Public Policy Sources 13:
The Case for Cost-effective Government

Big government in Canada - the strategy and the record

The size of government in Canada expanded much in line with other industrial countries after World War II, rising from about 22 percent of GDP in the late 1940s to about 30 percent by the mid-1960s. Thereafter, the country embarked on a European-style commitment to big government. Measured against the total economy, public spending reached a peak level of 52 percent of GDP in 1992. Although most major countries experienced comparable government sector expansion, Canada’s two largest trading partners, the United States and Japan, were exceptions to the rule and today still have comparatively small public sectors at 33 and 36 percent of GDP respectively.

Notwithstanding the immense expansion of the government sector in Canada over the last generation, the size of government and its relationship to economic progress and related social conditions has received remarkably little attention. Generally, analysis and debate regarding Canada’s fiscal structure have focused on individual programs and particular tax and financial strategies, presumably because these have been seen as more consequential than the question of the size of government itself. However, recent research indicates otherwise; namely, that the size of government has a substantial influence on the trend rate of economic growth, living standards and social conditions.

Although the literature on the size of government question is not large, findings are consistent and persuasive to the effect that excessive expansion of the government sector both reduces the trend rate of economic growth, and produces little or no measurable improvement in social indicators. For recent estimates of the impact of large government on economic performance in Canada, and for a review of the literature on size of government, readers are referred to Emes and Samida (forthcoming).

Having much higher government overheads and taxes relative to our principal trading partners adversely affects Canada’s productivity, international competitiveness and the exchange value of the currency. Canadian policy makers often take comfort in the fact that Canadian economic particulars tend to be in line with OECD averages. However, Canada does not compete with an average of OECD countries. More than 90 percent of Canadian trade is with the United States and Japan, countries which have the lowest government overhead and tax levels among the principal industrial countries.

By virtue of supporting government overheads proportionately about 40 percent larger than those supported by our principal trading partners adversely affects productivity and makes the Canadian economy less competitive internationally. Over the last 20 years, there has been a secular decline in the exchange rate that has amount to more than 30 percent. We believe that a lack of international competitiveness springing from excessive government overheads is a principal underlying cause of Canada’s chronic exchange rate weakness.

The size of Canada’s government sector was last in the area of 30 percent of GDP in 1966. Sustained expansion over the next generation increased government to a peak of more than 50 percent of the economy by the early 1990s. Expanding the take of government to more than half of everything being produced in the national economy begs the question; what have been the costs and benefits of this immense expansion of the government sector at the expense of the private sector?

We examine below the record and results of Canada’s experiment with big government over the last 30 years. The salient feature is that the introduction of big government in Canada has come at very high cost. Principal costs include high taxes, overextended social programs, vast government indebtedness and, for the first time in history, levels of unemployment nationally that are nearly twice those prevailing in the rest of North America. Beyond these immense costs, big government has delivered little or no identifiable improvement in economic and social conditions beyond what would have been provided by a smaller more efficient government fiscal structure.

Table 1 displays changes in principal government spending categories between 1966 and 1996. The 16.5 percentage points, or 50 percent, increase in the size of the government sector relative to the economy is almost entirely accounted for by increased transfers to persons and debt interest. The strategy, if one can call it that, looks simply to have been an uncritical commitment to the welfare state, and to the political convenience of deficit finance. If there was a more sophisticated strategy, it is not apparent.

The record shows that, beyond the commitment to the welfare state and deficit finance, other spending priorities changed little. As displayed in table 1, of the total rise of 16.5 percentage points of GDP in government sector share between 1966 and 1996, fully 14.9 percentage points were accounted for by increased transfer payments to individuals and increased debt interest. Goods and services spending did increase by 3.6 percentage points over the period, but this was largely offset by a fall of 2.3 percentage points in investment spending. Other spending changes were inconsequential.

