Section 3: Feature article
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
Slowdowns during periods of economic growth
by Philip Cross 1
In the four months ending in October, employment growth totalled only 0.1%. This follows a four-month period ending in July when preliminary estimates showed real GDP growth slowed to 0.3%. 2 As noted in the CEO overview last month, four-month pauses in both output and employment actually have been quite common in the last three decades. Since 1981, there have been six distinct periods when, over four consecutive months, output has grown by 0.3% or less and been accompanied by job growth of 0.1% or less, excluding recessions (when obviously jobs and output contracted for extended periods). As well, there were four other periods when real GDP growth slowed markedly but when employment growth did not.
The purpose of this paper is to examine periods when growth slowed or turned slightly negative, without the economy falling into recession. The fact that these slowdowns often occur in both output and employment shows that the economy is clearly pausing, and these episodes are not just 'noise' in the data. The paper finds that slowdowns occur in response to both cyclical and irregular factors but rarely, if ever, develop into recessions. This provides insight into the fundamental differences between recessions and slowdowns.
Between June 1983 and October 1983, output posted a 0.1% gain, and registered monthly declines in July and October. Employment followed suit, recording no net change between September 1983 and January 1984, with small declines in two of these four months.
The 1983 stall in growth is surprising because the recovery from the 1981-1982 recession is widely-cited as the gold standard for recoveries, with growth in 1983 outstripping the recoveries after the 1990 and 2008 recessions. This pause is easily overlooked, as brisk growth quickly resumed in both output and employment and quarterly GDP growth remained near 1.0%. This period seems to represent the economy shifting from the rapid rates of 2% growth early in the upturn to a more sustainable pace, and thus had no lasting significance since quarterly GDP growth remained strong.
In late 1984, growth slowed to just 0.3% in the four months ending in September for real GDP, including a sharp 0.6% drop in September. A strike in the auto industry was partly the source of the slowdown, which probably explains why employment was less affected (people on strike are still treated as employed in the Labour Force Survey). The slowdown in output had no lingering effect, and rapid growth quickly resumed over the next six months.
There was a pronounced cyclical slowdown which persisted for most of 1986, with eight monthly declines in real GDP during the year. Overall, real GDP rose only 0.2% over the four quarters of 1986. The weakness in output led to almost no employment growth, although jobs fell outright in only one month in 1986.
The main source of the weakness in 1986 originated in a rapid drop in the price of oil. This quickly led to cutbacks in the oilpatch, which triggered a 15% drop in business investment in structures in the last three quarters of the year. The stimulus to other sectors from lower energy prices and interest rates took longer to emerge in 1987 and 1988.
The 1987 stock market crash had little perceptible impact on output and jobs. After rapid gains in 1987 and 1988, growth slowed briefly early in 1989. Real GDP did not grow at all in the four months ending in June, while jobs stalled in the four months to July. However, quarterly GDP and jobs continued to grow slowly. The weakness was concentrated in exports, as the US economy slowed at the same time as the Canadian dollar appreciated. As such, this pause was clearly due to cyclical and not irregular factors. Growth quickly resumed in the second half of the year. The economy eventually slid into recession in 1990, when oil prices jumped in response to Iraq's invasion of Kuwait.
The weak recovery of the early 1990s was followed by a pronounced slowdown in 1995. Real GDP stalled in the four month period ending in April, fell in three of the last four months of 1995 and dipped twice more early in 1996. The net result for quarterly GDP was no growth in the second and third quarters of 1995 and only a 0.1% gain in the first quarter of 1996. Employment also rose by 0.1% or less in the four months ending in May 1995, and remained weak thru the end of the year. A slow upturn in output in 1996 was not enough to induce higher employment, which stalled in the last quarter of 1996.
The 1995 slowdown reflected turmoil in the global economy, notably rising interest rates in the US which helped to precipitate the Mexican peso crisis. Exports fell in the first half of the year, while higher mortgage rates led to a 20% drop in housing in Canada from its high in 1994. At the same time, the federal government in Canada embarked on an austerity program to eliminate its budget deficit. The slack late in 1996 was partly related to a strike in the auto industry.
