab licenses and the higher fee paid, this item is expected to yield NIS 110 million more in revenue in 1998 than in 1997.

Increase in frequency use license: Additional revenue of NIS 30 million is expected from the raised frequency fee.

State tax revenues in 1997 grew by 4.6% in real terms, to NIS 109.1 billion, up from NIS 95.3 billion in 1996 (see Table 5). Assuming that GDP rose by 2% in the same period, the share of State tax revenues in GDP rose from 31.4% in 1996 to 32.0% in 1997. This increase is fully reflected in the tax burden, including national insurance contributions and municipal rates, which increased by 0.5% of GDP, from 39.5% in 1996 to 40.0% in 1997.

*1997 data are revenue estimates made in September 1997.

---

It should be noted that the real change in tax revenues is an estimate received from deduction of the change in the CPI from the nominal change, despite the fact that every tax base has a specific price index, and this is generally different from the CPI. For instance, the most relevant price index for VAT revenues is the index without fruit and vegetables, which rose by 11.6% in 1996 in comparison with 1995; in the same period the CPI rose by 11.3%. The use of the CPI in this case gives a lower tax increase on account of price rises and a higher real increase. In 1997, the result is expected to be the direct opposite.

Development of taxes in 1997 was influenced by three factors:

Legislative amendments: Legislative amendments added NIS 2.35 billion (in 1997 prices) to tax revenues in 1997: NIS 1.37 billion in income tax (principally due to non-updating of personal tax brackets, and the raising of deduction at source on certain income to 50%); NIS 640 million in customs taxes (increase of tax on cigarettes and excise on fuel, and taxation on car air-conditioners); and NIS 340 million in fines and fees charged by government ministries.

Increased tax transfers to the Palestinian Authority: Under the economic annex to the agreement signed in 1994 between Israel and the PLO ("the Paris Agreement"), Israel undertook to transfer to the Palestinian Authority taxes collected by the Tax Divisions on imports to the Autonomy and on trade with residents of Israel. 1995 was a transition year, in which the transfers totaled NIS 763 million (in current prices). In 1996, with maturing of the agreement, NIS 1.357 billion (in current prices) were transferred, a real increase of NIS 500 million, which were included in the 1995 tax revenues, but were deducted from the tax revenues in 1996. In 1997 the transfers increased to NIS 1.6 billion, a NIS 100 million increase in real terms from 1996.

Postponement of refunds: In early 1996 exceptional tax refunds, which had been postponed from 1995, amounted to NIS 500 million. In order to counter this factor, this sum must be added to the 1996 tax revenues.

With deduction of these three factors, we find that the "combined" collection in 1997 was 2% up in real terms from 1996, like GDP. It should be remembered that in 1995 and 1996 tax revenue increases were lower than the potential (which is a combination of the rise in GDP and the progression of the tax system).

There are three explanations for the gap between actual and potential revenues:

Delay between date of creation of the tax liability and date of actual collection. As an example, companies and self-employed workers pay advances in the current year on the basis of profit estimates based on past data. The balance of the tax is paid (or alternatively the surplus advances on the tax liability are refunded) at the time of filing tax returns for a year or two years. In 1997, in contrast to the two previous years, this factor led to accelerated growth in revenues - tax refunds fell by 20% in real terms in contrast to 1996, contributing over a billion shekels to the net collection.

Specific tax bases (property taxes and capital gains tax) are outside GDP. Thus, in 1997 the number of real estate deals declined despite the rise in GDP, leading to a decrease in purchase and betterment tax collection. On the other hand, increased capital gains led to an exceptional increase in collection of advances from companies.

Since different tax rates apply to different uses, tax revenues will rise more slowly than GDP if tax-intensive uses increase less than average. For this reason import tax revenues fell by 4%, as a result of fewer vehicle imports.

Estimated State tax revenues for 1997 were, as stated, NIS 109.1 billion, NIS 4.3 billion less than the original forecast. In 1995 and 1996 actual revenues were also NIS 2.3 billion and NIS 3.9 billion less than forecast, respectively.

In light of these significant differences in recent years, the forecasting model was comprehensively reviewed by the State Revenues Administration from 1992 to 1997, with comparison between forecast and actual revenues. The full analysis appears in the annual report of the State Revenues Administration for 1996. Its results are presented below:

In the inspection period, the difference between the original forecast and the actual revenues varied from NIS +4.1 billion to NIS -5.6 billion, or in percentages of actual revenues, from +5.5% to -10.9% respectively. The absolute average difference was 5.1%.

