« Unemployment, aggregate demand, and search/matching | Main | The Real Minimum Wage »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

One story I vaguely remember hearing, from one of my economics profs as an undergrad, I think: a union that gets its members a pay raise won't be remembered, because who can remember getting an extra dollar 50 years ago, and try to figure out how much of that extra dollar they are still getting now? But health insurance is something permanent. Like getting your name on a university building, rather than giving the university a million in cash. We do not observe the counterfactual.

Ironically, I can't remember the name of the prof who suggested this theory.

Nick - there's something to that, but I think it's slightly different from what the unremembered prof had in mind. Health benefits are an inflation-protected form of compensation, that is, if the price level doubles the cost/value of employer funded health benefits doubles. Monetary payments, however, are vulnerable to inflation.

This would suggest that health benefits would be preferred form of compensation in inflationary times, or any other time when there was downwards pressure on real wages.

Course the permanence of health benefits isn't such a great thing if the cost shifts onto the employee, and the health benefits don't offer very good value for money.

This is a fascinating topic. Another option is a mix of 6 and 7: people are lousy at estimating the economic or even simple cash value of their benefits. They see the total package and vaguely sum it up, but don't think of the total cash value of the benefits they are likely to consume in a year, calculate the equivalent in taxable W, and compare it to what they spend on the plan (or what else they could do with the money).

Which raises the question: why do people even like benefits? For the median worker types (with families, low risk tolerance, etc.) my hypothesis is that people like the idea of benefits because they see it as the employer expanding the security of compensation into other aspects of their lives. Going to the dentist or physio and being charged zero or $3.75 feels like bonus pay to lots of people. What they don't track is the speed at which they exhaust the relatively narrow band of benefits they do use, and how much that really cost.

If this is true, perhaps unions and employers also like them because they can deliver increases in compensation that employees/members will vastly overestimate.

Shangwen - Interesting points, thanks for commenting.

"people like the idea of benefits because they see it as the employer expanding the security of compensation into other aspects of their lives"

Do you think there's also a good-jobs-pay-benefits-therefore-if-I-get-benefits-it-means-I-have-a-good-job thing going on here?

What about simple inertia?

From what I have heard, employer provided health insurance in the U. S. started as a way to offer more compensation when wages were capped during WWII. Over (a short) time workers began to view such benefits as rights.

Why the association with unions? Maybe with de-unionization, employers could more easily do away with compensation that is hard to value than with actual wages.

Related questions: Are executives more likely than non-union workers to receive health insurance and similar benefits? If so, why?

Frances: "good-jobs-pay-benefits-therefore-if-I-get-benefits-it-means-I-have-a-good-job"

If it's the case the union jobs (almost a synonym for public sector jobs) have better benefits, and the median worker is risk averse and not analytically oriented, then I can see how benefits would be seen as a tautological proxy for job goodness. It may also be that, since benefit coverage is relatively rigid across salary bands (and also across collective agreements), workers will also view benefits as less evaluative of the market value (unlike their wages) and more as a cocoon or security blanket. This is a particularly crappy form of egalitarianism that has wide appeal.

I wonder how a lot of employees would feel if you demonstrated in detail that their benefits are mostly foregone wages, with a portion of it going to administrators in remote cities. The rest of it--and one of my main reasons for disliking benefits plans--is that it's essentially a subsidy to high-income regulated professionals and an upward force on their fees.

Frances,
I work for a benefits consultancy that works with many kinds of employers (500 employees - 50,000+). I think that a combination of 4/5/6 is the most likely. The biggest difference that we see between unionized and non unionized plans is in the composition of benefits rather than cost per say. Union plans are usually loaded with generous paramedical and/or vision benefits while non-union plans often opt for large Health Savings Accounts. Also, union plans tend to be more generous to families while non union have greater individual coverage (all else equal). I think there is a substantial component of self delusion on the part of members as union leaders annually "sell" delivering higher benefits to their members while knowing perfectly well the trade-off between benefits and salary. On an anecdotal basis, despite my attempts to persuade her, my mother who is a long term public sector worker does not believe that there is a tradeoff between salary and benefits. She seems to place benefit costs in a separate "bucket" from salary.

