Five years of the Working Income Tax Benefit

Quietly, without (much) fanfare, Stephen Harper's Conservative government has been gradually promoting a new model for income support programs: the Working Income Tax Benefit, or WITB.

On the face of it, WITB looks very similar to the Liberal government's signature program, Canada Child Tax Benefit (CCTB). Both WITB and CCTB provide cash support to low income households (in the case of CCTB, through the National Child Benefit Supplement). Taken together with other federal and provincial programs, they provide a poor working family with thousands of dollars a year in income support.

The two programs, WITB and CCTB, share a number of basic goals. Both aim to reduce poverty, and improve people's lives. Both aim to reduce dependence on provincial social assistance programs, by making it possible for people to survive on low wage or part-time work, supplemented WITB, CCTB and other refundable tax credits, without resorting to welfare. Finally, by making work pay better than welfare, both programs encourage labour force attachment.

Yet WITB differs from CCTB in a number of respects. The first is how it is framed and marketed. The stated goal of WITB is to provide "tax relief for eligible working low-income individuals". This makes little sense, as many of those eligible for WITB will pay no federal income tax. The WITB recipients who do pay federal income tax will find that WITB adds  to their marginal effective tax rate, because an extra dollar of earnings will reduce their WITB payment. But I guess everyone wants a tax break, even people who don't pay much by way of taxes.

Second, WITB is available to low income singles and childless couples, as well as families with children. This matters because unattached people under the age of 65 are the Canadian demographic with  the greatest risk of having low incomes. Whether one looks at the numbers before or after taxes and transfers, singles under 65 have a higher rate of low income than single parent families, people over 65, or any other group. The WITB isn't much, but it's a start. 

The third important feature of the WITB is that it is conditional on undertaking some paid employment, at some point in the calendar year. The WITB is structured as follows:

• People with no earned income receive no WITB. 
• For each dollar earned over and above $3000 per year, WITB increases by 25 cents (under the national program - BC, Alberta, Quebec and Nunavut have their own programs, with slightly different rules)
• Benefits continue to increase with income they until reach the maximum: a benefit of $1762 per year for couples or single parents, a benefit of $970 per year for singles (2012 benefit numbers). 
• Recipients continue to receive the maximum benefits until their net income reaches $11,011 for singles, $15,205 for couples or single parents (in 2012).
• Once this net income threshold is reached, benefits are phased out at a rate of 15 cents per dollar of income.

In a picture:

One thing worth taking some time to understand is the difference between the way that WITB encourages work, and the way National Child Benefit does so. NCB encourages work by shortening the welfare wall. Provincial social assistance programs provide financial support, often known as "welfare", to people with no assets, no income, no other means of getting by. As a rough approximation, these programs have 100 percent tax back rates. They reduce benefits received by a dollar for each dollar earned.

Before the National Child Benefit was introduced, a single parent in need might receive, say, $15,000 a year in social assistance - and have to earn at least $15,000 before work paid better than welfare. With NCB, that same person might receive $11,000 in social assistance and $5,000 in NCB. If he found part-time work paying $15,000 a year, he would still lose the social assistance payments, but he would be able to keep the NCB. Work now pays a little better than welfare.

WITB, like NCB, can be retained when a person moves from social assistance into paid employment. But WITB adds in an extra kick - the 25 percent add on to any earnings - to strengthen the incentives to take up paid employment. Indeed, because WITB is conditional on having some earnings, long-term social assistance recipients might not receive any WITB at all - unless they are prepared to undertake some paid employment.

A final difference between WITB and child benefits (or other refundable credits such as the GST credit) is timeliness. People can apply to receive up to 50 percent their WITB payments in advance. This makes WITB into a more effective anti-poverty measure: people can get benefits when they need them. By way of contrast, if a person loses her job in January, 2012, her child benefits will not adjust until June, 2013, considerably lessening the ability of child benefits to fight poverty and reduce reliance on social assistance.

WITB aims, in part to encourage other Canadians to enter the workforce (that "other" conveys a subtle message: you're hard working and deserve tax relief, other people need to be encouraged to work harder). But is it effective? In theory, the effects of WITB on work incentives are ambiguous - the 25 percent subsidy for people earning just over $3,000 a year encourages paid employment, but the phase out adds to effective marginal tax rates, discouraging work effort. 