Three quarters of this difference derive from factors external to the model. When these are deducted, the difference is far lower: from +3.3% to -1.9% of the actual revenues, with a periodic average of 1.3%.

The main factors identified as external to the tax model, which would help to produce a forecast closer to actual revenues if they were accurately known, are:

Actual revenues in the year prior to the year of forecast - for purposes of revenue forecast, actual revenues are generally estimated a few months before the end of the year, for submission of the budget to the Knesset.

Legislative amendments - at the time of the forecast, legislative amendments that might be carried out in the budget year are not known and the exact influence on a given year is difficult to estimate.

Modifications of the base - e.g. tax refunds for previous years, transfers to the Palestinian Authority etc.

Economic activity - real macro-economic indexes such as the anticipated rise in GDP and different uses, and forecast of price evolution relevant to the different types of taxes.

In 1997 the difference between the original forecast and actual revenues (estimated in September 1997) was NIS 4.3 billion, or 3.9% of actual revenues, similar to the difference in the last two years. However, this year the entire difference is explained by the acute recession in activity relative to the budgetary forecast - both in prices and in real activity. In fact, this accounts for a NIS 5.0 billion difference(a third of which is explained by lower than anticipated price rises), so that actual revenues are NIS 0.7 billion higher than the amended forecast, in accordance with development of the external factors.

Total Division taxes in 1998 will be NIS 67.1 billion, 3% higher in real terms than in 1997, and the same as the anticipated increase in GDP.

Legislative amendments will add only NIS 70 million, as a result of legislation enacted in the past.

Income tax revenues in 1998 will be NIS 57.5 billion, up 3% in real terms. An income tax increase of 2% is anticipated from wage-earners due to their increased numbers, while wages will not change in real terms relative to 1997. On the other hand greater profitability is anticipated in the business sector, leading to a 4% increase in income tax in real terms.

Revenue from employment tax, VAT from nonprofit organizations, and from financial institutions will be NIS 6.2 billion in 1998, up 2% in real terms, like the number of employed persons.

Total 1998 revenue from property tax, purchase tax and betterment tax will be NIS 3.7 billion, up 3% in real terms. This is a result of the 3% increase in real estate transactions, following a cumulative 30% fall in apartment sales in 1996 and 1997.

Total Division revenues in 1997 will be NIS 52.1 billion, up 3% from 1997 in real terms.

Legislative amendments enacted in the past will reduce the revenues by NIS 20 million only.

VAT revenues in 1998 will be NIS 34.6, up 3% in real terms from 1997, like the anticipated growth in economic activity. VAT revenues will be two-thirds of the Division collection.

Revenue from customs and purchase tax on imports will be NIS 9.2 billion, up only 1% in real terms, as a result of the slow growth of imports of durable goods. Legislative amendments will reduce customs revenues by NIS 50 million, due to execution of a further stage in the exposure plan and international trade agreements.

Fees charged by government ministries will generate NIS 3.3 billion in 1998, 9% higher than in 1997 in real terms.

Legislative amendments will increase these revenues by NIS 140 million, as a result of the increased rate of sale of taxi cab licenses and the rise in the frequency use fee.

The share of the fees in total State tax revenues rose by 1.2% in 1985 and by 2.7% in 1998.

According to the multiyear budget forecast, revenues will grow in 1999 and 2000 on average by 4%, like the anticipated growth in economic activity, and will total NIS 128.2 billion in 1999 and NIS 133.3 billion in the year 2000 (see Table 6).

Until now no legislative amendments are scheduled for 1999 and 2000, apart from amendments enacted in the past and which will affect revenues in these years - a net decrease of NIS 1000 million in 1999 and of NIS 60 million in 2000 (see tables 7 and 8).

*Including taxes received in foreign currency, apart from VAT on defense imports
**1997 in current prices; the other years in 1998 budget prices.
***Estimate based on data of January-September 1997. collection

*Including taxes received in foreign currency, apart from VAT on defense imports.
**1997 in current prices; the other years in 1998 budget prices
***Estimate based on data of January-September 1997. collection
****Transfer of 25% from property tax to fund for compensation and timing differences in revenues of
government ministries through payments.

*Including taxes received in foreign currency, apart from VAT on defense imports
**1997 in current prices; the other years in 1998 budget prices.
***Estimate based on data of January-September 1997. collection