Here's an idea ... public sector unions aren't just bargaining with their employer, they're bargaining with the taxpayers and public opinion. Because health benefits aren't explicitly denominated in dollars, the public will perceive remuneration as lower than it actually is, and the government will be able to increase compensation without public outcry.

That would also explain defined-benefit pension plans, which (to me) appear to be quite generous. And training, and vacations, etc. Especially when these things are seen as "rights" rather than (say) "luxuries" or actual pay. The latter two "count," while the former doesn't, in many minds.

This also applies to the private sector; any union on strike has to be seen to be making reasonable demands. Keeping the paycheque lower, and receiving compensation in a form that's more difficult to measure, helps there too.

Funny hoe people talk about public sector union member having paid health benefit since I have none.
My members however are weel aware that "benefits" and pensions are part of their compensation and come from their overall pay since, during negociations , we have numerous general assemblies where we discuss the mix of compensation.

How about a collective bargaining / collective benefit argument? A sort of coordination change. I've always been a non-union employee and for the most part never had any collective employee bargainer. In order to make workplace insurance viable, all employees must be covered (or at least a lot of them). So there's never been any way for me to negotiate on health plans - it's always "here's what you get". I can easily negotiate on wages because (as non-union) everyone is paid differently. As faculty I'm now somewhere in the middle - faculty representatives have limited ability to negotiate and wage differentials exist but are smaller than private companies I was at.

I assume that a union position changes that all around: I wouldn't be able to negotiate wages, but my representative has ability to negotiate a whole range of compensation - health, dental, vacation, and wages - where as an individual I would only be able to negotiate some of them.

By that argument one would expect that median non-union workers are getting a higher wage/benefit ratio than they want. That would fit the data fairly well.

Great comments!

Shangwen: "is that it's essentially a subsidy to high-income regulated professionals and an upward force on their fees."

Both the fees and the quantity of treatment - dentists who treat to the maximum a plan will allow are almost a cliche.

honeyoak: " union plans tend to be more generous to families while non union have greater individual coverage (all else equal)" Fascinating observations. This fits with the theory that management designs salary packages with mobile workers in mind.

Phil "Because health benefits aren't explicitly denominated in dollars, the public will perceive remuneration as lower than it actually is" Interesting. Though it could work the other way, too - people might see benefits as being worth more than they actually are.

Jacques Rene - "since I have none" Does this mean there's no coverage, or that it's entirely employee paid? The degree of cost sharing is really important. But do you think the typical employee is aware of their cost-sharing arrangements? To be honest with you, I had to look mine up.

Squeeky Wheel: "my representative has ability to negotiate a whole range of compensation - health, dental, vacation, and wages - where as an individual I would only be able to negotiate some of them." That's basically the story Richard Freeman is saying with his "voice" argument. But, at the same time, wouldn't profit maximizing employers have an incentive to try to design the most attractive compensation mix possible? Otherwise, an employer could offer health benefits + a lower money salary and scoop up all the best workers/offer a slightly lower total compensation package (to offset the more desirable mix of compensation). There has to be something else going on here, e.g. employers tailoring packages to the marginal not the median worker, or people underestimating the value of benefits, or a less compressed wage structure.

Honeyoak: interesting comments. I would have assumed union plans were more lavish but apparently not. I wonder how much of this is supported by the perception by (mostly public sector) union workers perceiving the private sector to be a hellhole of immiseration.

Theory 2--overcoming adverse selection--is appealing to me. But how much of the efficiency gains are offset by rising costs and prices due to supply-sensitive consumption?

Frances, on a related note, here's a systematic review that suggests that routine dental care "could be extended to once every two years".