The onset of the Asian crisis in mid-1997 had direct and indirect impacts on the Canadian economy. While the US sustained our export growth in 1998, exports to Japan and non-OECD countries fell about one-third from their peaks in 1997, shaving over 3% off total exports. Moreover, oil prices fell steadily (to under $10 per barrel in late 1998), and output growth slowed in the first half of 1998. In addition, Eastern Canada saw a severe ice storm early in the year, which hampered all types of investment projects. All these factors resulted in a large build-up of non-farm inventories in the first quarter, which were drawn-down over the summer. Real GDP fell 0.1% in the four months ending in July, but jobs rose steadily. The slowdown in GDP was confined largely to inventories in the second quarter. Both output and job growth picked up sharply in the second half of the year, as the Federal Reserve Board's cuts to interest rates were sufficient to ward off the dampening effect of Russia's default on its debt (which helped precipitate the collapse of the giant LTCM hedge fund). 3
The bubble in high-tech stocks finally burst in late 2000. The large loss of wealth as these stocks collapsed and the sudden drop in investment by these companies sent the US economy into recession in the first three quarters of 2001, and Canada's exports fell 5.3% during this period. The extraordinary stimulus from monetary and fiscal policy after the September 11 terrorist attacks on the US helped revive demand late in 2001, followed by an upturn in jobs early in 2002.
In Canada, the impact of these events was muted, with no outright recession partly because the ICT sector comprised a smaller share of Canada's economy. Growth was weak enough that real GDP posted no net change over the second and third quarters. This cyclical slowdown was reflected in slower job growth only in the second quarter. Canada's economy quickly turned up as US demand revived, notably for autos.
Growth continued steadily in 2002 and into early 2003. Then followed a series of unrelated shocks to the economy, ranging from outbreaks of SARS and 'mad cow' disease to Hurricane Juan and Ontario's electrical blackout. 4 The net result was to lower four-month real GDP in March and April followed by a 0.9% drop during the August blackout. The one-quarter 0.1% dip in GDP was accompanied by no change in jobs in the second quarter. This period is clearly the sixth growth pause covered by this paper, but one due to a series of irregular events, not to a cyclical slowdown.
Canada then entered into one of its steadiest periods of economic growth, interrupted only by a sudden dip in output in March 2005.
However, output growth began to falter in the spring of 2006, followed by a deceleration in employment over the summer. As a result, real GDP growth slowed to 0.1% in the second and third quarters. The slowdown's origins appear to be irregular events, including a lowering of the GST rate on July 1 which shifted spending from spring into summer, and the winding down of Census operations. 5 However, this growth stall was mild and growth quickly rebounded. Job growth was less affected, with the low being a 0.3% gain in the third quarter.
Growth was steady in late 2006 and into 2007, despite the onset of turmoil in global financial markets in the summer of 2007 related to the deterioration of housing loans in the US.
In December 2007, several auto firms simultaneously closed plants in order to reduce inventories or to retool for new models. The effect of the unprecedented 26% drop in auto assemblies was to lower real GDP by 0.6% in the month. However, the one-month drop in output was not enough for employers to reduce their payrolls, and employment growth was steady well into 2008. After a rebound in January, output in the first half of 2008 was hampered by a series of irregular events, ranging from record snow in parts of the country to a new February holiday in several provinces and significant production problems in the oilpatch.
This paper has identified 10 periods since 1982 when either output or both output and employment growth essentially stalled over a four-month period. Excluding recession years, growth pauses occurred on average almost every two years, making them a regular, if not well-understood, feature of the economy. Of these 10 episodes, four can broadly be attributed to cyclical slowdowns in both output and jobs (1986, 1989, 1995 and 2001). The other six were related to a variety of irregular events, some of them economic (such as strikes) and some of them not (such as extreme weather or the electrical blackout), including some too mild to be reflected in employment.
Looking back at these periods of slow growth or no growth, a few common features become apparent. There were seven periods when both output and employment slowed, and three periods when only output slowed markedly (1984, 1998, and 2007). In two of these seven episodes when both output and employment paused, the slowdown in output occurred before employment stalled, while in the other five the slowdown occurred in the same quarter.
In all of the episodes when both GDP and jobs stalled, the slowdown in output was more pronounced than the slowdown for employment. In fact, employment did not fall for even one quarter during the slowdowns identified in this paper. Output is more volatile than employment: during short periods of lower production, especially those due to unexpected or irregular events such as poor weather or maintenance, it would not be rational to lay-off employees and to incur severance costs as well as the cost of searching and training new employees when output resumes. More fundamentally, quarterly employment in the last three decades has only fallen during outright recessions, a reflection that employers adapt to brief fluctuations in output by adjusting hours rather than employment. 6
Slowdowns versus recessions
How do these periods of slow growth compare with recessions? The declines that mark recessions are several magnitudes greater, in both duration and the severity of the weakness in GDP and in jobs. The slowdowns this paper examined showed, at most, small declines over a period of one or two quarters per episode, with no net decline during the slowdown, after which the economy resumed growing.