What about empirical evidence? I haven't been able to find any. William Scarf and Lei Tang have a paper in Canadian Public Policy on the WITB. Like a lot of Canadian policy papers, it's a "pre-valuation"; an assessment, before the policy is actually implemented, of its probable impacts. They predict that a WITB-type policy would cause unemployment to fall by between 0.2 and 0.17 percentage points, and the average incomes of WITB recipients to rise by 3.9 to 8.6 percent.  Their model is interesting in that the the decline in unemployment is driven by efficiency wage type considerations. The idea is that WITB makes minimum wage employment more attractive (relative to the alternative of not having a job), so people exert more effort in minimum wage jobs, productivity increases, firms hire more minimum wage workers, and unemployment falls. I'm not sure if I buy their story - but I couldn't find any other ones out there.

Even though it is a small program, and it is hard to say anything concrete about its impacts, the WITB reveals a great deal about the workings of Canadian public policy.

First, there are those, including some who write regularly on this blog, who would like to see Canada introduce a negative income tax, or a guaranteed annual income. They would like to take the alphabet soup of programs - WITB, CCTB, NCBS, HST refund, GST refund, GIS... - and roll them all into one simple straightforward program. That will never happen, and the WITB shows why. Every government prefers taking credit for a new initiative to enriching a program associated with their predecessors. Governments want programs that reflect their own values and priorities. They tweak program design to favour particular voting blocs, or to subtly encourage people to make certain types of choices. Even if we could magically create a guaranteed annual income tomorrow, the next budget would contain some new measure, an "announcable" that the current government could claim credit for.

The WITB also reveals a great deal about the way policies are evaluated in Canada. Why are there so few published evaluations of the WITB? One reason is that programs like the WITB are hard to analyze (see Kevin Milligan's comments in the discussion below). A scientific experiment would randomly assign some people to receive WITB. Others, the "control group", would receive a "placebo" treatment. The difference between the two groups could be used to measure the impact of WITB. Yet because WITB was introduced in all provinces at the same time, there is no control group, no basis for comparisons. True, there are small differences in the structure of WITB across provinces - as noted above, BC, Alberta, Quebec and Nunavut have their own programs - and perhaps at some point some one will work out how to use those differences to assess the impact of WITB. But for now, we really can't tell if WITB has been successful in achieving its stated goal of encouraging work effort.

The other reason that there have been few evaluations of WITB is that no one has an incentive to carry one out. Academics are unlikely to try to assess the program. There have been so many studies of the US Earned Income Tax Credit that such programs are old news. A WITB evaluation would be hard to publish in a top journal - and top journal publications are what academics seek. Private consulting firms are profit-driven, and the money is in pre-valuations - policy makers will pay, prior to the implementation of a program, for an estimate of its probable effects. There is no incentive to find out what happens next.

What about governments? Perhaps there have been in-house evaluations of the WITB, but none have been made public (even the Department of Finance's tax expenditure analysis, usually so useful, has nothing). Since WITB was introduced in 2007, any positive effects of the program would have been swamped in the 2008 economic downturn, so I suspect it has been hard to find evidence of its success. This is another difference between the WITB and the National Child Benefit. The Liberal government published regular National Child Benefit Progress Reports. The present Conservative government has provided a series of reports on Canada's Economic Action Plan, but they are written in budget-speak, for example:

In 2012, if the WITB had not been introduced, a typical low-income single parent in Nova Scotia would have only kept about 28 cents of each additional dollar earned between $3,000 and $10,000, due to reduced benefits from federal and provincial income-tested programs and taxes. As a result of the enhanced WITB, the same family will keep about 53 cents of each additional dollar earned.

There is no mention of the increase in that single parent's tax rate when WITB is phased out...

Still, while WITB has value as a metaphor for Canadian public policy, it is vital not to lose sight of the bottom line: WITB provides support for some of the most vulnerable Canadians, and this is a good thing.

This post has been updated in response to comments made in the discussion below.

Comments

[Kevin Milligan]

Hi Frances,

You're definitely right about the academic incentives. But, there are other barriers to evaluating the WITB.

a) for its first few years, it was very small. It applied in a very narrow income range, based on income from 18 months earlier. The probability of still being in that income range when you were actually receiving it was not super high.

b) It is now a bit bigger, but data takes a while to come out. I (along with Mark Stabile) have had our eye on evaluating it and have now that data are available it is now becoming possible

c) the inter-provincial variation is fairly weak and so it makes inferences difficult.

d) we do have some Canadian evidence on a program with very similar incentives--the National Child Benefit. So, a 2nd evaluation of a very similar program in Canada--especially when the US evidence is so deep on this issue--is a hard pitch for a top journal. Mark and I wrote the paper on the NCB in 2004 and it was published in 2007 in JPubE: http://faculty.arts.ubc.ca/kmilligan/research/ncb.htm

[Kevin Milligan]

Oh--and for provinces that 'clawed back' social assistance dollar for dollar with the NCBS, you effectively didn't get it if you were on SA but you did if you left SA for a low-wage job.