1) Perhaps the process of collective negotiation via unions causes the members to reflect more thoroughly on compensation and notice issues like the tax advantages of taking some compensation as health insurance vs trivially comparable wages. The time pressure on the evaluation goes away.

2) Perhaps they also reflect more thoroughly on the problems of individual insurance (at least in previous years). At a minimum they can, as a group, actually discover what a pain individual health insurance is and negotiate for group coverage in the next contract.

4) Certainly the group of all current employees is going to be older on average than most new employees, so the real value of spreading health insurance costs across the whole group is going to be higher to the former.

7) One sticking point is the word "undervalue". It isn't necessary that the new employees be incorrect, just for them to value it more than current employees. I think that it is true that prospective employees being mostly younger and healthy will objectively undervalue insurance. Current employees, being older and on average probably a bit sicker, will put a greater value on it. It's hard to imagine how this could not be the case. One might expect this to result in decreased wages for older workers, but that would go against longstanding custom, and is likely illegal (wages adjusted based on age or health).

In addition perhaps the power of collective bargaining makes it more likely for companies to go through the hassle of managing comprehensive benefits and the union sometimes takes on some of that benefit management overhead.

"for them to value it more than current employees"

Less, of course. Less.

Interesting post & discussion. As a non-union employee in a professional job I have seen a different trend in recent years, which fits what HoneyOak says above - large Health Savings Accounts as well as a wide (and confusing) set of options that seem to demand that I know what health problems I am likely to experience in the coming calendar year.

Perhaps a separate post on the apparent irrationality of what non-unionised employees are offered would be interesting too? The approach seems to walk head-on into adverse selection issues. If I had to guess, I'd say that employers in non-unionised workplaces overestimate the value of private choices.

Academic economists suffer from the "which is the reason" problem.

The correct answer is: All of the above (except maybe #7). There is not a single reason.

If you are smart, when you negotiate (say for example, a car) you generally negotiate for the *whole package cost/benefit* (including fees and the trade in), not just for a subset. That way they do not nickel and dime you on the trade in, or dealer fees (this technique is also known as "nibbling" - asking for "small stuff" at the end - a good negotiator can get 3-5% more value by nibbling). You don't want to negotiate higher wages and then have (even a vocal minority) pissed that the company took away the wages in the form of higher deductibles.

Also, keep in mind, *by law, if its not written in the contract, the company can do whatever they want*. That means, work rules, seniority, and hiring/firing decisions, and disciplinary rules are often there too. The union will often also bargain for the *number of positions* and what to do with vacancies (I can cite some examples). There are about 3000 ways a company can screw you if you *only* put wages in the contract. Put wages and benefits in the contract, they screw you on hiring/firing and eliminate positions. Put wages, benefits, hiring/firing in the contract, they screw you on work rules. Even if what goes into the contract is "no change in the current benefits/work rules/etc." Sometimes the contract will even say, "when you open another plant you must do xxx" or "you may not move work to South Carolina or other right to work state."

In reality, union negotiators are trying to please multiple constituencies: some who overestimate the benefits, some who are older and value them more, others who appreciate the tax subsidies, and others who just want higher wages. Union bosses are also monopolists who favor a compressed wage structure, their goal is to maximize the benefits to workers.

The other benefit by the way is that unions negotiate for the same health care to be carried over in retirement (whereas in many cases, corporate health care ends at retirement and you have to sign up for Medicare).


I see the same issue in the minimum wage literature, economists puzzle over *which way* employers react to an increase in the minimum wage - redistribute schedules, cut non-wage benefits, use technology, cut costs elsewhere, lay people off, etc. It is rarely the case that there is only one reason employers (unions, groups of people) do something, or even that one reason dominates. There is probably an analogue of the Coase Theorem in here - all costs/benefits in a group are internalized. Typically people in charge have to satisfy multiple constituencies at once, even if each of them in itself is a minority.

