Download Slides 3

1
3. Closing the 49th Parallel: An Unexplored Episode in the
Political and Economic History of the Canadian Welfare State
This lecture makes use of: James Davies and Stanley L. Winer. "Closing the
49th Parallel: An Unexplored Episode in Canadian Economic and Political
History". Canadian Public Policy, forthcoming
Stanley Winer
Canada Research Chair Professor
Carleton University
2
The near closing of the Canada-U.S. border to emigration in
the late 1960’s:
•
U.S. Immigration Act of 1965: A radical change in immigration
regulations affecting Canada, coming into force in 1968.
•
Near closure lasted until the onset of the mobility provisions of the free
trade agreements in the early 1990’s.
•
Historically, this is a very unusual episode in relations between Canada
and the United States, one whose implications for Canadian public
policy have received little attention.
3
Figure A1
Permanent Emigration to the U.S. as a Percent of Population
Canada, 1870 - 2000
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1875
1900
1925
1950
1975
2000
4
Official Migrants from Canada to the United States, 1946 - 2005
Restriction effective 1968
FTA / NFTA
5
Implications ?
(a) for labour markets and inequality
(b) for the fiscal system
6
Why is this investigation controversial?
7
a. Implications for Canadian Labour Markets
(i) implications for labour market integration!
Table 1
Correlations of Mean Weekly Wages and Salaries
Canada and the United States, 1921 – 2008
Period
1950-1968
1969-1990
1969-2008
Correlation of
Mean Wage
Levels
0.993
0.190
-0.235
Correlation of
Percent Changes
in Mean Wages
0.635
0.426
0.338
8
(ii) implications for inequality in Canada
Gini Coefficients for Family Income, Canada and the U.S.
Five Year Averages, 1970 – 2005
Years
Canada:
Market
Income
Canada: Total
Income
U.S.: Total
Income
1965-1969
na
.328
.352
1971-1975
.378
.330
.356
1976-1980
.378
.326
.363
1981-1985
.396
.329
.381
1986-1990
.402
.331
.395
1991-1995
.422
.336
.415
1996-2000
.430
.352
.429
2001-2005
.417
.347
.437
Notes:For Canada 1965-1969 includes the years 1965, 1967 and 1969. Those are the only years available from 1965 to 1970. For the U.S. 1965-1969 includes each year from 1965 to
1969. Statistics Canada (2005) reports a historical revision for the period from 1990 forwards that increased estimated Gini coefficients.The above table uses the Gini coefficients
published by Statistics Canada for 1965–2002 before the 2005 revision. For 2003-2005 this table uses the Gini coefficients published in Statistics Canada on the revised basis (2007), but
reduces each one by the difference between 2002 Gini coefficients reported on the old and new bases.
Sources:For Canada, 1965-1969 Love and Wolfson (1976. Table 5); 1971-2000 Kesselman and Cheung (2006, Table 12.4); 2001-2005 Statistics Canada (2007). For the U.S.: U.S. Census
Bureau, Income website, http://www.webcitation.org/query?url=http%3A%2F%2Fwww.census.gov%2Fhhes%2Fwww%2Fincome%2Fhistinc%2Ff01AR.html&date=2009-04-12
9
(iii) implications for the elasticity of labour supply
OLS Regressions of Annual Changes in the Log of Civilian Labour Force
on Changes in the Log of Wages, Canada, 1946 – 1995
(Absolute value of t- statistics)
Variable
Constant
D (Log of Canadian
Wages)
D( Log of U.S. Wages)
R2
1946-1968
0.0208*
(4.84)
0.4324*
(3.49)
-0.3988*
(3.46)
0.539
1969-1990
0.0239*
(10.39)
0.1640***
(1.77)
-0.0338
(0.30)
0.151
*(**)*** = Significant at 1% (5%) 10%.
1969-1995
0.0201*
(7.73)
0.1971***
(1.76)
-0.0287
(0.20)
0.126
10
How important is Canada-U.S. migration in overall Canadian labour supply?
• Assume immigration to Canada from non-U.S. sources, F , is exogenous, and
neglect Canadians emigrating to countries other than the U.S.
• Aggregate labour supply elasticity, e, is then a weighted sum of the domestic
labor supply elasticity, eD, the wage elasticity of immigration from the U.S., eI ,
and the (negative of the) wage elasticity of emigration, eE
e = wDeD + wIeI - wEeE
where the weights wD, wI , wE are shares of end-of-period labour supply, D + F + I – E.
• During1960-68 there were an average of 44,869 official migrants from Canada to
the U.S., = 0.64 % of the average labour force for this period => wD = 0.0064. In
the same period immigration from the U.S. averaged 14,536 => wI = .002.
• Using these weights, and with eI = 5.7 and eE = 9.0 from Hunt and Mueller
(2010), we have that migration components accounted for 0.069, or 16 percent
of the 0.43 wage elasticity of aggregate labour supply in Canada before the
border was closed.
11
(iv) quantitative importance of cross-border migration for Canadian policy?