If the immense investment in big government had actually produced stronger economic growth and lower unemployment, the big government initiative might have been self-financing, but it was not. The absence of any economic or social return on the investment in big government presented a serious financial problem; who pays for the new government largesse? Canada’s answer was a heavy reliance on deficit finance which was politically more attractive than actually paying for government largesse with appropriately higher taxes. It was a fateful choice, and the country now carries one of the world’s largest public debt burdens to highlight the error.

Rather than putting the immense costs of big government to the political test of higher taxes, successive Canadian governments chose to temporise and borrow. The strategy was to borrow today, and trust in deficit-financed economic momentum to deliver future growth and prosperity with which to service accumulated borrowings. Lenders and credit rating agencies went along with this for years, but a crisis of confidence finally developed in the early 1990s, first for provincial borrowers and then for the federal government. Faced with the threat of imminent financial crisis, governments finally hauled their fiscal structures into line, but the mountain of debt remains. This whole dreadful episode has provided a very expensive lesson in elementary finance and the dynamics of compound interest.

Government size and unemployment - G7 countries

In general, among small, medium and large size governments, there appears little systematic difference in respect to absolute or relative performance across a wide range of economic and social indicators. There are, however, three very notable exceptions: taxes, economic growth and unemployment. In these three critical dimensions small governments tend to do significantly better than large governments without the debilitating costs and impositions associated with oversized government.

The modern relationship between big government and high unemployment is quite apparent in table 2. For example, among the G7 countries the two smaller government countries (United States and Japan) posted 1996 unemployment rates of 3.4 and 5.4 percent compared to the two largest government countries (Italy and France) which posted unemployment rates of 12.0 and 12.4 percent. In fact, the very close relationship between size of government and level of unemployment among all the G7 countries is quite striking.

In contrast to present circumstances, the G7 countries had much smaller government sectors and much lower unemployment in the mid-1960s. Additionally, the relationship between the size of government and unemployment appears to have been less systematic in the earlier, smaller government era than it is today. In any event, big government and high unemployment are closely associated today and it is on this modern circumstance that we focus our attention. We examine in particular the influence of big government on taxation levels and labour market flexibility. Additionally, we examine the disincentives and dependency inherent in the large-scale recourse to income redistribution typically adopted by big government.

Three key factors appear to lie behind the high rates of unemployment associated with modern big government. First, the punitive tax and regulatory burdens associated with big government adversely affects output, labour market flexibility and employment. Second, big government raises costs and reduces efficiency by forcing economic activity out of the private sector into the government sector. Universally, the production and delivery of goods and services under government monopoly is inefficient and expensive compared to delivery via private sector free markets. Third, generous government transfer programs, which are the common currency of big government countries, carry disincentives and the moral hazard of dependency regardless of how well designed and well intentioned they may be.

Without exception, big government overheads require high taxes, but direct measurement of tax burdens and tax incidence is complicated and to a degree is subject to misrepresentation. Deficit finance, off balance-sheet financing and pay-as-you-go social programs all have the capacity to keep cash taxes somewhat below current spending levels, but not for long. At the bottom line, big government requires high taxes.

Among G7 countries (see table 3), the two largest government countries, Italy and France, have high top marginal tax rates (51 and 67 percent), high consumption tax rates (15 and 19 percent) and high social security tax rates (13 and 19 percent). The smaller government countries generally have substantially lower tax rates. The tax consequences of big government are unavoidable, namely high taxes across the board on income, consumption and payrolls. (It is on this account that the Government of Canada has been unable to honour electoral commitments to get rid of the goods and services tax (GST) and is so reluctant to reduce clearly excessive, job-killing payroll taxes.)

Top marginal tax rates are very important to the economic process because they bear directly on that segment of the population best able to save and invest, and thereby to create jobs. Additionally, top marginal rates are levied on those best able to remove their wealth or themselves from high-tax jurisdictions. It is difficult, however, to measure and compare directly the incidence and effects of top marginal tax rates among countries principally because of complicated, politically-inspired exemptions and thresholds. To overcome these limitations the OECD and the UN studies cited below have prepared various estimates of actual or effective marginal tax rates bearing on major segments of the labour force. Two of these measures, the marginal tax wedge statistic and high-income marginal tax rate, are displayed in table 3.