This might possibly make sense when employers pay for the cost of health insurance. Employer health insurance contributions are not a taxable benefit, so the tax savings from paying for routine dental exams through insurance might outweigh the administrative and moral hazard (overuse) costs of using the insurance system. But insuring routine dental care makes no sense when the insured individual pays the full cost of the insurance plan - for students, for example.

When employer health plans first became widespread, it was explicit government policy in both the United States and Canada to make them a "pseudo-social programme". That is, the benefits were received tax-free by the employee but still counted as a corporate income-tax deduction to the employer. With the very high tax rates in effect during and after WWII, this provided a strong financial impetus for such plans.

Defined Benefit Pensions were subject to the same policy. In fact I recently came across one of my grandfather's tax returns from 1960 when cleaning out my grandmother's home. Aside from ancestor of the Basic Personal Exemption, the only other personal deduction available was for a DB pension contribution. (It's also only one page.)

Make no mistake, employer benefit plans (aside from Disabilty Insurance) are entirely creatures of the Income Tax Act. Without the tax support, they would disappear quickly.

If you model DB pensions was a pseudo-social programme paid for out of tax revenues, then it becomes apparent that the corporate income tax cuts of the past 30 years would do and have done the DB pension no favours. Each CIT cut transfers more of the risk off the government and onto private employers.

I find the contrast with the UK intriguing where employer health plans are much less widespread due to the NHS. AIUI, "Health Insurance" in the UK usually refers to what we call "Disability Income Insurance"; to replace loss of earnings due to inability to work.

I think it's all about risk. Health insurance is not the only way in which firms insure workers: although a worker's productivity is highly volatile over time, the vast majority are paid constant salaries or wage rates, implying that the employer has absorbed essentially all of the risk. No doubt they do this because workers are risk-averse, so in exchange for taking a lot of extra risk, firms earn a hefty risk premium by paying workers less. Similarly, firms can earn another risk premium by insuring workers' health risks as well.

The fact that union workers are more likely to have health coverage does not imply that unions are specifically bargaining for it. Health insurance represents a per-capita cost to the firm--no matter how productive a worker is, their health insurance will cost the same. The costs of health benefits are ultimately deducted from employee compensation, meaning that lower-paid workers will be less interested in health benefits that will subtract more from their already low wages. And if the worker is also low productivity, the firm would rather not hire the worker than pay the insurance. Union firms, however, already have relatively high pay for all their workers, meaning both that 1) workers are more willing to take insurance instead of cash and 2)low-productivity workers have already been driven out of the firm by the high wages.

So it needn't be that unions want employer sponsored insurance, just that already-unionized firms have the characteristics that make them good candidates for ESI, which employers prefer to paying cash because of the risk premium.

But, to the extent we are talking about pre-Obamacare USA, we also need to remember that for the vast majority of the population, employer-sponsored insurance was basically your only option if you're not old and not a dirt-poor single mother (until Obamacare, most states restricted medicaid so that an employed childless adult could not qualify under any circumstances).

Jeff, interesting comments. In terms of subjective valuation by various groups (new/current, old/young) I think it takes place on two planes: (1) you can accurately judge whether you prefer ESI at all, as you say, but (2) as a modal employee you will be a terrible evaluator of the monetary value of the plan; likely you will overestimate your prospective consumption and the related co-pay out of the ESI. So the marginal changes impress at the symbolic level but the value of those changes are obscure. I can imagine many circumstances where an expansion of benefits--ultimately leading to higher premiums--would leave many worse off.

Matthew and dwb, correct about the American context, but in Canada ESI covers a relatively small set of mostly outpatient benefits: Rx and dentistry account for most of the expenditures, with the residual going to physio, massage, vision, etc. So the protection that is gained by additional coverage is, from the perspective of total health, infinitely small.

Determinant, I don't know the tax issues well, but would they really disappear? Certainly not from the public sector. I stick by my theory that people like the idea of going to the dentist on a Saturday and seeing a little "extra pay" in the form of a $195 contribution from the boss.

