This Selected Issues paper for Canada presents comprehensive and broad-based analysis of the role of domestic and external shocks. Canada’s economic history illustrates the important role played by external as well as domestic macroeconomic disturbances. Canada’s economy slowed in 2001 because of the global slowdown, although by less than in many other countries. In 2003, the recovery has been interrupted by a series of shocks that moderated growth. Fluctuations in Canadian real GDP are explained by external and domestic cycles.

Abstract

This Selected Issues paper for Canada presents comprehensive and broad-based analysis of the role of domestic and external shocks. Canada’s economic history illustrates the important role played by external as well as domestic macroeconomic disturbances. Canada’s economy slowed in 2001 because of the global slowdown, although by less than in many other countries. In 2003, the recovery has been interrupted by a series of shocks that moderated growth. Fluctuations in Canadian real GDP are explained by external and domestic cycles.

PART III: STRUCTURAL ISSUES

V. How Flexible is The Canadian Economy? An International Comparison65

A. Introduction

1. The openness of the Canadian economy leaves it susceptible to external shocks and puts a premium on economic adaptability. Accordingly, this paper compares the degree of flexibility of Canadian economy with that of other major industrialized countries. Flexibility is an important economic concept, but a difficult one to measure. Rather than focusing on one approach, this paper uses a range of different measures of economic adaptability:

  • Industry-level data on real value added is used to compare the degree to which countries have changed their economic structure, both over long periods of time and from year to year. The latter is a measure of the amount to which resources flows across sectors in response to changes in economic conditions, which can be characterized as economic “churning.”

  • A survey of microeconomic studies compares rates of entry and exit of individual firms as well as gross job creation and destruction across industrial countries, and supplement the industry-level analysis of churning.

  • Estimates of Phillips curves across countries measure the speed with which the economy responds to macroeconomic disturbances.

2. The results uniformly suggest that Canada is characterized by a relatively high degree of flexibility, of magnitude comparable if not larger than many other industrialized countries, with the likely exception of the United States. Industry-level data suggest Canada has undergone a deep transformation of its industrial structure over the last 20 years and that it has been able to respond rapidly from year to year to changing circumstances. In addition, a relatively large number of firms enter and exit the market every year, while the share of gross jobs created and lost in the Canadian manufacturing sector every year has been comparable to that in the United States. Phillips curves also point to a relatively strong ability to respond to macroeconomic disturbances.

B. Industry Data

3. The Canadian economy has undergone substantial structural changes over the past 20 years, largely reflecting the declining importance of the primary sector. Figure 1 shows that the share of value added of the Agriculture and Mining sectors has declined more rapidly in Canada than in other industrialized countries over this period (Figure 1). In contrast with other countries, the manufacturing sector’s share has increased over time, while the service sector’s share has not increased as rapidly.

Figure 1.
Figure 1.

Sectoral Change Across Countries

Citation: IMF Staff Country Reports 2005, 116; 10.5089/9781451806984.002.A005

Source: OECD, Structural Analysis Database.

4. Standard indexes suggest that Canada has been relatively successful in shifting resources across sectors. Two measures of structural change in response to disturbances are the Structural Change Index (SCI), which measures changes over extended periods, and the Lilien index, which focuses more on short-term churning (Box 1, Table 1). The Lilien Index suggests that Canada has experienced more churning across sectors than other G-7 countries, as well as the Netherlands, and Spain, but Canada scores lower on the SCI measure of long-term structural change.66

Table 1.

Indexes of Structural Change

article image
Sources: OECD, Structural Analysis Database; and Fund staff calculations.

Indexes of Structural Changes

The Structural Change Index, which measures change over extended periods, is given by:

SCI=12Σi=1N|lnsi,Tlnsi,T|

where si,t is the real value added of sector i = 1…N at time t. The index is calculated over three different time periods (for example, the index for the period 1980-2000 is based on the difference between shares in 2000 and those in 1980). To make the index less dependent on the business cycle, industry shares are averaged over 4 years at the beginning and end of the relevant period (for example, averages over 1980-83 and 1997-2000).

The Lilien Index, which focuses more on churning across sectors, is given by:

L=1T1Σt=1T[Σi=1N(si,tst)(Δ1nsi,tΔ1nst)2]1/2

where st is the real value added for the economy/manufacturing.

5. Estimates that adjust for secular trends and persistence in shocks suggest that Canada has experienced more churning than other countries except for the United States. The Lilien Index results for Canada could reflect the fact that it has been subject to relatively persistent shocks, leading to secular trends and more persistence in economic adjustment. To test for this, this paper estimates simple univariate regressions of the form:

ΔVAit=ci+ριΔVAit-1+ϵit(1)

where VA is the logarithm of sectoral real value added.67 The autoregressive coefficients (ρt), which measure the persistence of sectoral shocks, indeed suggest that Canadian shocks are relatively persistent across industries (Table 2), while the coefficient ci takes account of secular trends in value added for each sector.68 The standard deviation of the error term is then a measure of the churning across sectors (Table 3). The value for Canada is estimated to be lower than in the United States but somewhat higher than Japan, and considerably higher than the European countries.