By comparison, the 1981-1982 recession saw six straight quarterly declines in both output and employment, with a total drop of 5% for both. In the 1990-1992 recession, output fell 3.4% over four quarters, while jobs dropped 3.2% over eight quarters. In 2008-2009, both output and employment posted three consecutive quarterly losses totalling 3.3% for GDP and 1.8% for jobs.
Another difference between recessions and slowdowns is the breadth of the decline which results from each. Slowdowns are confined to one or two sectors of the economy, usually exports or inventories and sometimes housing. The recessions starting in 1981, 1990 and 2008 saw marked declines everywhere except for government: consumer spending, housing, business investment and exports all posted declines. Moreover, all began to contract in tandem with the onset of recession, except exports in 1990 (when they fell one quarter after the rest of the economy).
The broad-based nature of recessions reflects a fundamental difference between the source of these recessions compared with the source of growth pauses, irrespective of whether the latter was due to cyclical slowdowns or irregular events. All three of the recessions starting in 1981, 1990 and 2008 were due to massive shocks that reverberated throughout the financial system. The 1981-1982 recession was driven by interest rates in North America hitting over 25% as the inflation of the 1970s was decisively wrung out of the economy. In 1990-1991, a spike in oil prices related to the Iraq invasion of Kuwait helped to send interest rates above 15%, triggering a recession even before the introduction of the GST and the Gulf War further depressed economic activity early in 1991. The 2008-2009 recession had its origins in the seizure in global financial markets after the failure of Lehman Brothers and AIG in the US.
The events that triggered the slowdowns identified in this paper were different from the systemic shocks to the whole financial system that lead to full-blown recessions. The oil price shock in 1986, the stock market collapse in 2001, the 1998 Asian financial crisis, the bond market sell-off and Mexican peso crisis in 1994, the 1998 ice storm and the 2003 electrical blackout were each a major event. However, the effect of these events on the economy was mostly a loss of potential growth, not a contraction in output and jobs. This difference probably is because none of these events sent repercussions throughout the financial system such as those that occurred in 1981, 1990 and 2008, although there were shocks to individual financial markets (such as stocks in 2001 and bonds in 1994). This is a difference worth remembering in assessing the impact of future shocks to the real economy and to the financial system.
The argument could be made that slowdowns sometimes weakened the economy enough to make it more vulnerable to a large scale financial shock, but these shocks that occurred in 1981, 1990 and 2008 were so large that they would have led to a recession. In Canada, one example of slow growth preceding recesssion was in the first half of 2008, when real GDP stalled over a six-month period, although employment growth remained vigorous. However, the ensuing recession late in 2008 was clearly the product of the conflagration in the international financial system in September, not the inevitable outgrowth of the irregular factors that checked Canada's GDP growth in the first half of the year. 7 This is borne out by the steady gains in both jobs and in the leading indicator into the fall of 2008.
The US provides another example of a growth pause preceding a recession caused by a sudden shock to the financial system. Real GDP levelled off thru most of 1974 at a time of rising energy prices, but the economy remained strong enough to sustain employment growth and commodity prices stayed high. However, the 1974 slowdown turned suddenly into a severe recession in the US after the Franklin National Bank went insolvent in October 1974, one of four major banks with over $1 billion in assets to disappear during the recession. 8 The real economy quickly unravelled. Employment fell 2.8% between October 1974 and April 1975, the worst six months in post-war history before 2008. The rate of decline of real GDP in the US tripled to 1.2% in the first quarter of 1975. This financial stress and bank failures in Europe led to the creation of the Bank of International Settlements to address the so-called problem of Herstatt Risk. 9
The implication is that the recent slowdown in output and employment growth has few lasting consequences for future growth and almost no bearing on the risk of a renewed recession. Such pauses in growth are a recurring feature of the economy outside of recessions, and happen for both cyclical and irregular reasons.
It is unclear whether the stall in GDP and job growth in mid-2010 was driven primarily by cyclical or irregular factors. It may represent the end of the inventory rebuilding part of the recovery. Alternatively, this slowdown could reflect the economy down-shifting from its initial rapid recovery from the 2008-2009 recession, aggravated by the end of irregular events that raised growth early in the year, such as the Vancouver Winter Olympics and a record warm winter and spring. The introduction of the HST in Ontario and BC also helped dampen demand for services, which were unusually weak over the summer.
- Date modified: