Israel Finance Ministry, Economic Outlook

ECONOMIC OUTLOOK

September 13, 1999

Non-Israeli Workers in the Business Sector

	1995	1996		1997	1998
Thousands
Total labor permits for non-Israelis	101		119	123	115
Thereof: Territories	31		24	31	35
Foreign workers	70		95	92	80
Total illegal workers	51		60	68	78
Thereof: Territories	29		17	17	18
Foreign workers	22		43	51	60
Percent of non-Israelis in total business-sector employment	9.6		10.8	11.4	11.4

Sources: Employment Service, Bank of Israel

Many structural changes proposed in past years have not advanced to the implementation stage. Take foreign workers, for example. Actually, this issue is not representative of ordinary structural changes because there seems to be a wall-to-wall consensus about the importance of tackling it. For the very reason of this consensus, one might think that the change would be pulled off successfully. One would be wrong. The number of permits for foreign workers has been cut back, as the government resolved, but the numbers of illegal foreign workers have grown steadily. Much ink has been spilled in retelling how problematic the protracted growth of the foreign-worker population is, especially at a time of high unemployment and especially when unskilled workers are conspicuously represented among the jobless.

Composition of the Budget

After the government set the 2000 budget-deficit target at 2.5 percent of GDP, and after the effects of demographic and legislative changes that made it necessary to increase the 2000 budget by NIS3.9 billion relative to the 1999 budget were factored in, the budget deficit had to be reduced by NIS5.8 billion to meet the new target. Obviously, the need for so large a budget cut makes it very difficult to change priorities abruptly-a fortiori at a time of economic downturn, when tax revenues are low and welfare and social-service expenditures are high and rising. In choosing the budget lines to be slashed, the guiding principle is to cause minimum harm to public services, to streamline the public sector, and to realign the government's priorities. Accordingly, several resolutions were taken, foremost an increase in the budgets of the ministries of Education and Health and a NIS500 million increase in the transport infrastructure budget (on top of the existing NIS4 billion total budgeted investment in infrastructure), despite the need to reduce the total budget.

To recompose the budget as the government demanded (in reference to the proposal presented to it) without deviating from the deficit target stipulated at the first meeting, it was necessary to expand sources by NIS1.75 billion. This sum is needed to pay for the decisions to reduce the cutback in the Defense Ministry budget (from NIS1.2 billion to NIS950 million), to rescind the NIS400 million cutback that had been planned for the Ministry of Education, to rescind a cutback in child allowances (which would have reduced budget expenditure by NIS600 million), and to invest more (NIS500 million, as stated) in infrastructure.

To finance these decisions, four sources were chosen:

1. An extra NIS200 million cutback across the board.

On can always prune the budget by reducing all ministries' budgets at the same rate; this method attracts the smallest measure of public and political resistance. However, it has the disadvantage of failing to reflect changes in the economy's needs or the government's priorities.

2. Raising the 1999 tax revenue estimate by NIS750 million.

The forecast of revenue from taxes and other sources (National Insurance, Israel Lands Administration, royalties, etc.) in 2000 dictates the level of the budget in view of the deficit set forth in the Reduction of the Deficit Law (2.5 percent of GDP in 2000). Considering the many stages that the 2000 budget must pass until it is sent to the Knesset for approval, it was necessary to create a revenue forecast by July, on the basis of estimated revenues in the first six months of 1999. Since the revenue level rose in June-August-after the 1999 revenue forecast was written-it seems that tax revenues in 1999 will exceed the July estimate by NIS750 million. In response to information obtained in the past few months in this respect, it was decided to raise the estimated 1999 revenue base by this sum. The predicted growth rate of revenues between 1999 and 2000 was not changed.

3. Raising the ceiling for payments of National Insurance contributions and health tax (employee's share only) from four times the national average wage to five times, thus augmenting revenues by NIS450 million.

