How OECD governments generate tax revenues

I did a post last year documenting the choices OECD countries have made when it comes to tax rates - you might want to take a look at it before continuing. I'll wait here.

Okay, welcome back. Given the recent attention ([1],[2],[3]) given to the link between tax rates and tax revenues, this post is going to compare how much revenues these tax rate choices produce.

My data source is the OECD, and the measure is share of GDP, averaged over 2005-09. The revenue sources are broken down into these components:

• Goods and services. This includes retail sales taxes and VAT/GST/HST revenues
• Corporate income taxes (CIT)
• Personal income taxes (PIT)
• Social security contributions (SSC), broken down into employer/employee components. 'Social security' is what the OECD calls the series; but it's better to interpret it as the tax wedge between what the employer pays and what the employee receives.
• Other

 

Here are the numbers:

	Goods and services	Corporate income	Personal income	Social security	Other	Total
Australia	7.81	5.93	10.69	0	3.87	28.31
Austria	11.76	2.19	9.47	12.58	6.17	42.17
Belgium	11.05	3.25	12.43	12.56	4.64	43.93
Canada	7.94	3.53	11.87	4.71	4.73	32.77
Czech Republic	11.43	4.44	4.08	13.58	2.96	36.49
Denmark	16.00	3.56	25.33	1.02	3.21	49.10
Finland	13.31	3.22	13.26	11.28	2.16	43.24
France	10.89	2.56	7.61	15.15	7.40	43.61
Germany	10.48	1.88	8.94	12.54	2.24	36.08
Greece	11.25	2.68	4.86	9.10	3.41	31.29
Hungary	15.03	2.43	7.24	12.25	2.04	38.98
Ireland	10.96	3.16	8.31	4.65	2.88	29.96
Italy	10.80	3.39	11.12	11.34	6.00	42.64
Japan	5.17	4.04	5.35	9.44	3.79	27.79
Korea	8.24	3.86	3.81	4.63	4.97	25.52
Luxembourg	10.21	5.33	7.46	9.09	4.35	36.44
Netherlands	12.00	3.10	7.62	10.90	5.10	38.72
New Zealand	11.49	4.93	14.20	0	3.91	34.53
Norway	11.91	11.51	9.60	8.62	1.73	43.37
Poland	12.76	2.53	4.76	9.44	4.07	33.55
Portugal	12.99	3.15	5.50	8.27	1.83	31.73
Slovak Republic	11.28	2.85	2.57	9.51	3.53	29.73
Spain	8.94	3.53	6.88	10.59	4.77	34.71
Sweden	12.88	3.41	14.57	12.01	4.69	47.55
Switzerland	6.61	3.04	9.50	6.35	3.69	29.19
Turkey	11.53	1.70	3.90	3.78	2.44	24.34
United Kingdom	10.43	3.40	10.60	6.51	4.71	35.64
United States	4.66	2.64	9.66	6.20	3.50	26.67
OECD average	11.03	3.49	8.87	8.36	3.00	34.75

 

I've been trying to think of a good graphical representation for the data in this table (I even considered bubble graphs), but I can't. I decided on a sequence of bar graphs, where the ordering is done according to how important the various sources are.

The first graph sorts the countries by the size of all tax revenues as a share of GDP. These graphs are all pretty busy; clicking on them will produce a larger version:

 

Canada shows up as a relatively low-tax country, with the Nordic countries dominating the top of the list.

In this post - and expecially this graph, we saw that even though corporate rates vary greatly across time and across countries, there is remarkably little variation in CIT revenues as a share of GDP.

 

As noted in the other post, Norway is the big outlier here, even though its CIT rate is in the middle of the OECD pack - oil revenues are the explanation. CIT revenue shares are only weakly related to total spending; Norway aside, the CIT shares generated by the big-government countries are around the OECD average. Note also that Ireland - which has an ultra-low CIT rate of 12.5% - generates CIT revenues comparable to countries that apply much higher rates. Even in 2010, Irish CIT revenues were greater than in Germany as a share of GDP.