•
Were cross-border flows large enough to matter for policy choices in
Canada?
• Caution: Small flows before 1965 don't necessarily reflect what could
happen after new, 'big' Canadian policy changes (e.g., socialized
medicine).
• Develop sense of the constraints on Canadian policy choices imposed by
the threat of emigration in the following manner:
12
First, note that the gap between Canadian and U.S. total govt. expenditure to gdp
ratios widened to 10% by early 1990's:
Comprehensive Measure of Government Size: 1929 - 2004
Defense and Depreciation Adjusted with Nonprofit Hospitals and Universities Included Consistently
0.55
0.50
Proportion of GDP
0.45
0.40
38.5%
0.35
37.0%
0.30
0.25
0.20
0.15
0.10
0.10
0.05
0.00
1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004
Canada
United States
Source: Ferris and Winer (2007, Figure 8 )
Canada / U.S. (Equality = .10)
13
Then we need to make some detailed (back of the envelope) calculations:
• Suppose the burden of financing govt. expansion falls entirely on labour,
through income taxes, payroll taxes, or other taxes shifted onto labour => a rise
in the effective tax rate on labour = about 15 percent, since labour income = 2/3
of GDP.
• Starting from yearly movements between Canada and the U.S. at their 1960-68
level, and using elasticities from Hunt and Mueller (2010), this => an annual
reduction of about 13,000 in the number of people immigrating from the U.S.
and a rise in emigration by about 68,000 for net emigration of 81,000 people
per year, or 1.1 % of the labour force.
• Over the typical four-year life of a government, this would add up to 4.4 % of
the labour force.
• A counterfactual: from 1968 to 1990 the female LFP rate in Canada rose from
37.4% to 58.5%, for an average annual increase of 1.0 percentage points. In
1968 such an increase represented 71,000 women whereas by 1990 it meant
108,000. These numbers are of a similar magnitude.
14
b. Implications for the Canadian Fiscal System
(i) tax structure
(ii) overall size of government
15
Equilibria in a Competitive Political System Before and After the Closing of the Border
M arg in al
P ol iti cal
C os ts and
Benef its
P er D ol la r
T ax B ase A: E m pl oy m ent In com e
T ot al m arg in al
p ol it ical co st s
p er $
T ot al m argi nal
p ol it ical
b enefi ts
p er $
T ax B ase B: Capi tal In com e
M ar gi nal p ol it ica l co s ts pe r $
M argi n al p o li ti cal cos ts p er $
1
2
0
2
2
Ra
tA
1
1
Rb
tB
L affer cu rves n et of adm i n cos t s
Ta x R a tes
L affer cu rve net o f ad m in co st s
R1
R2
Ra + Rb
16
Table 1: The Tax Mix in Canada, Consolidated Public Sector, 1950 – 2000
As A Percent of Total Taxes
Personal Income Tax
Year
1950
1960
15.0
21.1
Corporate Income
Tax
24.3
17.0
Social Insurance
Taxes
na
na
Indirect Taxes
Local Taxes
48.4
44.9
12.37
17.02
1970
34.2
11.6
7.1
33.1
14.06
1980
35.3
13.3
9.3
30.3
11.83
1985
1990
1995
2000
35.5
40.3
37.3
37.4
10.4
7.1
7.8
12.5
11.8
12.1
14.2
12.9
31.6
30.0
30.3
28.5
10.69
10.42
10.45
8.57
As A Percent of GDP
Year
Personal Income Tax
Corporate
Income Tax
Social Insurance
Taxes
Indirect Taxes
Local Taxes
Total Taxes
1950
1960
3.2
5.1
5.2
4.0
na
na
10.3
10.6
2.60
4.00
21.4
23.7
1970
10.0
3.4
2.1
9.7
4.10
29.3
1980
1985
1990
1995
2000
10.2
10.9
14.1
13.1
13.4
3.9
3.2
2.4
2.7
4.5
2.7
3.7
4.3
5.0
4.6
8.7
9.7
10.6
10.6
10.2
3.40
3.30
3.70
3.70
3.10
28.9
30.8
35.1
35.1
35.7
17
Marginal Income Tax Rates, Canada, 1947 – 2000*
* MTR = Marginal effective tax rate. P99 refers to the 99th percentile tax bracket.
18
(ii)
Overall size of government?
19
Comprehensive Measure of Government Size: 1929 - 2004
Defense and Depreciation Adjusted with Nonprofit Hospitals and Universities Included Consistently
0.55
0.50
Proportion of GDP
0.45
0.40
38.5%
0.35
37.0%
0.30
0.25
0.20
0.15
0.10
0.10
0.05
0.00
1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1962 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004
Canada
United States
Source: Ferris and Winer (2007, Figure 8 )
Canada / U.S. (Equality = .10)
20
A broader view?