Israel, like quite a few countries, has a wage ceiling on payments of health tax and National Insurance contributions. The existence of such a ceiling distinguishes between ordinary taxes and payments that have an insurance element. Thus far, revenue from wages in excess of four times the national average (NIS23,924 per month, gross) have been exempt from withholding of 10 percent for National Insurance contributions and health tax. After the change, income between NIS24,000 and NIS29,900 will be taxed as well. The correction will affect 4 percent of wage earners (those who earn more than four times the national average wage), who will pay NIS450 million more. It will also have the effect of broadening the highest marginal tax bracket of 60 percent (50 percent income tax, 5 percent health tax, and 5 percent in National Insurance contributions) to five times the national average wage. Beyond that level, the lower marginal tax rate of 50 percent will continue apply. As stated, many other countries have a "hump" of this kind, reflected in lower marginal tax rates for higher income earners. Although the marginal tax rate on very high incomes declines, the system as a whole remains progressive, since the average tax rates rise commensurate with income.

As a rule, it is bad for the labor market to raise the marginal tax rate. Such a measure increases the marginal utility of tax evasion, prompts taxpayers to do more tax planning, and encourages businesses to transfer economic activity to countries where tax rates are lower. From the macroeconomic standpoint, these effects are especially severe at a time of downturn in growth and high unemployment, when it is so easy (especially in high-tech fields) to move activity to other countries. (Many high-tech companies already have overseas subsidiaries or offices.) Furthermore, raising the tax on labor is contrary to the requirements of the direct-taxation reform plan. This reform, to be presented to the government for approval in the course of 2000, is predicated on broadening the tax base to include non-labor income and lowering marginal tax rates on labor. This also explains the decision to raise the wage ceiling for payments of National Insurance contributions and health tax by means of a one-year ad-hoc provision (i.e., the decision will lapse automatically within one year). This decision may provide evidence of some degree of commitment to the direct-taxation reform.

4. Raising taxes on Diesel fuel for transport, for a net revenue increase of NIS350 million

The 1998 report of the State Revenues Administration presents a lengthy description of a planned reform in fuel taxation: an increase in the excise tax on Diesel fuel for transport and a decrease in taxation of Diesel fuel for industry. This will eliminate a distortion in the production of energy (in which administrative intervention or imports or exports of distillates are sometimes needed to regulate the market) and internalize the negative external costs of transportation. To carry out this measure, Diesel fuel for industry must be separated from Diesel fuel for transport in some technological way, so that the industrial Diesel fuel cannot be used for transport. The purpose is to reduce the "seepage" between uses that would occur in the case of differential taxation. Most countries use differential taxation of this kind and have solved the problem of distinguishing between types of Diesel fuel in various ways, such as extreme differences in quality and color and chemical marking (which makes it possible to identify non-transport Diesel fuel even when it is highly diluted). In July 1999, Israel began to distinguish between the two types of Diesel fuel by means of an order from the Fuel Administration setting the maximum sulfur content of transport Diesel fuel at 0.05 percent and that of industrial Diesel fuel at 0.2 percent. This is part of a larger process in which international standards that reduce emissions of fuel pollutants will be adopted. Notably, the differences in quality are not yet meaningful, since the economic incentive to engage in "seepage" will come into play only when the aforementioned proposal is applied.

The process is programmed to generate net revenue of NIS350 million. The gross revenue has to be much greater because the change will entail additional budget outlays: subsidization of public and other transport from the state budget, the railroad budget, lowering of fees for Diesel-powered motor vehicles (a measure meant to express the new price differentials between Diesel fuel and gasoline), reducing the excise tax on Diesel fuel for industry and heating; reducing the excise tax on kerosene and aligning it with the taxes on other types of industrial fuel, and administrative actions to enforce the new regime and deter would-be violators.

The effect of this measure on the Consumer Price Index is not significant because Diesel fuel accounts for a small share in household consumption, especially when the reduction in taxation of industrial Diesel fuel is taken into account. It should also be borne in mind that the share of fuel in the per-kilometer cost of transport to business-buses, taxis, and trucks-is 7.5 - 8.2 percent. The share of fuel in the "basket" of expenditures presented to the price control committee is slightly higher, at 11 percent.

The entire measure is definitely favorable and will have significant collateral effects on the manufacturing of distillates, the costs of production in industry, and the behavior of consumers. Concurrently, it will internalize some of the externalities that pertain to the use of Diesel fuel in transportation.

In sum, the measures shown-raising the expenditure side and recomposing and raising the revenue side-will have the following total effect:

When we compare the composition of the 2000 budget with that of 1999, we find that two budget lines stand out. Although some ministries will find their budgets reduced, the government decided to increase transport infrastructure investment significantly and to expand the budget of the second-largest ministry, that of Education. The benefits of larger infrastructure investments are undoubted. An increase in education investment is a growth-supportive measure, but the steps previously proposed, which would have cut the education budget, had been meant to make this sector more efficient.

The budget increases will be funded by measures that are undeniably growth-suppressive: tax hikes. This is especially so when some of the hikes will occur in direct taxes, which affect the labor market directly.

To determine the net outcome of the two clashing effects, the question to answer is whether the positive return from the extra allocation for infrastructure and education surpasses or falls short of the negative return that occurs when money is transferred from the business sector to the public sector. The inquiry should concern itself with both the extent of the effect-positive or negative-and its magnitude. In respect to taxation, the burden will increase by 0.2 percent of GDP. In education, a programmed reduction of 2 percent has been waived. Transport infrastructure will get a 40 percent supplement (and total infrastructure investment will gain by 13 percent). From the macroeconomic point of view, it certainly would have been better to finance the extra allocations for education and infrastructure from budget items that contribute less to growth and other goals than the favorable contribution of education and infrastructure and contribute less than the negative contribution of raising taxes. Efficiencies and cutbacks in these items would have obviated the need to increase the budget and raise the tax burden.

Monetary Policy

A very tough monetary policy has been applied in 1999, as reflected in the real interest level (nominal interest less inflation expectations for the coming twelve months). In the first eight months of the year, the real interest rate oscillated around an average of 7 percent. Households and businesses, of course, paid much higher rates: 11 - 14 percent (depending on the customer). These high interest levels have helped the appreciable disinflation that has occurred but have also contributed to the economic slowdown. One may dispute the degree of correspondence between inflation and growth-the cost of each percentage point of disinflation in terms of the percent of growth lost-and in Israel's case, in 1995 - 1999 one may also debate the short-term cost of the disinflation process as against the long-term benefits of low inflation. There is no doubt, however, that the tough monetary policy, now in its fifth year, has dampened economic growth.

The goal of monetary policy in 1999 is to meet the government's 1999 inflation target of 4 percent. Several weeks ago, the Bank of Israel, pursuant to a government resolution, released its estimation that inflation in 1999 will be more than 1 percent under the stipulated target, i.e., less than 3 percent. The steep currency depreciation in the autumn of 1998 has taught us to beware of congratulating (or cursing) ourselves beforyear's end, but for the purpose of discussing the 2000 - 2001 inflation target, let us assume that 1999 will end with no surprises and that Israel will close out the year with less than 3 percent inflation.

Interest and Inflation Rates-Annual Averages

	Real interest rate	Key lending rate	Inflation	Inflation expectations

1990 - 1994 average	0.2	12.9	14.3	13.6

1995	4.6	14.6	10	10.6
1996	4.3	15.1	11.3	11.6
1997	5.0	13.6	9.0	9.1
1998	5.8	11.6	5.4	6.2
*1-8.99	7.1	12.5	6.5	5.9

1995 - 1999 average	5.4	13.5	8.4	8.7

* The average inflation rate for 1999 was computed as the average in the first eight months of the year relative to the corresponding months in 1998.

Sources: CBS, Bank of Israel

The low inflation that will probably be attained this year will be the result of the negative indices early in the year. (In the first seven months of 1999, the CPI decreased by 0.1 percent.) One should not regard the negative indices in early 1999 as a phenomenon that will recur in coming years, just as one should not regard the very high indices of 1998 as permanent. (As we recall, in late 1998, after the steep currency depreciation and despite the high indices, the government set the 1999 inflation target at only 4 percent.) In today's inflation environment-4 - 5 percent-a target of 3 - 4 percent signals the government's intention to keep the disinflation process going. The target was set for two years, 2000 - 2001, because of a simple truth: it is easier to reduce inflation from 10 percent in 9 percent, for example (and, therefore, less costly in terms of lost product) than to reduce it from 4 percent to 3 percent (because of inelasticities in nominal wages). Furthermore, disinflation relative to today's present inflation environment is of less long-term utility to the economy than disinflation that begins at a higher level. In view of these two remarks (which reflect the thinking of a school in which Paul Krugman and Lawrence Klein are two of the most prominent exponents), the current inflation environment, and the state of the economy, the government decided that, for the next two years, the Bank of Israel shall apply the tools at its disposal to continue lowering the inflation rate and to create a lower inflation environment of 3 - 4 percent. Another point emphasized, in the context of monetary policy, was the importance of a target that represents progress toward fulfilling the economy's growth potential-especially considering the protracted downturn in activity and the rising level of unemployment.

Why is our growth forecast lower than the predictions of other economists?

As stated, the Finance Ministry growth forecast is the most conservative in the market. The predictions of other macroeconomists, in Israel and abroad, fall into a range of 3.2 percent to 4 percent. Such a disparity between our prediction and the average forecast ought to be addressed. Several forecasters presumed that monetary policy will take an expansionary turn; we do not share that assumption. However, a brief telephone survey showed us that the main difference is rooted in one basic assumption, that pertaining to peacemaking developments with all of their ramifications. Below, we relate to this assumption before all others.

Underlying assumptions and risk factors in the 2000 forecast

No forecast is complete unless it explicitly mentions its underlying assumptions and lists factors that may result in a different (higher or lower) actual growth rate. These factors are much more numerous and meaningful this year than in the past. Three of these factors (the Y2K bug , bimillennium tourism, and the new government's economic policy) are peculiar to the current year.

The peacemaking process-The peacemaking process and the security situation have an immense effect on future growth. Rapid and stable progress in the process will contribute favorably to expectations and actual growth in the coming year. The forecast assumes that in 1999-2000 nothing extreme will happen in this respect, for better or worse. One may certainly debate the reasonability of this assumption, but the debate over the reasonability of any other assumption would be more acrid. As stated, different assumptions among forecasters concerning developments in this regard explain many of the differences among their forecasts. Most forecasters assumed that favorable progress in peacemaking will occur and that this will affect exports, investments, private consumption, and-especially-expectations of future prosperity. The forecasters who based their prognoses on such a development did so in view of a particular principle: the more progress in peacemaking one expects, the stronger the contribution to growth will be. Such an assumption is permissible in forecasts that serve as a basis for short-term (quarterly or annual) economic decisions. Such forecasters will weight all information and assessments available to them at any given moment, formulate their predictions on their basis, and adjust them whenever the wind changes direction. If within a few months they believe that peacemaking is advancing more rapidly or more slowly than predicted, they can always adjust their predictions with no harm done and no cost incurred. Those whose forecasts dictate longer-term economic decisions do not have such flexibility and, therefore, must limit themselves to much less assertive assumptions.

The Y2K bug-The forecast assumed that the Y2K bug will have no significant repercussions in the financial domain (crisis in world stock exchanges, abrupt changes in exchange and interest rates) or in real activity (economic slowdown prompted by uncertainty), both in Israel and in its trading partners. This is a very strong assumption; experts on the matter consider its fulfillment quite unlikely. The repercussions are hard to estimate for the time being, but undoubtedly there will be implications-probably significant ones. Furthermore, with three and a half months to go until zero hour, it seems certain that even if the turn of the millennium passes uneventfully and the reason for the world panic proves to have been a conspiracy by unemployed programmers, the effects of the expectations and preparations will be felt sharply in the world's capital and currency markets.

Bimillennium tourism-The forecast assumes that the bimillennium events will give incoming tourism a powerful shot in the arm. Insofar as this does not occur, for whenever reason, growth will definitely be affected for the worse.