Here are the same data again, ranked by revenues from taxes on goods and services:

Canada is well down the list here. To provide some perspective, each GST point generates roughly half a per cent of GDP in revenues, so it would take the equivalent of a six percentage point increase in the GST to reach the OECD average.

Here are the data sorted according to personal income tax revenues:

Canada ranks above the OECD average in personal income tax revenues. Oddly enough, so does the US. Denmark looks to be a huge outlier here, and I'll get back to it later.

Sorting by the tax wedge:

Australia and New Zealand have no tax wedge revenues, and Canada's is below the OECD average. These taxes are easy to overlook, especially the employer payroll taxes that workers don't see. But since the incidence of payroll taxes falls mainly on workers (I had a really good link to a HDRC explainer by Bev Dahlby making the point, but it has gone dead. [Update: Kyle Hall has tracked it down in the comments - it's here.]), it is perhaps better to view Social Security taxes as a form of income tax. Of course, it should also be noted that these measures are invariably less progressive than income taxes. For example, contributions in Canada are capped, so the average rate paid by high earners declines with income.

When you do this, Denmark's relatively high PIT revenues don't lok so extreme, because its tax wedge is so small:

Canada's combination of relatively high PIT and a relatively small tax wedge puts it just under the OECD average when the two are combined. Social security taxes also explain how France can rank so highly in the total tax rankings with below-average personal income tax revenues.

The revenues generated by taxes on corporate income and on labour income put Canada around the OECD average; it is the relatively small amount of revenues from taxes on goods and services that put Canada below the average.

Comments

[Jacques René Giguère]

If ordinary taxpayers compute in aftertax income ( and most do even if they don't even make the difference between taxes and their dental insurance plan...), then any form of taxation causes wedges...

[Stephen Gordon]

Which is why it's probably better to think of them as a form of PIT.

[Frances Woolley]

Stephen: "But since the incidence of payroll taxes falls mainly on workers "

Not obviously. For minimum wage workers, the economic incidence of payroll taxes will be the same as the statutory incidence - because an employer can't reduce wages below the minimum, they must absorb the payroll tax.

"Canada ranks above the OECD average in personal income tax revenues. Oddly enough, so does the US."

That must be a sarcastic "oddly enough"?

[Frances Woolley]

My point being that Canada's tax structure is heavily influenced by the US tax structure - we have relatively high PIT and relatively low taxes on goods and services in part because they do - it's far from obvious that European levels of VAT would work in this country given the ease of cross-border shopping.

It's great to get this data out there, Stephen, and jump start an informed discussion.

[Stephen Gordon]

Yeah, the 'oddly enough' bit was a reflection of my surprise. I was expecting the US to be below average one all these measures.

And the point about minimum wage is well-taken. OTOH, something less than 5% of workers are paid the minimum wage, so to a first approximation, it seems reasonable to just add payroll taxes to the PIT.

(That number came from this ancient post, which links to a StatsCan document which is no longer there. This really is getting to be a problem.)

[Frances Woolley]

Stephen -Was it one of these?

a 2004 Perspectives on Income and Wealth article by D. Sussman
a 2008 Library of Parliament survey on minimum wages.

The % of employees earning minimum wage is somewhat higher now than it was then due to a combination of the economic downturn and the increases to the min wage around 2008/9 in most parts of the country, but I can't find any numbers suggesting that it's radically higher.

[Frances Woolley]

I'm wondering if perhaps we should start just uploading .pdf files rather than linking to government web sites - we can do so without violating copyright - that way there's less risk of links disappearing.

[blink]

Do the data reflect combine taxes at all levels of government, or only by the national government? I cannot tell from the description alone.

[Stephen Gordon]

Frances: Yes, IIRC, it was a Perspectives on Income and Wealth article.

blink: Data are for all levels of government

[Determinant]

What's "Employer and Employee Social Security Contributions" in Canada? CPP, EI or both?

Let's not make category errors about CPP, CPP is not a tax. That Dental Insurance quote was also a category error. According to this view, any plan charging a risk-based premium with pooling among participants is a tax. That is conceptually wrong, life insurance is not a tax, nor is disability insurance.

CPP is a pension which is a form of longevity insurance, a life annuity.

[Bob Smith]

Stephen,

The relatively heavy reliance on PIT isn't all that surprising, given that it is a (relatively) small government jurisdiction. If you're going to be spending 54% of your GDP you need more efficient revenue raising tools than PIT (i.e., VATs and payroll taxes (or taxes on other immobile factors of production). On the other hand, if government spending is 35%, you can survive with relatively inefficient revenue raising tools (although since US spending is now well above that, that's a problem).

One of the weird results of the US reliance on inefficient taxes(apart from keeping us government spending relatively low), is that its after-tax (and transfer) income distribution is far closer to that of some European countries than market income figures would suggest (and then some European would like to believe) since the US relied so heavily on sharply progressive PIT, while the Europeans rely on far less progressive (if not regressive) VAT or SSC. A few years ago I read a fascinating criticism of the French welfare state which suggested that (a) France was less equal on an after-tax basis than the US and (b) whereas the US tax and transfer system mitigated the inequalities of the market, the French system made them worse (making France, literally, a reverse Robin Hood state).

There's a political dimension to this too. The rabid opposition to a US VAT is driven in part by a concern (probably quite justified) that if it was easy for the US government to finance more spending, they would. Of course, the lack of revenue hasn't really kept the us from spending.

[Bob Smith]

Determinant, the fact that we may get benefits from CPP doesn't make it less of a tax. If you think its not a tax, try not paying it and see how that turns out.

[Nick Rowe]

Bob: a better analogy(?) might be to say that CPP is a "forced purchase". The more income we earn the more we are forced to purchase. If we would want to purchase CPP anyway, that creates no disincentive effects. If we didn't want the purchase, because we subjectively evaluated the benefits as less than the cost, it acts like a tax in having disincentive effects. But if we really wanted the purchase, because we subjectively evaluated the benefits as greater than the costs, it acts like a subsidy in having incentive effects.

What I wonder is: how does the total complexity and lack of transparency of CPP (nobody except a couple of public finance economists actually understand the individual's marginal costs and benefits of CPP) affect people's calculations of their subjective costs and benefits?

[Mike Moffatt]

"Bob: a better analogy(?) might be to say that CPP is a "forced purchase"."

EI would certainly fit this description as well.

[Stephen Gordon]

Bob - yes, that's Peter Lindert's point in "Growing Public": if you're going to have a large govt, you can't afford the luxury of a dumb tax system. I wasn't surprised that US PIT was large relative to the entire US tax take, but that it was large even relative to the OECD average.

[Jacques René Giguère]

Determinant: I am the union representative on health insurance and retirement. My members are college professors. Most of them can't make the difference between taxes and insurance premium, let alone QPP and RREGOP ( pension) contributions.
Out of the crooked timber,,,

[Joe]

Stephen,

Do you have any data presenting what portions of tax revenue are obtained from each quintile along the distribution curve, including the top, say, 1%. Such info would be very useful to compare amongs countries.

Even more useful would be to compare this data across time.

[Nick Rowe]

Jacques Rene: I'm ahead of most, in that I understand that there is *some sort* of connection between my CPP contributions and my CPP benefits. But I have no idea what that connection is. I figure it is probably weakly monotonic, but I'm not even really sure about that, if I keep on working too long, and if I'm already maxed out, and get to collect CPP for a shorter span.

My guess is that Frances maybe knows, but I doubt many others do.

A regression with errors in variables biases the estimeted beta towards zero, IIRC. Which makes CPP more like a tax.

[Bob Smith]

Nick,

I take your point, but at least with CPP, part of the rationale for making it mandatory (apart from the intra-generational transfer) is that people otherwise wouldn't save adequately for their retirement - i.e., they wouldn't voluntarily save that money (even though they should). So there's a case where mandatory CPP may well be welfare improving (if you believe that people don't adequately save for retirement), but in which it's regarded by employees/employers as a tax.

Also, how far do we stretch the "forced purchase" analogy? I'm forced to purchase health care, education, roads through, amongst other things, the GST and PIT system. Those are all things which, if not provided by the government, I'd probably consume myself. Do those cease to be taxes?

Mike,

In theory, EI might fit within the "forced purchase" analogy, but in practice? Let's face it, it's a system for transfering money between regions and industries. Moreover, for many, it's a "forced purchase" that the "purchasers" will not be entitled to consume (i.e., a good chunk of the unemployed in Ontario). Who would ever purchase such an "insurance" policy?

[Bob Smith]

Don't get me wrong, CPP and EI may be taxes, but to the extent they're borne by employees, they're probably relatively efficient taxes if you believe (as I think most do) that the labour supply elasticity for low/middle income individuals is relatively low. Regressive? Sure, but in the greater scheme of things, relatively efficient - hence the European predilection for payroll taxes. Plus, politically, politicians much prefer to levy "contributions" than "taxes" - it's all about branding.

[Nick Rowe]

Bob: "So there's a case where mandatory CPP may well be welfare improving (if you believe that people don't adequately save for retirement), but in which it's regarded by employees/employers as a tax."

Good point. Let me restate it: if it didn't act like a tax, it wouldn't have needed to be mandatory.

"Also, how far do we stretch the "forced purchase" analogy?"

Hmmm. Only as far as we can stretch the link between: an individual's choices, an individual's costs, and that same individual's perceived (by him) benefits. It doesn't work for education, healthcare, and roads, since what I get is unrelated to what I pay. I think the link exists (but is weak) for EI.

[Stephen Gordon]

Joe: No. But there are some data on how the redistributional effects of the various tax systems; see here.

[Frances Woolley]

Nick: "What I wonder is: how does the total complexity and lack of transparency of CPP (nobody except a couple of public finance economists actually understand the individual's marginal costs and benefits of CPP) affect people's calculations of their subjective costs and benefits?"

This is such an important point, and something public finance economists haven't in the past paid nearly enough attention to. Think, e.g., of the entire negative income tax movement. "Let's scrap the entire existing tax/benefit structure and replace it with a simple formula, taxes = -a+bY where a=value of refundable tax credit, b=marginal tax rate." No, we do *not* want people to be able to work out what a and b are. To counteract the negative incentive effects that would inevitably accompany an increase in transparency, any such NIT scheme would have to be accompanied by a significant reduction in the value of b relative to what we have right now at the low income end, accompanied by a corresponding reduction in the value of a.

It's sort of the economic equivalent of putting cookies in an opaque container on the bottom shelf of the cupboard, as opposed to a clear glass container on the counter. Sure, you still know they're there, but you'd rather not be reminded of them.

I suspect this might be one of these economists-see-through-framing-and-can't-understand-that-other-people-don't things.

[Nick Rowe]

Frances: and the counterpart, if we don't want CPP contributions to have disincentive effects like a tax, we want to put the cookies in a clear jar and tell people exactly how many extra cookies they get if they earn an extra dollar. CPP contributions are clear (or, rather, pay net of CPP etc contributions is very clear). We don't see the benefits.

Does anyone tell their kids: "If you do the washing up, I may give you mumble mumble things in opaque jar on the bottom shelf"? You wave the cookie in front of their little noses.

[J.V. Dubois]

Great post, however I start to think that all these comparisons are totally worthless. Here is why

1) Social security is a mess. I will give you just two examples

- Outright taxes masked as social security contributions. For instance, in Slovakia where I live, we have a so called "health insurance" that really is just a tax. Every employee pays 14% (10% employee + 4% employer) of his gross salary in this "health insurance" up to a cap (. Then everybody in the country has access to free healthcare - no matter how much did he pay as par of his "health insurance". So in reality this is basically a special personal income tax for emploeyed people.

- Partial taxes. In Slovakia government uses SSC to cross-subside other parts of its social system. There are mild ones, such as that government pays "maternity allowance" - that is basically 50% of the previous income of the mother based on a government disability insurance. So if you are single male you know that your contributions are also used to pay allowance to mothers even if you know that you never want to have a child. But this pattern is repeating and over one-third of payments gathered as part of the SSC are not used on for the purposes they were created for.

2) Sometimes taxes are meaningless. Also let me have an example

- Imagine that you live in a city where parking is free. Now the city hall wants ti increase the revenue and they can choose from several possibilities

a) Create a congestion tax as it was introduced in London. You pay a fee to be admitted into a city and then you can park for free (or cheaper) - because there is less cars in the city.
b) Create City-Owned parking company that will gather parking fees from people willing to park their car in the city.
c) Rent the parking lots to a private company for 99 years for a lots of cash.

Now only choice a) shows in the OECD tax paper as part of the "Other taxes" column. But it does not say anything about how people live in the country, if the revenue is used well or if it is wasted, if taxes are general (used for transfers) of if they are akin to fees for particular government services.

[Frances Woolley]

Nick: " if we don't want CPP contributions to have disincentive effects like a tax, we want to put the cookies in a clear jar and tell people exactly how many extra cookies they get if they earn an extra dollar."

In this case it's a special jar made of distorted glass, that makes the cookies seem bigger than they actually are. By which I mean for every $4.95 the employee puts in, the employer also puts in $4.95. Because the employee only sees their $4.95, it looks like a better deal than it actually is. (How's the marking going?)

J.V. Dubois - so true. I think this is one problem with public discussions around taxation - the more you know about taxes, the more you find yourself shrugging your shoulders and saying "it's complicated". Hence the people who shout the loudest are sometimes the people who know the least.

[Stephen Gordon]

one of these economists-see-through-framing-and-can't-understand-that-other-people-don't things

QOTD

[Patrick]

"Does anyone tell their kids: "If you do the washing up, I may give you mumble mumble things in opaque jar on the bottom shelf"? You wave the cookie in front of their little noses."

If only that worked! I have to promise to go outside and play soccer with my kid. I want him to do stuff that's good for his soul and increases my leisure, not the other way around.

[Bob Smith]

Joe,

I think this is what you're looking for:

"Growing Unequal? Income Distribution and Poverty in OECD Countries," Organization for Economic Cooperation and Development, 2008. p. 112. http://dx.doi.org/10.1787/422013187855.

Note, in reading it, apparently income taxes refer to both personal and social insurance taxes.

What's intriguing is that the US appears to impose the heaviest taxes on high income earners in the OECD - measures as the proportion of taxes born by the top decile relative to their share of market income. In contrast, the Scandinavian states generally impose much smaller tax burdens on high income earners (indeed, perversely, in Norway, the top 10% pay less than their "share" of income taxes, relative to their share of market income). This makes sense if you consider that countries like Sweden generally have relatively low tax rates on capital income (which, you'd expect, would generally earned by higher income earners).

You wonder to what extent adding VAT would affect that analysis. Obviously, high income earners would pay more VAT than low-income earners. On the other hand, low-income earners are likely to pay VAT on non-market income (social assistance, refundable tax credits, etc.). So it's not clear that the ratio of VAT paid by high income earners to market income would be greater than 1.

Anyhow, very interesting.

[Kathleen]

Sorry this is off topic but: I have always found it puzzling that CPP is considered to be non-transparent. We know what our contributions are (printed on our pay stub), we know what the employer contribuion is, we know what the amount of the benefit. It works out to a roughly 3% internal rate of return on a highly safe investment that is indexed to inflation (reference: Jamie Golombek). The only drawback that I can see (and it is a big one) is that the government can decide at any moment to cut the benefit.

[Bob Smith]

Quick, Kathleen, tell me how much is in your CPP "account"? If you retire in 2033, what will your monthly payments be? Don't worry if you can't answer that, I doubt anyone here could.

[Kathleen]

@ Bob Smith,

If what you say is correct then it is simply that people are generally not that interested. I, like many Canadians, have access to the e-services website run by the government of canada. On that website there is a full record on every single contribution that I have made. There is also, what I assume to be, an accruate figure given for my benefit at age 65 and 60 and 70. I go to that website every year and note how much my benefit has increased. With that information I was able to calculate an internal rate of return of around 3%, a figure that was then confirmed by an article published by Jamie Golombek. Because I am a small business owner, I actually have a choice of whether or not to contribute to CPP and therefore I wanted to know what my return was. CPP is government backed and tied to inflation so I regard a 3% return to be pretty good.

[Bob Smith]

And how many Canadians have established e-service accounts? That the information may be available really isn't the question, the question is whether it is readily accessible?

As an aside, how annoying is it that the feds can't link that account to your CRA account, so you only need one login code and password to access all of your personal information.

[Frances Woolley]

Kathleen: " It works out to a roughly 3% internal rate of return on a highly safe investment that is indexed to inflation (reference: Jamie Golombek)."

That number varies enormously according to personal circumstances.

[Andrew F]

CPP is not quite government-backed. Government of Canada does not guarantee benefits, and any shortfalls in the CPP fund would need to be made up in increased contributions or reduced benefits. It is pretty safe because the Government of Canada guarantees that Canadians will be contributing to the fund (or else).

[Kyle Hall]

Stephen: The HRSDC paper is archived here.

[Stephen Gordon]

Fabulous! Many thanks! Update has been added.

[Determinant]

@Bob Smith:

"Quick, Kathleen, tell me how much is in your CPP "account"? If you retire in 2033, what will your monthly payments be? Don't worry if you can't answer that, I doubt anyone here could. "

Irrelevant. CPP is a defined-benefit pension plan. It pays out based on your last earnings and longevity. As such it relies on the insurance principle of aggregation and pooling, that the longevity of an individual is unknown and highly uncertain, but collectively our longevity is far more regular and predictable, thus enabling risk-sharing and pooling.

Private DB plans don't have individual present value accounts either; if you are discharged from one and take a Commuted Value settlement is is expressly a "best guess" and can fall short.

@Bob Smith and Nick Rowe:

Good point. Let me restate it: if it didn't act like a tax, it wouldn't have needed to be mandatory.

Sorry Nick, you lose. Quick, can you opt out of the Carleton pension? No, of course not, contributions are mandatory because it is a Defined-benefit plan with the opportunity for Additional Voluntary Contributions from you which are cash payments. The same goes for your supplemental health insurance and your disability insurance. All of these are defined benefit plans without individual present-value accounts but with individual defined benefit entitlements. All are mandatory to enable the pooling to work. None are taxes.

You Nick, France and Stephen all pay into these plans (Laval has similar plans) and none are taxes. Category error again. Perhaps Frances is right in that economists have a framing problem.

By way of anecdotal comparison, my church hosted the United Church of Canada local Presbytery meeting six months ago. As one minister mentioned, in the United Church you have to be in "Essential Agreement" (Essential is very, very loosely defined) with the Twenty Articles of Faith, our official doctrine document, but you MUST participate in the DB pension plan.

@Jacques:

Determinant: I am the union representative on health insurance and retirement. My members are college professors. Most of them can't make the difference between taxes and insurance premium, let alone QPP and RREGOP ( pension) contributions.
Out of the crooked timber,,,

Very much so, very very much so. That's why I am posting the way I am. Economists seem to have a knee-jerk reaction against collective pooling arrangements which have defined benefits and variable premiums, notwithstanding that these are significant private-sector economic activities in Canada, see Manulife, Industrial-Alliance & friends.

[Bob Smith]

"Sorry Nick, you lose. Quick, can you opt out of the Carleton pension?"

Sure he can, he can work for UofT. Or he and his like-minded colleagues can lobby his union to give up the pension plan in exchange for higher wages (he might be crazy to do that, or maybe not, if he thinks that Carleton will ultimately run the pension plan and itself into the ground).

More to the point, Nick's point (and mine) is more subtle than that. People will treat something as a tax if (i) there isn't a clear link between what they pay and what they get and (ii) if they would otherwise pay for what they get themselves.

In most cases, I'd put it to you that employee benefits aren't likely to be perceived as a "tax", in part, because most employees don't perceive themselves as "paying" for them (because they don't see themselves as having given up higher cash earnings to "get" that benefit - it's an interesting framing problem, if you presented workers with pay package X and employer contribution to pension plan Y, they'd probably be more responsive than if you gave them pay package X+Y and said they had to make mandatory contibutions Y to the pension plan, even though it's the exact same offer). Furthemore, those are often benefits that people would otherwise pay from themselves and to the extent there is a perceived cost it's tied to a clear well-defined benefit - the health club you use, the insurance forms you submit, the pension statement you get every 6 months or year.

Now, none of that is self-evidently true in the case of either CPP or EI.

First, the cost is clear and apparent, it's right there on your paystub every month. Granted, that's only part of the cost, but it's still a cost.

Second, the "benefits" of CPP and EI are ones that people often wouldn't pay for themselves. In the case of EI, there is hefty redistributive element to it, so that people who are statistically likely (if not statistically certain i.e., a good chunk of the working population either can't collect EI benefits - anyone working part time or contract jobs in places like Ontario or Alberta- or are highly unlikely to - tenured professors, anyone?) not to collect benefits, it's a loser investment, one they'd never make if they weren't compelled to. In the case of CPP, that's in part because part of the rationale for CPP is a recognition that, absent government compulsion, people simply won't save enough to fund their retirement (and I believe there's a large body of research on this point, but Nick or Frances may be better suited to discuss it). In that case, even if you believe that what you pay is what you get, that isn't the perception of the payer - they're giving up hard cash today, for a benefit in 30 years that they don't value today. Compounding that is a widely held view (and one promoted, somewhat inproperly, by investment advisors) that, in practice, in 30 years what you're going to get won't amount to a hill of beans.

Third, and finally, the link between the benefit and the cost is not at all clear. Most people have no clue what they would collect on EI if they were to lose their job (although this works both ways, some people think that EI pays far more than it does, or mistakenly believe that they're eligible to collect it), most people probably have no clue what their CPP entitlement will be when they retire (assuming that they believe that they'll get anything).

That isn't to say you couldn't change that perception. The government could send out CPP statements every 6 months showing Canadians how well their "investment" is doing (there's a reason investment advisors do that - apart from regulatory reporting - it's to encourage people to keep saving). They could shift the payment obligation entirely to employers. They could offer Canadians choice in how the government invests "their" money (sort of like what the SPP does) to provide a clear link between the costs and the benefits. But since they don't...

Put it this way, if CPP and EI are no different from defined benefit plans, supplemental health insurance, and disability insurance, isn't it odd that they have to be collected with the threats of imprisonment for non-compliance?

[Bob Smith]

Or consider another example, is a carbon tax a tax? In theory, with an optimal carbon tax the benefits arising from imposing the tax should be equal to the cost. And with a sub-optimal carbon tax, the benefits may well outweigh the costs by several orders of magnitude. But try telling that to Joe Q. Voter when he asks why his price of gas has gone up.

[Patrick]

Bob: To compound the OT ... could it be that linking them precluded by legislation?

[Determinant]

Again, category error Bob. Nick can't stay at Carleton and opt out of those benefits. That's my point, Carleton deducts those premiums from his paycheque. The fact that he could opt out by moving to UofT is akin to saying that to opt out of CPP you have to leave the country. Your point about the union is akin to arguing that tomorrow we could abolish EI and CPP by getting them repealed by Parliament.

Or is it a premium when levied by the union but a tax when levied by the government? Is tax an irregular verb, I tax, he benefits, they pay premiums, in the Yes, Minister sense then?

More to the point, Nick's point (and mine) is more subtle than that. People will treat something as a tax if (i) there isn't a clear link between what they pay and what they get and (ii) if they would otherwise pay for what they get themselves.

Point 1 is rebutted by the entire insurance industry. A risk-based and therefore unclear cost on an individual level between premiums and benefits is how insurance works, the solution is that it works in aggregate.