The UN Human Development Report 1997 estimates high-income marginal tax rates for production workers in the G7 countries. These high-income marginal tax rates apply to a one-earner couple with two children and an income twice the average production worker income. The rates are generally below the top marginal rate, but in most of the larger government countries these rates are still quite high, even for better paid production workers. Overall, the influence of the size of government on the level of high-income marginal tax rates is quite apparent. (The rate for France is an exception reflecting special preferential rates for child rearing couples.)

Marginal tax rates on income do not account for the considerable additional tax burden of consumption and payroll taxes. These latter taxes are particularly onerous for workers in big government countries. To get a more comprehensive measure of the total marginal tax burden bearing on taxpayers the OECD Jobs Study has calculated a marginal tax wedge statistic. This statistic, which is presented in table 3, measures the marginal tax burden of all forms of taxation levied on an average production worker. It is a telling statistic, and very illustrative of the impact of big government on tax rates.

The marginal tax wedge is calculated as the tax gap between the after-tax purchasing power of an average production worker compared to the total wage and payroll tax cost to an employer of employing the worker. The gap is taxes; payroll taxes paid by the employer plus income, payroll, consumption and social security taxes paid by the worker. The OECD estimates indicate that for each dollar expended on employing a new worker in big government countries more than 60 cents goes to taxes and less than 40 cents goes to the worker. The 60 percent marginal tax wedge in the big government countries compares to a wedge of less than 40 percent in the small government countries. The wedge in Canada at 55 percent is not much below the highest tax jurisdictions.

In a major study of industrial country labour markets (The OECD Jobs Study 1994), the principal cause of high and persistent unemployment in OECD countries was identified as “... an inability of OECD countries to adapt rapidly and innovatively to a world of rapid structural change...”. In addition to labour market rigidity, the study also identified high levels of taxation as a serious impediment to full employment. However, the study did not identify or examine the size of government as a critical factor in producing high taxes, labour market rigidity or high unemployment. The study was simply silent on the size of government question, possibly reflecting a lack of interest in the subject on the part of big government member countries. In any event, we believe it important to examine the influence of government size in the determination of taxation levels, labour market flexibility and the level of unemployment.

In general, labour market flexibility declines as the size of government increases (see table 4). One also observes that big government countries tend to have large, unionised public and quasi-public sectors that enjoy considerable insulation from market forces. These large, non-market sectors create a momentum for dispute resolution by regulation, arbitration and litigation rather than the more flexible market-driven solutions common to more open competitive markets. Labour market rigidities are further compounded by dependency-inducing transfer payments which are common to big government. As noted in The OECD Jobs Study, generalised labour market inflexibility reduces the ability of an economy to adjust to external shocks and is reflected in reduced output and higher levels of unemployment.

Overall, we are persuaded that the heavy tax burden of big government stands as a substantial impediment to economic growth and job creation. Additionally, labour market rigidities and dependency-inducing transfer payments common to big government countries compound the adverse impact of high taxes on the rate of unemployment.

Government size and income distribution in Canada

An important goal associated with social programs in Canada has been the redistribution of income from upper to lower income Canadians as a means to reduce poverty and dependence. Our interest is in whether or not expansion of government programs over the last thirty years has been accompanied by improved circumstances for lower income Canadians.

Table 5 compares the pre-tax income distribution of 1965 with that of 1995, and on a post-tax basis between 1971 (earliest data available) and 1995. Readers will note that the 1995 after-tax income share of the lowest income quintile compared to pre-tax share is improved by one percentage point, from 4.7 to 5.7 percent. However, it is important to recognise that even this modest improvement in post-tax income share is not related or attributable to expansion of the government sector.

Tax structures favouring lower income groups are independent of the size of government. Namely, such redistribution via the tax system is equally feasible with small or large government and, therefore, should not be credited to big government. To the extent that more government spending were able to improve the incomes of lower income Canadians, such improvement would show up in pre-tax income. Accordingly, the impact of big government on income distribution is best measured in terms of pre-tax income distributions, which is the measure we employ in the analysis below.

