Kevin Milligan

[PDF version of this letter]

My conflict disclosure is here.

This is the text of a letter written by me to Liberal policy advisor Tyler Meredith.

Re: Basic Personal Amount

Tyler Meredith
Liberal Party of Canada
tmeredith@liberal.ca

September 21, 2019

Dear Mr. Meredith,

I am writing about your request to provide information about the proposal by the Liberal Party of Canada to expand the Basic Personal Amount. Under your proposal, the Basic Personal Amount will increase between the 2020 and 2023 tax years to $15,000. In this letter, I am happy to provide the information requested about your proposal.

As a professor at a public university, engaging with public policy is part of my job. In this role over the past four years, I have provided advice to a Member of Parliament (CPC) on his Private Members Bill, the Government of British Columbia (NDP) on tax policy and poverty initiatives, and to the Government of Canada (LPC) on several matters of public policy.

In this same spirit, I hope the analysis I provide will better inform the voting public as it considers policy options during the current election campaign. I invite you to quote this letter in your public communications, and I will make this letter available to the public on my website as well.

During this election cycle, the Parliamentary Budget Office is providing cost estimates for policy proposals. I fully support this development, but the PBO is limited to providing total cost estimates and is not providing estimates of the impacts on different populations of Canadians. I therefore view the information I am providing with this letter to be complementary to the fine work of the Parliamentary Budget Office.

My efforts here should not be interpreted as an endorsement of your party or its proposal by me, my employer, or any institution to which I am affiliated.

The main finding of my analysis is that the proposal to expand the Basic Personal Amount will remove about 690,000 Canadians from paying federal income taxes by 2023, and lift 38,000 Canadians above the poverty line (MBM).

I also show that Canadians from all family income groups benefit from this tax cut and provide a comparison to the recent tax cut proposed by the Conservative Party of Canada.

These estimates are offered with a moderate level of uncertainty. The uncertainty arises from assumptions about the future incomes of Canadian out to 2023. More detail on the findings and the methodology are in the attached memo.

It has been my pleasure to assist you with this request. I sincerely hope it helps to inform the public debate during this election campaign.

Yours truly,

Kevin Milligan
Professor of Economics
UBC Vancouver School of Economics

Details and Methodology

The proposal is to increase the Basic Personal Amount and the Spouse or Common-Law Partner Amount according to the schedule set out below. It is phased out over the 2^nd highest tax bracket according to individual income.

The ‘Default BPA’ is the default path from 2020-2023. The ‘proposed BPA’ is the proposed schedule you provided me. The BPA gives rise to a 15% credit and is non-refundable. These are not changed under your proposal.

My analysis makes use of two sources of information:

• The Canadian Income Survey microdata.
• The Canadian Tax and Credit Simulator (CTaCS).

I project forward to future years the pattern of incomes using the Canadian Income Survey. Then, these incomes are processed through the Canadian Tax and Credit Simulator to generate an estimated tax liability, including all federal and provincial taxes and benefits. This process is repeated first under the status quo and then with the proposed policy in place. The change in taxes is used to assess the size of the tax cut and whether any federal income tax is owing. After-tax income is then used to assess each individual and family’s position relative to the poverty line.

The poverty concept used for this analysis is the Market Basket Measure, which has recently been adopted as the official poverty indicator for Canada.

The estimates do not account for behavioural response.

The results are below in Table 1 and Table 2.