Determinant, I don't know the tax issues well, but would they really disappear? Certainly not from the public sector. I stick by my theory that people like the idea of going to the dentist on a Saturday and seeing a little "extra pay" in the form of a $195 contribution from the boss.

90% of employee benefit plans are sold to the sponsor on the basis of the tax treatment. Remember the employee does not purchase the plan, the employer does.

In the public sector, there are still income tax issues as each level of government pays *some* income tax to the other, and that's a cost to the entity paying.

Matthew and dwb, correct about the American context, but in Canada ESI covers a relatively small set of mostly outpatient benefits: Rx and dentistry account for most of the expenditures, with the residual going to physio, massage, vision, etc. So the protection that is gained by additional coverage is, from the perspective of total health, infinitely small.

Stop, stop right there. ;) You forgot Disability Income insurance. In Canada this is near-universal for any full-time job. It does not receive any preferential tax treatment, just the standard insurance treatment (you either tax the income to pay the premiums, or the benefits, but not both). The employee benefit market is broken into segments, but insurers generally insist that you purchase it as part of any benefit plan. It is standard for smaller "retail" plans for small businesses. And a good thing too, it protects against a genuine catastrophe, the loss of income due to impaired health.

And as a Type 1 Diabetic, prescription benefits are no small matter. The fact that they are not part of the public health system like they are in the UK is something I will never forgive Tommy Douglas and Woodrow Lloyd for.

Since we were discussing mostly "employer paid" benefits, I meant none of them ( apart a guarantee of not losing your job and still having a (reduced) salary for the first two years of a medical leave.) Noe we have a compulsory employe paid disability scheme,kicking in atfter the second year. (Cegep professors , however have an administrative understanding that after two years of disability, you can voluntarily resign or stay on the list. We have a case of a member coming back after 11 years of disability. No other "corps d'emploi" in the QC public service has that possibility. They are definitely discharged after 2 years.
Our employee-pais compulsory drugs and dental plan is multi-tiered. Almost half a dozen combinations are possible, chosen by the member and modifiable under certain conditions at certain time of the year or modification in family events or employment status change, such as birth of a child or getting tenure.
For people of my professionnal rank (economist) private sector is not viewed as a hellhole but as a paradise. Only ousider think of QC public servants as privileged. Usually readers of newspaper articles writen by less qualified but better paid than we are journalists.

I think there is another issue, the principal-agent problem.

The union has to seem like it “wins” in any negotiation lest the leadership loses its grip on power. The best way to achieve a “win” is to ensure that the compensation package is as complicated as possible. The more difficult to evaluate, the better. If the union leadership only negotiated wages, it would be very clear to the membership whether the union won or lost in bargaining - nothing is so simple to understand as your take-home pay. The leadership knows far more about the details of a negotiated compensation package than the rank and file membership. The leadership can use its information advantage to argue that the compensation package represents a great win when in fact an equivalent increase in wages would have disappointed the membership. Union leadership has little incentive to simplify the compensation package just as the government has little incentive to simplify the tax code.

tomslee "large Health Savings Accounts as well as a wide (and confusing) set of options that seem to demand that I know what health problems I am likely to experience in the coming calendar year."

This is in the US? The Canadian health insurance landscape is so different it's hard for me to compare/comment - the big things covered by private insurance are dental, pharmaceuticals, and then (to varying extents) things like therapy, vision, physio. Would you say this is an example of people being made worse off by having too many choices? Does the wide set of options nudge people in particular ways?

Determinant "then it becomes apparent that the corporate income tax cuts of the past 30 years would do and have done the DB pension no favours" This is an important point, but I think that you need something else to make employers favour DB contributions over salary payments, e.g. possibly the pension plan contributions can be put in a plan that generates some kind of surplus one way or another for the employer

Matthew "although a worker's productivity is highly volatile over time, the vast majority are paid constant salaries or wage rates, implying that the employer has absorbed essentially all of the risk"

And in a unionized workplace, it's much harder for an employer to let an employee go if they are no longer productive. So it makes sense for employers to invest more in their employee's health care, because health and productivity tend to be correlated.

