Presentation by Wayne Simpson on Guaranteed

Guaranteed Annual Income (GAI),
Basic Income (BI) and the Manitoba
Basic Annual Income Experiment
Wayne Simpson
Department of Economics
University of Manitoba
Key Points:
• GAI and BI are not the same thing (and there
are important differences for policy)
• Mincome GAI Experiment was much more
than Dauphin (contrary to Wikipedia)
• GAI is gaining acceptance in many forms
What is a Guaranteed Annual Income?
– an income guarantee (G) based on a family low
income/poverty level, e.g. the Statistics
Canada Low Income Cut-off, the Government
of Canada’s Market Basket Measure. A family
with no other income would receive G, e.g. the
LICO was $22,361 for a family of 2 in 2008
– a (negative income) tax rate (t) that reduces
the amount of the benefit paid as income is
earned, e.g. at a 50% tax rate a family that
earns $10,000 would receive a benefit of
What is a Basic Income?
• a specific amount paid to each individual and
subject to tax
• Alaska pays a Basic Income out of saved oil
revenues ($2,069 in 2008, $1,174 in 2011);
Western Australia considering similar scheme
• Canadian examples:
– Old Age Security used to be paid to everyone
over 65, regardless of income (like a BI); now it
is clawed back (like a GAI)
– Family Allowance used to be paid to all
mothers; now the Child Tax
Benefit/Supplement are clawed back
GAI vs. BI
• The Guarantee G: GAI targets poverty, as we
measure it, based on family income, where BI
does not
• The tax rate t: GAI targets poor households by
“clawing back” benefits as incomes rise, while
also providing incentives to work; BI goes to
everyone (pure income transfer) and is taxed in
the normal way
• BI does not target poor families as well and is
much more expensive than GAI
Manitoba Basic Annual Income
Experiment (Mincome) 1974-79
• A GAI experiment (not BI)
• 3 components:
– Winnipeg dispersed sample of low-income
households with 3 levels of G x 3 levels of t
– Dauphin “saturated site” sample with single G, t
– Rural dispersed sample (never analyzed)
• Mincome comparable to four U.S. Negative
Income Tax Experiments
• but Dauphin sample unique
Mincome(Winnipeg)/U.S. results
• “25 word summary”: “Few adverse effects
have been found . . . Those adverse effects
found, such as work response, are smaller
than would have been expected without
experimentation” (Hum and Simpson, Journal
of Labor Economics, 1993)
• no examples of “full blown” GAI but the
concept of income-test (targetted) benefit
programs is now widespread:
GAI has been gaining acceptance in
social policy programming
• programs like the Guaranteed Income
Supplement for seniors, the National Child
Benefit, and the GST credit are based on family
income and clawed back as income rises
• should think of Social Assistance programs in
terms of the level of the guarantee, G, but also
the nature of the claw back, t, in terms of the
incentives provided (e.g. work and the “welfare
wall” of high claw back rates)
Yukon House Motion Nov 27/07
• THAT this House urges the Yukon government to research
and develop a policy on a guaranteed annual income, as
recommended by the Royal Commission on the Status of
Women, the Macdonald Commission, the National Council
of Welfare, the Special Senate Committee on Poverty and
the federal working paper on social security, that would:
• (1) be a simple, non-taxable basic income available to all
• (2) be recoverable through the personal income tax system
for those earning over a certain amount;
• (3) eliminate or significantly reduce the social assistance
• (4) simplify administration and reduce administrative costs