While the pre-tax income share of the lowest quintile increased very modestly from 4.4 percent to 4.7 percent between 1965 and 1995, the income share of the highest quintile also increased, and by much more, from 41.4 percent to 44.1 percent. Broadly measured, the income share of the bottom 40 percent actually fell over the period, from 16.2 to 14.9 percent, while that of the top 40 percent rose from 65.9 to 68.6 percent.

Available evidence indicates that the modest improvement in pre-tax income share of the lowest quintile was more than accounted for by increased transfers, and not by higher earned income and greater independence. Data from 1980 (earliest data available) to 1995 records an 8.9 percent decline in real income before transfers for the lowest quintile income group. Over the same period, transfers increased by 29.4 percent producing a net increase of 10.9 percent in total real income of the lowest quintile. Falling earned income of the lowest income quintile being offset by rising government transfers reflects increased dependency and significant policy failure.

Independence or self-sufficiency as measured by the percent of income earned before transfers is displayed in the top panel of table 6. All income groups experienced a decline in independence because of the large and widespread rise in transfers. However, the sizeable decline in independence of the two lowest income quintiles is remarkable at a time when immense efforts were under way to improve the circumstances and independence of lower income groups.

Dependency measured as the percent of transfers in total income is displayed in the lower section of table 6. Steady increases in already high levels of dependency among lower income Canadians were recorded over the 1980 to 1995 period. Transfers accounted for 60.4 percent total real income of the lowest quintile in 1995 compared to 51.8 percent in 1980. Dependency also rose sharply in the second lowest income quintile, from 21.6 to 34.8 percent.

Overall, it is quite apparent that the circumstances of the lowest income quintile actually deteriorated through falling earned income and rising dependence on transfers over the 1980 to 1995 period. Accordingly, based on the 1980-95 data, it is probable that the very modest improvement in lowest quintile income share between 1965 and 1995 was more than accounted for by increased transfers. Based on this evidence, we would submit that the public policy objective of higher earned income and less dependence on transfers for the lowest income quintile was not met.

The role of income redistribution in big government countries

Transfer payments are central to big government operations everywhere. Table 7 displays government spending in the G7 countries broken down between transfer payments and other government spending, which is comprised of goods, services and investment spending. The relationship between size of government and the level of transfer spending appears quite closely related to the size of government. The two smallest government countries support transfer payments amounting to only 12 to 14 percent of GDP. The two largest government countries support staggering levels of transfer payments amounting to 28 to 29 percent of GDP, about twice the level of transfers supported by small government countries.

Comparing columns two and four in table 7 indicates that there is no apparent systematic relationship between size of government and conventional, non-transfer-payment government spending. Note in particular that both the two-smallest and the two-largest government sector countries record very similar levels of non-transfer payment spending. All seven countries record conventional government spending within a relatively narrow range, 23 to 31 percent of GDP compared to a range of 12 to 29 percent for transfer payments.

In practice, transfer payments are big government commitments to improve the circumstances of identifiable groups. Conventional government spending does not produce immediate, specific and identifiable results by way of improving the circumstances of target groups. Transfers, on the other hand, produce immediate, immediate income or other benefits. Accordingly, transfer payments have become an important means of maintaining support for big government through the provision of income or other identifiable benefits to target groups. Political attractions aside, the economics of income redistribution are quite unsatisfactory. Income redistribution initiatives tend to be expensive, expansive and ultimately counter productive. First, income redistribution is expensive because it is difficult to identify and target the needy, or the worthy, with precision. Accordingly, redistribution schemes are open to cheating and subject to political pressure for ever-wider application. Second, income redistribution is counterproductive because it creates work disincentives. As discussed above, evidence in Canada suggests that income redistribution to the lowest income quintile over the last generation has increased dependency while simultaneously reducing work effort.

Despite serious drawbacks, income redistribution will likely remain a major element in the operation of big government because it can deliver politically where conventional government spending cannot. The political advantage of being able to deliver income and other benefits to target groups on a timely and fully identified basis is considerable. As a testament to the staying power of income redistribution schemes in Canada, the Atlantic region has been a victim of massive government transfers for decades despite the obvious moral hazard and compelling evidence of the damage being done.

Security of persons and property under big government

Security of persons and property is among the most important responsibilities of government. There has been a large, generalised increase in crime rates since the early 1960s across the western industrial countries. (Demographic factors contributed to some of the increase in crime rates, but beyond these influences the rise in crime rates has been very large and wide spread across the western industrial countries.) The rise was sharpest until the early 1980s, modestly higher over the next decade and followed by a modest decline since the early 1990s. Setting aside the complexities of crime rate statistics, the compelling development has been the inability of governments of any size to contain the vast increase in crime rates recorded over the last generation. The Canadian data presented below is representative of developments in other western industrial economies.

The massive rise in crime rates recorded over the last generation has been a sign of failure for Canada, and indeed for all of the western industrial countries. Such a large and widespread lapse in the capacity of western governments to maintain high standards of peace and security is regrettable. That it should have developed alongside a massive increase in government sector resources in most countries makes the crime explosion even more disconcerting. It is interesting to note that the failure of governments to limit the crime explosion did not stand in the way of an ongoing massive increase of new resources being delivered to most governments. In fact, the rising crime rate probably assisted governments in gaining more resources on the implicit assumption that more resources for government would somehow reduce or contain the rise in crime.

In retrospect, society’s approval of the post-1960 expansion of the government sector in Canada and elsewhere was not sufficiently critical and demanding of results in the key areas of government responsibility. As we observe elsewhere in this paper, it is results rather than vision and public policy planning skills that should determine public sector resource allocation. Governments unable to control massive increases in crime rates or discharge other core responsibilities are poor candidates for even more money. Against this backdrop, there is a case for a more transparent and critical appraisal of public-sector performance and resource allocations. Where public policy is clearly deficient, other arrangements have to be considered, particularly market-determined or market-assisted alternatives.

Measuring the return on public expenditures: Canada and abroad

We present below an examination of the record of economic and social achievements of six selected G7 countries. The countries chosen, United States and Japan, United Kingdom and Canada, Italy and France, are representative of countries with small, medium and large government sectors. In important instances such as economic growth, taxes and unemployment, the evidence indicates that expansion of the government sector has not met its objectives and has made circumstances worse.

Taken together, the evidence suggests that the costs of big government exceed the benefits. Specifically, outstanding economic and social outcomes have not been achieved in big government countries relative to small government countries. This failure to deliver results must be contrasted with the vast increases in government generated overhead imposed on big government countries over the last generation. Measured against results in Canada and other big government countries, there is simply no evidence of economic and social improvement on a scale that would warrant the costs and impositions associated with vast expansion of the government sector.

We examine below evidence of the impact of the expansion of the government sector on economic and social performance indicators in Canada and abroad. Our premise is that the large costs involved in a major expansion of the public sector can only be justified if they produce significant additional benefits. If the costs of big government fail to produce a measurable net benefit, they leave society poorer than it would have been had the resources appropriated by government been left under the control of those who produced them.

The measures and comparisons reviewed here are not precise, and comparisons made internationally are that much more difficult again. Accordingly, our approach is to rely on the weight of evidence, or lack of the same, in evaluating the costs and benefits of big government. Measurement difficulties aside, the immense costs and dislocations associated with public sector appropriation of half or more of total national output are real and unmistakable. Given the high costs of big government, it is appropriate to identify associated benefits, if any. An absence of clear evidence of significant benefits being generated by the added costs of big government destroys the case for big government, at least from the perspective of those who are paying for it.

Accordingly, we look for evidence of improvements in economic and social circumstances that could reasonably be attributed to big government and thereby be counted as offsetting the considerable costs and impositions of big government. If big government does add value, there should be some measurable evidence of improved performance in the economic and social indicators in big government countries as compared with small government countries.