Determinant "then it becomes apparent that the corporate income tax cuts of the past 30 years would do and have done the DB pension no favours" This is an important point, but I think that you need something else to make employers favour DB contributions over salary payments, e.g. possibly the pension plan contributions can be put in a plan that generates some kind of surplus one way or another for the employer

No, DB pension sponsors cannot, in general, commandeer pension surpluses for themselves by brute force. Generally you can take a "contribution holiday" but with government funding rules and rules about overfunding, you can never consistently take a profit that way. Unless you are the Federal Government and you write your own pension laws yourself.

What Canada did do was to make actual Pensions (not RRSPs) have far more contribution room and thus tax benefits than RRSP's. They were and are exempt from payroll taxes too. Contributions between RRSP's and Pensions were equalized in 1988.

That was a big impetus for the decline of DB pensions. Group RRSP's (ugh) were legalized at the same time too (by provincial governments, actually, by not calling them pensions)

"Determinant: "90% of employee benefit plans are sold to the sponsor on the basis of the tax treatment. Remember the employee does not purchase the plan, the employer does."

Sure, but its the tax treatment to the EMPLOYEE that is beneficial, not to the employer. The employer can deduct their compensation to employees, whether they pay cash or in kind through benefits. But whereas the employee has to pay tax on cash income, they can't claim a deduction for insurance premiums (although they may be able to squeeze it into the medical tax credit - which isn't as generous). So employee health benefits are worth more to employees on an after-tax basis than it costs the employer to provide - hence the incentive. This is why these sorts of plans make sense in the public sector where the government pays no taxes.

That being said, if you introduced a general deduction for health care premiums (which would equalize the tax treatment between employer provided and self-purchased insurance), most people would still get their health insurance through their employer for the usual risk-pooling, adverse selection related reasons - namely that employers will be able to get a better rate.

Determinant: "In the public sector, there are still income tax issues as each level of government pays *some* income tax to the other, and that's a cost to the entity paying"

That's not correct. In Canada each level of government is constitutionally exempt from tax by the other (though in some cases, for example, in the HST provinces and Quebec, the provincial governments have waived that exemption with respect to HST). Moreover, many government bodies that are not themselves constitutionally exempt (think municipal governments or crown corporations) are exempt from tax under the Income Tax Act. So they don't care at all about deductions.

Determinant: "If you model DB pensions was a pseudo-social programme paid for out of tax revenues, then it becomes apparent that the corporate income tax cuts of the past 30 years would do and have done the DB pension no favours. Each CIT cut transfers more of the risk off the government and onto private employers."

I suppose it is true that, as a general proposition, taxes involve the government taking on a degree of the taxpayer's risk (at least where you can have losses ), at least to the extent that loss can be deducted against gains (or carried back to generate tax refunds). I do seem to recall there being a paper showing that the taxation of capital gains should promote riskier investments (since the government shared in the loss). So, to that extent, yes, lower taxes might induce a corporation to take on less pension risk.

That being said, I doubt that tax changes (or at least changes to the corporate tax rate) would be a signficant story in the context of the decline of DB pension plans for a couple of reasons. First, in many instances, when a pension risk is realized (i.e., the pension liabilities are higher than expected and the company goes bust) the benefit of having government share risk by being able to deduct losses is worthless. Second, as you noted, the legislative advantages for employees associated with DB pension plans was largely eliminated in the late 1980s. Third, the decline in DB pension plans, at least in Canada, long pre-dates the relatively recent reductions in corporate income tax rates (which didnt really start until the late 1990s). A more likely explanation is that (a) people started living longer than expected, resulting in under-funded pension plans (and health plans), and (b) traditionally industries (think cars, heavy industry, etc.) faced with increased competition. Faced with that one-two punch companies decied they can't afford to take on that pension risk anymore.