So, I disagree with this sentence: "A key point to note is that WITB differs fundamentally from National Child Benefit Supplement, which has no work requirements."

I hadn't noticed the 'pre-pay' option for WITB. Has it always been there? I don't recall it being there when WITB was introduced.

[Frances Woolley]

Kevin, I wondered if you were working on something on WITB, but found nothing on your web page. Was that NCB paper harder to sell than you thought it would be? Personally, I like it, but you didn't get me as a referee...

A couple of points:

NCB doesn't have the kick into the labour market that WITB does, only the phase out, so its predicted impacts are different. Also WITB recipients (singles, childless couples, as well as parents) have different labour supply elasticities/reservation wages than NCB recipients (parents)

Because the WITB isn't available to students, to work out its impacts, you have to be able to identify students in your data - a lot of people in that WITB income range are students. This might be an issue with using something like the LAD to assess WITB, I don't know.

[Kevin Milligan]

"NCB doesn't have the kick into the labour market that WITB does"

Hi Frances, I don't understand this sentence.

The evidence suggested the intensive margin wage supplements or taxback don't matter much. It is the extensive margin (in/out) that matters. In that way, NCB (with SA clawback in place) provides a healthy kickup to the 'work' side of the 'work or welfare' decision. You get NCB if you are working, you don't if on SA.

[Frances Woolley]

Kevin, yup, I think that pre-pay was always available. It was one of the things I liked about the WITB.

On the sentence you disagree on - the "clawback" of NCBS from SA recipients means that people get off SA more quickly, so they face a 100% marginal effective tax rate for a shorter period of time. That's different from the WITB, which has that phase in explicitly designed to reduce the marginal effective tax rate by 25 percentage points right around the $3,000 earnings mark.

[Kevin Milligan]

I see--you are focusing on the marginal wage subsidy effect. Yes, you're right the NCB doesn't affect that.

But, in my take on the literature, most evidence finds little impact of the marginal incentives--most people face binding corner solutions: it is either work 0 hours or 20+ hours. The extensive margin is where much more of the action is, not moving people from working 20 to 25 hours.

This Saez paper is the framework I rely on for the intensive vs extensive margin responses. He suggests that at the low end, participation elasticities are more important.
http://qje.oxfordjournals.org/content/117/3/1039.abstract

ungated--http://elsa.berkeley.edu/~saez/botqje.pdf

and ps thanks for your kind words about the 2007 paper.

[Frances Woolley]

Kevin - "I don't understand this sentence."

As I say right at the beginning, "Finally, by making work pay better than welfare, both programs encourage labour force attachment." I agree that the NCB makes the welfare wall shorter, and so encourages labour force participation on the extensive margin. But it doesn't do as much on the intensive margin as WITB does. You can say that empirically this doesn't matter, and you may be right - but since no one has studied WITB it's hard to say this for sure.

[Kevin Milligan]

I was not a fan of the 2006-2009 WITB. With such a small amount over such a narrow income range (a range into which people bounce in and out at high rates), I thought it was not well designed. I think it would have been much better to consolidate the existing alphabet soup of programs than to add another. My thinking is aligned with Frances that the political return to something 'new' was an important consideration for why it was introduced.

Low income families with children receive something like 10-12 different child benefits. I spend a lot of time trying to figure them out and I find it very hard. I'm not sure how we expect busy single parents to wade through this morass. Simplicity is an undervalued virture here, I think.

The expansion of the benefit size and income ranges in 2010 does help make it more useful.

[Kevin Milligan]

Hi Frances,

thanks for reminding me about the advance payment option. Found the info here: http://www.cra-arc.gc.ca/bnfts/wtb/fq_qlfyng-eng.html#q10

Looks like a positive option thing. You have to fill in paperwork and send it in.

Are there any HRSDC administrative reports that mention how much take-up there would be for this option?

Anyone have anecdotal information about whether community centres, YWCAs, provincial SA offices etc put a push on for people to fill in these advance payment forms? I know that all of these kinds of institutions put a big push on for people to file taxes in general so they can get CCTB+NCBS etc. But I don't know if they also push people to fill in pre-payment option forms. In the absence of such a push, I think I would be surprised if people did go through the effort for the prepayment option--especially since mobility at those income ranges is fairly high.

[Kevin Milligan]

...but I agree with Frances that this prepayment option is a good one to have, rather than the long lags associated with CCTB/NCBS and the provincial work supplements administered by the CRA.

[Frances Woolley]

Kevin "Low income families with children receive something like 10-12 different child benefits"

I would really like to know how many of those HST and GST credit cheques that are mailed to university and college students end up going astray, getting lost, buried under a pile of dirty socks until it's too late to cash them, etc. Direct deposit is the best way to cope - I really would like to know how many people have signed up for that.

I think people just end up learning by trial and error how the system works. I remember reading in one of those evaluations of the Employment Insurance earnings exemption that people who are on EI for the first time are much more likely to take part-time employment than people who are on EI for the second time. People don't know the fine details, but they more or less figure out the system parameters.

[Frances Woolley]

Kevin - I'd be willing to bet that when people go down to the social assistance office they get told to fill out one of those WITB pre-payment forms....

[Kevin Milligan]

"Finally, by making work pay better than welfare, both programs encourage labour force attachment."

Yes indeed that sentence is there; I don't need to sound so critical--I know that you know your stuff on this!

Also--your point about the difference between single-no kids people and single parents having different elasiticities and circumstances is a very good one.

[Jacob AG]

I don't know much about the WITB, but it sounds like a Canadian version of the Earned Income Tax Credit (EITC).

The EITC has two opposing effects on the incentives facing a low-income worker: 1) the substitution effect and 2) the income effect. The substitution effect encourages the worker to work more by making the value of time spent working more valuable relative to time spent not working, while the income effect encourages the worker to work less by making it possible to maintain any given level of income while working less, because the value of each hour worked is higher.

So if the WITC is like the EITC (I think it is), then you should check out the empirical literature on the EITC. See here for example (PDF): http://www.cbo.gov/sites/default/files/cbofiles/attachments/10-25-2012-Recent_Research_on_Labor_Supply_Elasticities.pdf

The above literature review shows that the for low-income workers, the effective marginal tax rate at the lowest income levels is reduced significantly by the EITC, then plateaus as income rises, and then in the "phase-out" range they start to rise again. On net though the substitution seems to be typically larger than the income effect, so labor force participation increases. But the effects vary a lot by gender, marriage status, number of children, income, etc. Poor single mothers for example are much more likely to work with an EITC than without, for example; high-income married men, not so much.

[Frances Woolley]

Jacob - I appreciate you taking the time to respond, but this comment precisely illustrates why so people bother to go out and evaluate Canadian programs. There are dozens? hundreds? thousands? of papers evaluating the EITC. WITB is like the EITC, and we know what the EITC does, so we know what the WITB does.

Not obvious. I mean, it's obvious that high income married men won't be positively effected by the WITB (unless they have backwards-bending labour supply curves - someone's taxes have to rise to pay for these programs). It's fairly obvious how, in theory, these programs affect the budget constraint, and that they create income and substitution effects.

But it's not obvious that low income people will respond to the WITB in the same way that they respond to the EITC. WITB is not conditional on the # of kids, EITC, for the most part, is. WITB is smaller. It's nested within a very different set of social programs (universal health care, more generous social assistance, etc). Culturally the two countries are very different - we have had, until recently, a very high skill immigrant flow, and most of our immigrants are legal. The US has a much lower skill immigrant flow, and many US immigrants are illegal. This matters because immigrants make up a significant chunk those who would be affected by the WITB. The vast majority of the evaluations of EITC - the "stylized facts" about the program happened before the 2008 economic downturn.

[Mike Cornelia]

I actually really like this plan and think it could really help in eliminating back taxes. I have always thought this is one of our nation's biggest tax problems. A fair taxation program can help eliminate the amount of back taxes our country accrues each year. I really hope this helps.

[Ian Brodie]

I really enjoyed this post & the discussion to follow. Since I was around for the birth of WITB.

As I always say, it's the first time I've ever heard of a tax measure being named for the Finance Minister's home riding........

[Frances Woolley]

Ian - Excellent! I'm going to tweet that. B.t.w., which of the Ian Brodies on twitter are you - are you "logger, speaker, budding author and ever-hopeful Newcastle United fan" "Folklorist; bon vivant; cryogenecist; proud grandmother of quints" or "World class journalist with a splash is wrestling talent"?

[Bob Smith]

"As I always say, it's the first time I've ever heard of a tax measure being named for the Finance Minister's home riding........"

Brilliant!

[Allan]

Excellent tax credit for the working class family!

[Income Tax Calculation]

Great note that clearly highlights WITB and CCTB difference. Support for low income households is a must but should also be available when it is most needed and WITB seems to be a step in the right direction. Most attractive feature of WITB is that eligible people can get some portion of it in advance. This makes a big difference for families with children and i am sure it will be much appreciated.