Taken together, the economic and social indicators reviewed below provide no case for big government. Of critical importance, the weight of evidence indicates that small governments actually produce better results on key measures including economic growth, unemployment and taxes. On other economic and social indicators, beyond the output, employment and tax measures noted above, the evidence is that small government countries produce results comparable to big government countries without the debilitating costs of big government.

Looking forward, the critical factor is economic growth. Over extended periods of time, even small increases or decreases in the underlying rate of economic growth have a profound impact on future output and living standards. Small governments deliver more rapid, long-term economic growth and are thereby able to raise living standards faster than a larger government could.

Economic and social performance indicators - G7 countries

Table 9 presents a number of economic performance indicators. Where possible, we make historical comparisons, but historic data is not available for all of the indicators. We compare performance among six G7 countries, two with relatively small governments (the U.S. and Japan), two with medium-sized governments (the U.K. and Canada), and two with large governments (Italy and France).

The economic growth trends of the G 7 countries displayed in table 9 are representative of the relationship between government size and trend rates of economic growth across both the advanced and newly industrialised countries. Using G7 countries as a reference, trend rates of economic growth in the post-1966, big-government era are measurably more rapid in small government countries compared to big government countries, both on a total economy and per capita basis. As discussed in an earlier section, the slower economic growth trends in big government countries are accompanied by higher rates of unemployment.

Without exception, big government countries are high tax countries as a direct consequence of their high spending levels. In North America and Japan, high-income marginal tax rates are closely tied to size of government, below 30 percent in the US and Japan and above 50 percent in Canada. In Europe, the link between big government and high-income marginal tax rates is mixed depending on exemption structures and the degree of reliance on sales and payroll taxes. Regardless of how taxes are collected, however, big government universally requires high taxes. In this connection, small government countries score better than big government countries on measures of economic freedom, principally because they have lower tax and regulatory burdens.

While taxes are tied to the size of government, public indebtedness clearly is not. The lack of a systematic relationship between government size and public indebtedness indicates that the choice to finance big government with borrowing rather than higher taxes is more a political than a financial choice. Labour force structure as measured in terms of participation rate and number of discouraged workers also appears unrelated to size of government. Although not captured by the measures in table 9, big government tends to produce labour market rigidities that restrict economic growth and raise unemployment as has been discussed above. Saving and investment do not appear to be closely related to government size. However, measurement in both of these areas is difficult and even more so in respect to international comparisons. Finally, consumer price inflation also looks to be unrelated to government size.

Overall, it appears that economic freedom, non-punitive tax rates, economic growth and unemployment are all favourably influenced by small government. These small government benefits are delivered without the considerable additional cost of big government. In other areas such as labour force structure, saving, investment and price inflation, big government countries appear to perform no differently than small government countries. In these areas, the high costs of big government are wasteful because they do not deliver measurable benefits beyond those available with small government.

Table 10 sets out a number of social indicators for the countries discussed above. If big government overheads are improving social conditions, it simply does not show in the social indicators reported by the UN and the OECD. There are differences in the indicators among countries, but the differences are unrelated to the size of government. In general, the social indicators presented here suggest that factors other than size of government are instrumental in shaping social conditions. Family, tradition, religion and cultural values appear to be more important than size of government in influencing social conditions.

Of the social performance indicators presented in table 10, even the number of doctors appears to be unrelated to government size, with big government countries having about the same number of doctors relative to population as small government countries. Countries with small governments generally achieved good rankings in the United Nations measure of human development. However, of the six countries examined, the two biggest government countries received the poorest (Italy) and the second best (France) human development ranking. Overall, the size of government appears to be unrelated to UN human development measures.

Further reflecting the lack of a systematic relationship between government size and social performance, the two small government countries have both the highest (Japan) and lowest (U.S.) life expectancy at birth, and the highest (U.S.) and lowest (Japan) infant mortality. The two largest government countries recorded the highest (France) and lowest (Italy) number of suicides per 100,000 population. As well, the two mid-sized government countries had both the highest (U.K.) and lowest (Canada) number of prisoners per 100,000 population in 1993. Overall, social indicator measures do not appear to be much influenced by size of government.

A budget constraint for government: 30 percent of GDP

A number of studies (see below, and the bibliography) have examined the relationship between size of government and various economic and social performance indicators for a wide variety of countries. Generally, these studies indicate progressively more adverse consequences when the government sector is expanded beyond an efficient size, variously estimated in the range of 20 to 30 percent of GDP. We note in particular that none of the published research on size of government points to any net benefit associated with large-scale expansion of the public sector. In fact, the evidence suggests that the expansion of government beyond 20 to 30 percent of GDP range produces increasing under performance.

Scully (1991) examined 103 countries with respect to the impact of taxation on economic growth. His models estimated that economic growth was maximised when total tax revenue was limited to 19.3 percent of GDP. As taxes rise beyond this level, the trend rate of economic growth declines and approaches a zero rate when taxes reach 45 percent of GDP. Peden (1991) examined the influence of government size on productivity growth in the United States from 1888 to 1986. His analysis indicates that productivity growth increases with expansion of the government sector until the size of government reaches about 17 percent of GDP. He attributes the slowdown in US productivity growth in the 1970s and 1980s to a dramatic increase in the size of the government sector that rose from 17 percent of GDP at the end of World War II to 35 percent in 1986. He concludes that to raise the rate of productivity growth back to its historic trend the size of government would have to be reduced.

Grossman (1988) examined the impact of the absolute and relative size of government on economic growth in the United States over the 1929–1992 period. He found that, measured against absolute size, the benefits of increasing government size are offset by associated costs. With respect to increasing the size of government relative to the size of the economy, he found a significant net negative impact on economic growth.

International Monetary Fund and World Trade Organisation economists Tanzi and Schuknecht (1995, 1997), have collaborated on studies examining the growth of government in industrialised countries and concluded that there is considerable scope for reducing the size of the state, and associated overheads, without compromising economic and social well-being. This conclusion is based on two key observations:

• Most important social and economic gains can be, and indeed have been in the past, achieved with drastically lower levels of public spending than prevail today among most industrial countries.
• Government spending in excess of 30% of GDP produces little or no improvement in economic performance or social conditions compared to the record of small government countries which contain the size of the government sector to the area of 30 percent of GDP or less.

Tanzi and Schuknecht recognise the important, government-supported improvement in social indicators that was recorded between the late 19^th century and mid-20^th century. Impressive gains were associated with expansion of the government sector from about 10 to 30 percent of GDP. They argue, however, that continued expansion of government in big government countries post-1960 involved considerable cost and produced little or no measurable improvement in economic performance or social indicators. The Tanzi-Schuknecht results are consistent with Canadian experience.

The record in Canada certainly indicates that an approximate 50 percent increase in the size of the public sector over the last 30 years has simply not produced the results anticipated. Worse still, excessive government overheads have almost certainly retarded improvements in living standards as a direct consequence of costs associated with such a massive expansion of the public sector. Research at The Fraser Institute (see Emes and Samida, forthcoming) indicates that the sharp drop in the trend rate of economic growth recorded over the last 30 years in Canada is related to excessive expansion of the government sector. Below-potential output growth over the last generation has cumulated to produce current output levels well below what could have been achieved with a smaller more efficient government sector.

Econometric models developed by Scully and adapted for Canada by The Fraser Institute estimate the growth-maximising size of government in Canada at about 30 percent of GDP (see Emes and Samida, forthcoming). These models also indicate that Canadian output levels in 1995 would have been more than 50 percent above the levels actually recorded, if the size of government had been contained at the 30 percent of GDP level since the mid-1960s.

---

info@fraserinstitute.ca You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.

If you know someone who would be interested in this web page, please enter their email address below, and we will forward this URL to them: Email Address:

Last Modified: Thursday, May 20, 1999.