Point 2 about direct costs is rebutted by disability insurance premiums. Nick is a bit of an exception as Carleton has a specific form of plan, but most DI plans in Canada deduct a set premium from employee's paystub from after-tax dollars. This is because the benefits then received from the policy are tax free, so the total cost is lower. It's used all over the country. As a lawyer you may not have seen it but lawyers are a prime market for individually-purchased disability insurance policies.

Your points about CPP and EI benefit information are just wrong and I am soirry but I have no other way to put it. DB plans do not entitle an individual to a set, individual account. It doesn't exist, the money is all pooled. What you do get are benefit formulas. These are easily available from government websites, the formula isn't complicated and it's right there. DI benefits are often integrated with EI so the EI formula is right there in employee benefit booklets.

Standard Life has an excellent retirement planning calculator that I was introduced to when I joined a EC pension plan at a former employee. It explicitly calculates OAS and CPP entitlements. It's easy. It's widespread, it's handed out by sales staff. It counters your arguments about CPP completely.

CPP and EI are not taxes because contributions create benefit entitlements. Sure, claims aren't always easy but DI claims aren't easy either. It's possible, even probable that you'll pay a lifetime of premiums and never make a claim.

Microeconomics seems to have great conceptual problems with defined-benefit type plans the collective action assumptions they are based on.

"Put it this way, if CPP and EI are no different from defined benefit plans, supplemental health insurance, and disability insurance, isn't it odd that they have to be collected with the threats of imprisonment for non-compliance? "

Isn't it odd that DI premiums and DB pension plan contributions have to collected or your employment can be terminated? That seems like force to me.

[J.V. Dubois]

Determinant: good posts. I am surprised that so many people do think about taxes in such a narrow sense - as cash payment that needs to go from your wallet to a tax office. There is multitude of government regulations that can easily be considered as taxes. I do not know about Canada, but in my country there is a law about retailers being obliged to follow a government policy for return of defect goods. There is government agency tasked to follow on consumer complaints if retailers do not comply with this law, it can be enforced in the court of justice. This is basically government-forced form of insurance quite similar with some forms of social security. It surely generates costs for producers and retailers, yet I have never seen any tax analysis that would try to calculate costs of such regulation as form of taxation.

But I understand why thinking about taxes in this cost-benefit way generates so much resistance. You would be forced to calculate value of essential government services - such as being protected by army and police, having your contracts enforced by judiciary system - and substract it from your taxes. It would require analysts to take into account not only the level of taxation, but also the amount of government goods services people receive in return (and count only the difference). It becomes much more complex if these goods and services do not have market alternative, or if its value is hard to measure (as is with insurance policies).

Like Frances said - it's complicated. And I think it is so much complex an issue that it may be impossible to discuss it in a meaningful way - not with this kind of blunt analysis.

[Determinant]

Thanks J.V.

I've run into this thinking before, private/public cost-benefits. I am a diabetic and thus generate significant amounts of medical waste (used needles). Drug stores often provide sharps containers to dispose of this waste. But don't ever try to bring one shop's container to another shop, even if you are now a customer of that new shop because you moved and a different chain has all the stores in your area. They won't touch another shop's container and make clear that the program is to satisfy customers, not deal with a kind of waste that is unusual for households to have.

Even local household hazardous waste depots don't like to touch biohazardous waste, they are geared to taking old paints, oils and household chemicals.

I am of the view that if you deal in hazardous waste (needles) you should take the waste back, especially because it is a special class that requires specific handling. But no, shops ignore that cost. I've often wanted a provincial law saying that you can return needles to drug stores in sharps containers but of course that is coercion and restrictoin of the free market. Except the health of the garbagmen and the neighbours requires it.

[westslope]

Terrific set of posts Stephan. I share some of your surprise with the results.

Frankly, I find it "odd" that the US still has a dual mandate for monetary policy. American or American-based economists create the theory, and other countries import the technology. A lot of good work has been done on tax incidence in the USA. Same with property rights in general and common property in specific.

In practice the USA remains under the throes of an open-access ideology. Perhaps it is simply a legacy of colonial times? Canadians socialists and 'progressives' are also open-access ideologues but frankly I don't think they have simply thought it through.