Table 2:

Median Unbiased Estimates of First Order Autoregressive Coefficient, Real Value Added

(Δ V A = c + ρ Δ V A(-1) + ε) 1/

article image
Sources: OECD, Structural Analysis Database; and Fund staff calculations.

Data are annual, covering the period 1981-2000.

Table 3:

Standard Error of First Order Autoregressive Coefficient, Real Value Added

(Δ V A = c + ρ Δ V A(-1) + ε) 1/

article image
Sources: OECD, Structural Analysis Database; and Fund staff calculations.

Data are annual, covering the period 1981-2000.

C. Firm-Level Data

6. Microeconomic studies have compared responses of Canadian firms to changes in the environment. A scarcity of comparable cross-country data on firms’ behavior restricts the scope for international comparison. However, some analysis has been performed on the demographics of firms across sectors and countries using data collected by the OECD. In addition, a Statistics Canada study has compared gross job flows—a measure of the relative flexibility of the labor market—between Canada and the United States.

7. The OECD firm-level study suggests that Canada has a relatively high amount of firm entry and exit (Figure 2). During the 1990s, firm turnover (entry plus exit rates) in the business sector was on average around 20 percent in Canada, a rate second only to the United States among the countries covered by the database (Bartselman et al, 2003). Entering firms also had a relatively large probability of failing within the first two years.69 These results suggest that it is relatively easy for small Canadian firms to obtain external financing and enter the market, where they face strong pressures.

Figure 2.
Figure 2.

Turnover rates in OECD countries

Citation: IMF Staff Country Reports 2005, 116; 10.5089/9781451806984.002.A005

Source: OECD, Structural Analysis Database.

8. A study by Baldwin et al. (1998) also suggests that churning in the Canadian labor market is comparable to that in the United States. About one in ten manufacturing jobs was created in Canada each year between the early 1970s and the early 1990s, and roughly the same number was lost each year. The reallocation rate—the sum of jobs created and jobs eliminated—was about 22 percent in Canada compared with 19 percent in the United States.70 The study also suggests that much of the job reallocation in both countries reflected shifts in employment opportunities across firms in the same sectors, rather than between industries; and that the industries showing high (low) job creation and destruction in Canada also were the ones showing high (low) job reallocation in the United States. This supports that sectoral differences in the rate of job creation/destruction are mainly explained by common technological characteristics of the industries in the two countries.

D. Macroeconomic Data

9. An alternative approach to measuring flexibility is to look at relationships at themacroeconomic level. A particularly simple characterization of the economy involves a Phillips curve, IS curve, and monetary reaction function.71 A typical specification is:

Phillips curve:πt=βπ4,t+4e+(1-β)π4,t-1+γyt+ϵtπ(2)
IS curve:yt=δ0+δ1(it-πt+1e)+δ2yet+1+(1-δ2)yt-1+ϵty.(3)
Taylor rule:it=α0+(1-ρ1-ρ2)(α1π4,t+4e+α2yt)+ρ1it-1+ρ2it-2+ϵti.(4)

where πt is the annualized quarterly CPI inflation rate, π4t is the annual CPI rate of inflation, yt is the output gap (measured by detrending real GDP using a Hodrick Prescott filter with a smoothing coefficient of 1,600), and it is the policy interest rate.

10. The Phillips curve embodies the supply response of the economy and provides a measure of macroeconomic flexibility. A forward-backward looking Phillips curve is used, which is broadly consistent with new-Keynesian models of inflation, and measures responses to supply shocks in terms of the degree of inflation inertia (measured by the coefficient 1-β) and response of inflation to deviations of output from potential output (γ). In such a model, a lower level of inflation inertia implies a measure of the flexibility of the economy, as it raises the speed at which the economy is able to absorb aggregate demand and supply disturbances, as discussed in more detail in Bayoumi and Sgherri (2004). By contrast, the relationship between the coefficient on the output gap and flexibility is ambiguous. While prices respond more rapidly in sectors of the economy that wish to attract resources, prices will also fall by more in sectors that should release them for the same reason.

11. Accordingly, estimated Phillips curves were used to measure Canada’s flexibility compared to other major industrial countries over 1980-2003. Equation (2) was estimated using generalized method of moments (GMM) with the first four lags of inflation, the output gap, and the interest rate as instruments. The sample included the G-7 countries and Australia, which, like Canada, is a sizeable industrial country relatively specialized in commodity production. All of the coefficients in Table 4 are correctly signed, and those on forward-looking inflation are consistently highly significant. The coefficients on the output gap are estimated with considerably less precision and are not discussed further.

Table 4.

Phillips curve estimates 1/

π4=βπe4,t+4+(1+β)π4,t-1+γyt+ϵπt

article image
Source: Fund staff calculations.

The coefficients are estimated over 1980:1-2003:4 using GMM with instruments being the first 4 lags of inflation, the output gap, and nominal interest rates. Robust standard errors are reported in parentheses. ** and * indicate the coefficient is significant at the 1 and 5 percent levels, respectively.

12. The results suggest that European countries (France, Germany, Italy, and the United Kingdom) all exhibit less flexibility than the non-European countries (Australia, Canada, Japan, and the United States). The average of the coefficient on forward-looking inflation (β) is 0.54 for European countries and 0.68 for non- European countries. The four European countries also have the lowest coefficient estimates in the sample, which is strong evidence that flexibility is lower in Europe, as such an outcome for these countries would occur by chance less than one percent of the time.

13. Canada’s results are similar to most of the other non-European economies. Canada’s coefficient on forward-looking inflation is 0.67, slightly higher than those found for the United States and Australia, but, given the standard errors on the coefficient estimates, these differences could well reflect noise. The most surprising result given the earlier analysis is the extremely high coefficient on forward-looking inflation in Japan. This could well reflect the relatively flexible wage structure, where almost one-third of overall wages comprise annual bonuses that vary with the state of the economy. However, this flexibility has to be seen in the context of the “jobs-for-life” environment that limited real flexibility, helping to explain why Japan appears to be less flexible in the microeconomic analysis discussed in earlier sections.

14. IS curves were also estimated to examine whether there were any systemic differences in the response of aggregate demand, but no patterns were found. The IS curve relates the current level of the output gap to its expected future and past values (with the coefficients constrained to sum to unity), as well as the level of real interest rates.72 Estimated equations could reveal differences in flexibility—for example, using similar arguments to those used for the Phillips curve, more flexible economies might have lower coefficients on lagged output gap. However, the results (not shown) indicated remarkable stability in the coefficients across countries, with most estimates of the coefficient on the lagged output gap being close to one-half and real interest rate effects being small and insignificantly different across countries.

E. Conclusions

15. The results in the paper suggest that Canada has a relatively flexible economy compared with other major industrial countries, with the likely exception of the United States. This conclusion comes from measures of economic flexibility using a range of approaches, including the response of output at the sectoral level, microeconomic studies of firm behavior, and from macroeconomic relationships. It suggests that, while on a number of measures the flexibility of the Canadian economy is lower than that of its neighbor to the south, any gap is small, particularly when compared to other industrial countries.

References

  • Baldwin J., T. Dunne, and J. Haltiwanger, 1998, “A Comparison of Job Creation and Job Destruction in Canada and the United States,The Review of Economics and Statistics, Vol. 80, pp. 34756.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bartelsman E., S. Scarpetta, and F. Schivardi, 2003, “Comparative Analysis of Firm Demographics and Survival: Micro-Level Evidence for the OECD Countries,OECD Economics Department Working Paper No. 348 (Paris: Organization for Economic Cooperation and Development).

    • Search Google Scholar
    • Export Citation
  • Bayoumi, T., and S. Sgherri, 2004, “Monetary Magic? How the Fed Improved the Flexibility of the Economy,IMF Working Paper No. 04/24 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clarida, R., J. Galí, and M. Gertler, 1998, “Monetary Policy Rules In Practice: Some International Evidence,European Economic Review, Vol. 42, pp. 103367

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Davis S. J., J. C. Haltiwanger, and S. Schuh, 1996, Job Creation and Destruction (Cambridge: The MIT Press).

  • Hansen, B., 1999, “The Grid Bootstrap and the Autoregressive Model,Review of Economics and Statistics, Vol. 81, pp. 594607.

65

Prepared by Tamim Bayoumi and Roberto Cardarelli (RES).

66

Germany is excluded from the comparison due to the impact of reunification.

67

Annual data on sectoral real value added were taken from the latest version of the OECD’s STAN database for the period 1981-2000.

68

To avoid the finite sample downward bias affecting OLS estimates, the autoregressive coefficient has been estimated using Hansen’ grid bootstrap method, which produces median unbiased parameter estimates (see Hansen, 1999).

69

Results for Canada are discussed in Baldwin et al. (2000).

70

These numbers are quite similar to those reported by Davis et al. (1996) for a series of other industrialized countries, ranging from 16 percent in Germany to 23½ percent for Sweden.

71

See, for example, Clarida et al. (1998).

72

Such a specification can be derived from the Euler equation of consumption with habit persistence in preferences.

Canada: Selected Issues
Author: International Monetary Fund