Many factors affect government size and structure
21
Comprehensively Defined Non-Defense Government Spending as a Proportion of GDP
1929 - 2004
Source: Ferris and Winer (2007, Table 10b)
Table A1: A Preliminary Model of Federal Government Non-interest Expenditure Relative to GNP (FEDERAL SIZE), 1900 – 2000
OLS estimation (Absolute value of Newey-West HAC standard errors, unless otherwise stated)
Explanatory Variables
(in log form, except for:
WWI, WWII, WWII –
AFTERMATH, FIXED
EXCHANGE)
REAL INCOME
AGRICULTURE
%YOUNG
OPENNESS
FIXED EXCHANGE
WWI
WWII
WWII-AFTERMATH
IMMIGRATION
EMIGRATION_US
1900 – 2000
1921 – 2000
1921 – 2000
Saikkonen
t’s(a)
for col (2)
(3)
1945 – 2000
(2)
1921 – 2000
Using emigration
regimes
indicators
(2a)
(1)
0.37
(0.70)
0.09
(0.03)
-0.12
(0.31)
-0.614*
(2.77)
-0.18*
(4.05)
0.75*
(5.15)
1.52*
(5.52)
0.70*
(4.43)
-0.12*
(3.80)
-0.02
(0.63)
1945 – 2000
Saikkonen t’s (a)
for col (4)
(4)
1945 – 2000
Using emigration
regimes
indicators
(4a)
0.35
(0.71)
0.14
(0.51)
0.73
(0.23)
-0.69*
(4.42)
-0.15*
(5.45)
0.14
(0.32)
0.02
(0.23)
-0.28
(-0.90)
-0.66*
(3.46)
-0.14*
(3.56)
0.96**
(2.50)
0.16
(0.75)
1.94*
(5.03)
-0.36***
(1.97)
-0.06***
(1.97)
0.40
(0.66)
0.13
(0.44)
-0.27
(0.95)
-0.91*
(4.93)
-0.13*
(4.76)
-.02
(0.04)
-.04
(0.16)
-0.45
(1.47)
-0.77*
(-4.96)
-0.14*
(5.11)
3.54*
(4.31)
1.71*
(4.20)
0.50**
(2.03)
-1.54*
(11.87)
-0.07*
(5.45)
1.49*
(5.75)
0.69*
(5.42)
-0.06**
(2.11)
-0.18*
(5.22)
1.52*
(5.43)
0.86*
(6.53)
-0.20*
(5.70)
1.14*
(9.66)
0.06
(0.45)
0.03
(0.61)
-0.41*
(6.54)
1.33*
(13.09)
0.18*
(3.26)
-0.01
(0.22)
-0.13*
(3.29)
1.35*
(13.57)
0.10***
(1.74)
-0.16
(0.33)
1.85*
(18.8)
0.71*
(8.43)
-0. 02
(0.98)
-0.32*
(7.81)
(5)
22
1945 – 2000
Emigration
regimes
Saikkonen t’s (a)
for col (4a)
(6)
-0.25
(0.74)
-0.09
(0.56)
-0.72
-1.02*
(12.25)
-0.16*
(21.56)
1.41*
(19.42)
0.33*
(5.64)
-0.05**
(2.35)
-6.41
(1.274)
-6.94
(1.46)
0.09
(1.18)
0.07
(0.55)
-4.54
(1.05)
Observations
Adj. R2
Durbin Watson
101
0.93
1.01
80
0.95
1.45
80
0.92
1.27
78
0.97
1.49
56
0.86
1.27
56
0.86
1.17
54
0.95
2.27
LM serial correlation
(Obsvs * R2): P-value
0.00
0.01
0.00
0.002
0.01
0.00
0.02
Adj. Dickey Fuller
-6.10*
-5.05**
-6.39*
-5.74*
-4.51***
-5.51
-9.97
Phillips-Perron
Saikkonen (1991) adjustment
-5.72*
-6.99*
-5.80*
-4.46***
-4.29
-24.89
3.50
-26.85
3.50
EMIGRATION_1968_1990
EMIGRATION_FTA
Constant
-18.13*
(5.05)
-5.43
(0.97)
0.16**
(2.17)
0.12
(0.90)
-0.96
(0.19)
-35.09
(4.56)
0.06*
(3.68)
0.16*
(5.08)
1.44
(0.46)
54
0.95
2.61
Statistics
1.42
23
Predicted Federal Government Size With and Without Emigration Restrictions,
1968 – 1990.
.20
.19
.18
.17
.16
.15
.14
.13
.12
.11
1950
1960
1970
1980
1990
2000
Predicted federal government size without emigration restriction 1968 - 1990
Predicted federal government size in presence of actual emigration restrictions
•
Using the elasticity form of the effect of emigration in column (4), of - 0.13, the 80% fall in
emigration (relative to population) that occurred between 1965 and 1975 has effect of increasing
federal government size relative to GNP by about 10%.
24
Conclusions
•
Amazing uncoupling of Canadian and U.S. labour markets!
•
Consequences for inequality of incomes and the fiscal
system are apparent
•
Role in overall development of the Canadian welfare state
???
•
Further study of the role of international migration in
Canadian public policy is warranted.