Economic policy-The public is waiting to see where the new government's economic policy is headed. Lest we forget, the 2000 budget still has a long way to go until finally approved by the Knesset. Economic policy, resolve, and executive ability will have implications mainly for future years, but the power of the public's expectations should not be belittled; they may affect economic activity by 2000. The forecast assumes that the budget deficit will be 2.5 percent of GDP and that budget adjustments that are needed to attain this target will not be made on the revenue side, i.e., taxes will not be raised significantly. A slight departure from this assumption has already occurred; we discussed it above. The forecast also assumes that the government and the Knesset will approve the structural changes (at least the meaningful ones) and, more importantly, will implement them.

World crises-The past two years have taught us that financial or real crises may occur in any country and reverberate in Israel. The way policy responds to such events, insofar as they occur, also has substantial implications for future growth. Needless to say, the possibility of a financial or economic crisis in other countries, and the response of domestic economic policy to such a crisis, cannot be factored into the forecast even though it is a real danger.

Immigration-The forecast assumes that the level of immigration will not be much dfrom that in the past few years-about 60,000 per year. An increase in immigration will contribute to a higher growth rate on the demand side, by stimulating private consumption and homebuilding investment, but will raise the unemployment rate beyond the predicted level.

If these assumptions fail to materialize, the forecast and the requisite policy measures may be seriously affected.

Reduction of the deficit law

In the first government meeting on the 2000 budget-that devoted to the budget aggregates-it was decided to adjust the parameters of the Reduction of the Deficit Law. Since this is the third change since the law was enacted in 1991 (for fiscal 1992), we should briefly survey the various phrasings of this law and describe its advantages and drawbacks. An international comparison of similar laws in other countries elicits some interesting insights.

What the Law Says

The Reduction of the Deficit Law, as variously phrased, sets the maximum level of the budget deficit. This stipulation, coupled with the forecast of state revenues from taxes and other sources (royalties, profits of the Bank of Israel, National Insurance), determine how much money can be spent in the coming year's budget.

The Reduction of the Deficit Law, 1992, set the maximum domestic deficit in the state budget, not including net allocation of credit (hereinafter, the domestic deficit), as a percent of Gross Domestic Product in each of the years 1992 - 1995.

In 1993, the law was amended to state that the deficit in 1994 - 1996 shall decline each year but did not state by what rate.

In 1996, the law was amended to state that, starting in 1997, it shall apply to the total state budget deficit (excluding net allocation of credit) and stipulated the annual fraction of the deficit as a fraction of GDP in the years 1997 - 2001.

The law was amended again in 1999. In its current version, it sets the permissible deficit rate in 2000 and states that the deficit shall decline each year until 2003 in accordance with the pace of economic activity, provided that the annual decrease be no less than 0.25 percent, so that in 2003 the deficit shall not exceed 1.5 percent of GDP.

Advantages of the Law

In 1991, when the government's main goal was to take in hundreds of thousands of immigrants within a very short period of time, it was decided to send the business sector a calming and clear message: the immigrant-absorption enterprise would be carried out with a maximum of fiscal responsibility; the mammoth budget deficits that burdened the economy in the 1980s would not recur. It was immensely important to convince the business sector that the government was truly resolved to honor this commitment. If the business sector were not convinced of this, it would build up expectations-of an increase in deficits, an upturn in inflation, a future increase in the tax burden, and higher interest rates on the way-that would affect businesses' decisions. A credible statement of intent was also important for foreign capital markets, albeit immeasurably less so at the beginning of decade than today. To assure the credibility of this statement, the government accompanied it with an undertaking in the form of a law that tied its hands in planning the annual budget and, in the original phrasing of the law, requiring the government to reduce its domestic budget deficit each year along a given trajectory that would end at zero in 1995.

Although the frequent rephrasings of the law have eroded this credibility somewhat, the restraint created by the law has undoubtedly been decisive in strengthening budget discipline in Israel and gradually reducing the share of the deficit and the public-sector debt in GDP. The Reduction of the Deficit Law converges with several policy decisions made in the past few years (including those that set an inflation target) that have indisputably given the Israeli economy the contours that we recognize today.

Drawbacks of the Law

The law has pro-cyclical effects.

The main advantage of the law-the fact that it limits government spending-is also its main drawback: inflexibility resulting from the need to meet a rigid constraint that was imposed at a time when economic conditions may have been different. A law that imposes rigid rules is pro-cyclical: to comply with it during an economic slump, one has to boost revenues and/or slash expenditure-a policy response that aggravates the downturn. In times of economic upturn, the policy response would have the effect of intensifying the upturn. In the past decade, this drawback has become real twice. At the beginning of the decade, when growth rates and, in turn, tax revenue were very high, budgets were increased. This helped stimulate demand, which was high to begin with, without breaching the legally mandated deficit target. In the second half of the decade, when economic activity slumped, governments of Israel had to work prodigiously to make substantial reductions in the budget (which had ballooned during the years of prosperity); taxes were raised (1997) and deficits mounted (1996 and 1999). Thus, the desired effect was not attained in any of these cases nor do the negative effects balance out over time.

The phrasing of the new law, passed this year, mitigates this negative impact slightly by drawing a connection between the slope of the deficit's downward path and the pace of economic growth. The faster the growth is, the faster the deficit should contract, beyond the necessary minimum reduction of 0.25 percent per year. Within three years, the deficit should be cut by 1 percent of GDP.

Various countries have tried to solve the pro-cyclicality problem in various ways (see international comparison below) and very few countries, if any, totally disregard the cyclicality issue.

The law applies to the state budget but not to that of the entire public sector.

It is feared that ad hoc measures will be taken to comply with the law.

Composition of the Budget Cut

The previously described drawback concerned ad hoc decisions made with respect to revenue side in order to meet the legally mandated target. This drawback recurs in ad hoc decisions concerning the expenditure side. All countries that have similar laws encounter the problem of a less-than-optimal composition of the budget cut when a such cut is needed to comply with the law. It is much easier to postpone long-term projects and investment-intensive expenditthan to slash the current expenditures of government ministries. This is obviously a very significant disadvantage. Notably, however, Israel's infrastructure budgets have actually been raised substantially since the law was passed.

Ostensibly, one may apply the law to the regular budget deficit only and to exclude the investment and development items, but Israel, like other countries, waived this option lest it result in a spate of redefinitions. Budget lines for wages, for example, would be redefined as "investment in human capital"; budgets for office equipment would be termed "investment in personal well-being" or "investment in improving service to the public." This would create a license to increase spending without overrunning the legally stipulated targets.

Metamorphoses of Reduction of Deficit Law and Actual Deficit

		Deficit target, in accordance with the law:
	Actual deficit (domestic through 1996, total starting in 1997)	1991, domestic deficit	1993, domestic deficit*	1996, total deficit	1999, total deficit

1992	4.9	6.2
1993	2.3	3.2
1994	1.9	2.2	3.0
1995	3.2	0.0	2.75
1996	4.5		2.5
1997	2.7			2.8
1998	2.4			2.4
1999				2
2000				1.75	2.5
2001				1.5	2.25 at the most
2002					2.0 at the most
2003					1.5 at the most

Each year, the target for the coming year was set. The law did not stipulate a downward trajectory ab initio.

Source: Ministry of Finance

The table immediately calls our attention to several questions:

In 1995 - 1996, the domestic budget (in percent of GDP) overran the target stated in the law. In 1999, the total deficit is expected to do the same. Is an illegality being committed?

What is the significance of the changeover from a domestic budget deficit target to a total deficit target?

Do the advantages of the frequent amendments to the law surpass their drawbacks?

1. Compliance with the law-at the planning phase, or in performance as well?

According to the law, the domestic state budget deficit in 1995 was supposed to be 2.75 percent of GDP. In actuality, it came to 3.2 percent. A year later, the overrun was even greater. However, no illegality occurred, since the test of compliance with the deficit target is applied at the planning phase and not at the performance phase. The level of the budget should be set so that, according to all available information at the time of planning, compliance with the target will be obtained. If it transpires during the year that the target will not be met, measures to meet the target can be taken. Raising taxes and/or cutting the budget are the severe of these measures. Less severe measures are taken constantly by the Accountant General, who monitors the trends in revenues, expenditure, and the deficit on a daily basis, spots deviations from the deficit trajectory, and (to the extent possible) corrects them by regulating the monthly levels of expenditure. Control is weaker on the revenue side, of course.

A large majority of countries that have similar laws use this method, i.e., they measure compliance with the deficit target at the planning phase and not at the performance phase. The purpose is to mitigate the pro-cyclical effect described above. In years when the economic growth proves unexpectedly slow, the law has an "autostabilizer" effect. Since tax revenues are lower than predicted, the deficit will be larger than the programmed figure. Correcting this situation during the year by cutting the budget or raising taxes may aggravate the downturn in economic activity. Thus, no illegality occurred in 1995 - 1996 or in 1999. This is not to say that the target overruns had no effect whatsoever. They definitely came at an economic price-one cannot deviate so significantly from a deficit that had been announced up front (both in 1996 and in 1995) without repercussions from the world and domestic money and currency markets. The two governments that presided over these overruns also paid a certain price in prestige and credibility. The risk of loss of prestige may be a less effective (and certainly a less pungent) threat than the sanction applied in one of the provinces of Canada, where ministers' wages are reduced in the event of a deviation from the target that cannot be explained in terms of some external event-but it certainly has an impact.

2. Domestic Deficit and Total Deficit

The total deficit pertains to all government activities-domestic and abroad, in foreign currency and local currency-and reflects the contribution of government to surplus economic demand. This is the indicator that most countries use. The domestic deficit, in contrast, pertains only to government activities in local currency-on the revenue and the expenditure sides. This dichotomy is peculiar to Israel.

At a time of full employment, the distinction between the currencies in which activities are denominated is of little consequence; government procurements in Israel would cause the business sector to make purchases overseas, and vice versa. At a time of unemployment, however, demand referred to the domestic market would be important. (This is a very schematic presentation, but for our purposes it suffices.) In the past few years, revenues from abroad exceeded expenditure abroad. The surplus made it possible (in 1997 - 1998) to incur relatively high domestic deficits without infringing on the total deficit target; thus, the moderating influence of fiscal restraint on domestic demand was reduced. (The question of defining the Bank of Israel's profits and deciding how to treat them in the budget are issues for separate discussion.)

3. Advantages and Disadvantages of Rephrasing the Law

It is hard to ignore the high frequency of rephrasings of this law-three times within seven years-especially when in two of these cases it was decided to adopt a less pretentious target than that stipulated in the previous version. (The changeover to a total target in 1997 was a technical one. Furthermore, it was accompanied by the public unveiling of quantitative targets; previously, all that was required was some decrease in the level of deficit.) However, considering that many countries have suspended the implementation of their laws in this matter, repealed them, or simply decided to disregard them when discomfited by the target that they had set for themselves, rephrasing the statute seems to be a preferable option.

After one examines the revenue and growth figures in the first years in which this law was in effect, one may think that even the very ambitious stipulation of the original statute-a balanced budget by 1995-might have been attained had a different policy been applied. Lest we forget, the 1992 - 1995 period was a time of growth and prosperity, in which tax revenue (net of legislative changes) expanded very briskly. Immigrant absorption justified an exceptional increase in expenditure. However, it was not the heightened outlay for the absorption of mass immigration that prodded the legislator to amend the deficit law to stipulate a much less ambitious downward path; the main factor was the aberrant upturn in public-sector wages. Additionally, the rapid economic growth and steep increases in tax revenue contributed to decisions at that time to reduce the direct-tax burden and the cost of labor. The burden was reduced at the pinnacle of a period of growth that was much faster than the rate that the Israel economy could sustain. The result (as we observed it) was a shortage of sources coupled with a slowdown of growth, a steep increase in resources allocated for wages and pensions at the expensive of other budget lines, and difficulty in meeting the deficit targets.

The most recent rephrasing of the law, which raised the permissible deficit to 2.5 percent of GDP in 2000 as against the previous target of 1.75 percent, was virtually forced on Israel by realities. The intensity of the economic slump and the sizable lag in total state revenue (from taxes and other sources) entailed an adjustment of NIS9 billion to meet the original target (taking into account the budget increase required by legislative and demographic changes). It was feared that such a large adjustmwould bring all the drawbacks of the law into play: taxes would be raised, the budget cut would be applied to the wrong items (in terms of their effect on growth), and the magnitude of the cutback would aggravate the downturn-and that these drawbacks would outweigh the utility gained by having a smaller deficit. As we have seen, the option of raising taxes was invoked anyway, albeit at a relatively small 0.2 percent of GDP.

International comparison

The IMF survey makes no reference whatsoever to countries whose laws are in effect for a limited period of time; it deals only with countries that have open-ended laws in this respect. Thus, it excludes statutes on the books of many countries, including Israel.

Most of the rules and regulations invoked around the world have two points in common: they achieve flexibility by taking the business cycle into consideration, and they examine compliance with the target at the phase of planning, not performance.

Different countries take account of their economy's business cycles in different ways. One way is to stipulate a relatively long period of time during which budgets must be kept in balance. Another way is to include escape clauses, e.g., in cases of security risks, very low growth, or a steep decline in tax revenues; in such cases, the requirement to meet the balanced-budget target is set aside. Several laws include restrictions on funding the deficit and sanctions for failing to meet the target.

Conclusion

Our international comparison of legislation and practices with respect to reducing the budget deficit or bringing the budget into balance indicates that the phrasings used by different countries differ in minor respects only. One substantive difference that we cannot overlook, however, is the fact that countries to which we regularly liken ourselves have very small budget deficits and low public debt. Today, these countries are interested in sustaining the budget equilibrium that they have attained, whereas Israel is still struggling to reduce its budget deficit. This forces Israel to behave more ambitiously than countries that have already carried the process to its end.

Strong consideration for business cycles and the existence of escape clauses that permit a deviation from the target in the case of economic slowdown also stand out in the comparison countries. The existence of these clauses explains why changing the 2000 deficit target from 1.75 percent of GDP, as in the previous version of Israel's law, to 2.5 percent, did not rock the foundations of this edifice. Foreign investors and economists are familiar with the way their countries relate to deficit targets at times of economic slowdown; they took favorable note of Israel's decision to keep its budget deficit on a declining path until it reaches 1.5 percent of GDP in 2003.

Rule	Country	Reference to cyclicality and escape clauses	Remarks
Balancing the budget or achieving a surplus in the medium-term + limiting the deficit to 3% of GDP in any given year	Maastricht	To allow automatic stabilizers to function throughout the business cycle, a deficit of up to 3% of GDP is permitted.	Compliance is tested ex-ante (at the planning phase, not performance).
Balancing the budget in the medium term, after the public debt is lowered	New Zealand	In the short run, cyclical factors may cause temporary and desired deficits or surpluses.	Adjustments by raising taxes are forbidden.
A balanced budget at all phases of the business cycle	Switzerland	One of the proposals bruited is this: "If growth surpasses the average (above 1.8%), a budget surplus shall be mandatory. If growth is substantially under the average (less than 0.5 percent), a deficit is allowed."
A legislative amendment requiring a balanced budget each year has been submitted and rejected three times.	United States	In case of national security risk or a majority of 60% in the Congress, the law is set aside.	The question of when compliance is required-planning phase or performance phase-is not noted.
A balanced budget during a period of 4-5 years on average.	Various Canadian provinces	Escape clauses in cases of emergency, disaster, or (in one province) a significant decline in revenue.	Compliance is tested at the planning phase in some provinces and at the performance phase in others.
A balanced budget every year	Several American and German states	In several American states, reserves are kept for use only if it transpires during the year that no deficit has formed.	Applies to planning and not to performance.

.לארשי תנידמ ,1998/1999 ,תורומש תויוכזה לכ ©
webmaster@mof.gov.il :תבותכל םכיתועצהו םכיתורעה תא לבקל חמשנ
:ךיראתב הנורחאל ונכדוע םיפדה