Since insurance isn't a taxable benefit, it doesn't factor into who shows up on Ontario's sunshine list. It's probably not a good thing for groups like college teachers to have many members showing up on that list as it may increase public outrage. So there may be an element of "hiding" income, at least in the public sector unions with higher salaries.

As a trustee of a multi-employer public employee pension plan through
much of the 90's and early 00's I can testify to two things:-

Most plans generated a paper surplus in the late 90's due to the dot
com bubble.A lot of plan providers used this to take contribution holidays
which, in a lot of cases, led to the underfunding we are now facing. Who
knew that bubbles tend to burst? My own plan attempted to do this but
it was blocked by the trustees. The employers went ahead and did it
anyway ( it's our plan !). Only a serious threat of lawsuit persuaded
them to replace the funds.

Also, the paper surplus led to a parade of consultants prancing through
the system selling the idea of DC plans as the best of all possible
worlds. Independence! Wealth! Unfortunately, a lot of plans eagerly
bought that package and sold it to uninformed plan members.

And, for most plan members who typically live paycheck-to-paycheck,
the sudden impact of a large dental bill or prescription medication
can be a real hit. Like most insurance it is only loved when it's
needed. So, nobody likes paying for it. Fortunately, people who work
together tend to talk to each other .. especially, I think, people
who work in health care. So they come to know the horror stories of
what can happen to those without coverage.

Having been involved in numerous rounds of collective bargaining I can attest to some of the strategic elements mentioned. On occasions when government has imposed limits on wages, it is a way to provide alternate compensation. Within our workplace the employer negotiates with several different unions, and it is clear that in that context they often want to hold the line on percentage wage increases, but are willing to throw more money in for some if it can be hidden somewhere else. And in presenting demands it is usually thought wise to include some throwaways, which sometimes surprisingly get accepted - we finally got partial spousal and dependent tuition waivers last time, after asking repeatedly for almost thirty years.

Jim "but are willing to throw more money in for some if it can be hidden somewhere else"

This is something that I hadn't realized until recently. Universities are institutions with (largely) fixed budgets that go (largely) to salary costs. In this context, it comes down to a case of each union trying to get as big a size of the pie as possible, and some unions have more leverage than others.

Bill: "the sudden impact of a large dental bill or prescription medication
can be a real hit. Like most insurance it is only loved when it's
needed."

But the irony, at least with dental insurance, is that the large bills are precisely what's not covered. Anything over $500 has to be pre-authorized (in my plan), and not infrequently the requests are turned down.

"e.g. employers tailoring packages to the marginal not the median worker, or people underestimating the value of benefits, or a less compressed wage structure. "

I think I was trying to say that from the employee perspective. The company wants to bring in the marginal worker at the lowest cost and wages are more easily tailored to the individual marginal worker than benefits are. Same happens from the marginal employee's perspective (marginal as in the next one hired, not marginal as in lousy).

What would happen if non-union workplaces published all wages (http://www.npr.org/blogs/money/2014/07/02/327289264/episode-550-when-salaries-arent-secret)? Would that change things?

That's not correct. In Canada each level of government is constitutionally exempt from tax by the other (though in some cases, for example, in the HST provinces and Quebec, the provincial governments have waived that exemption with respect to HST). Moreover, many government bodies that are not themselves constitutionally exempt (think municipal governments or crown corporations) are exempt from tax under the Income Tax Act. So they don't care at all about deductions.

That is irrelevant, and you forgot the jurisprudence to boot. Employees of each level of government are still liable to each other's income tax (thanks to an Australian precedent from the Privy Council) and on net it is the employer on whom the incidence of income tax actually falls. An increase in income tax from the other level of government leads to a demand for a salary increase for public servants.

Third, the decline in DB pension plans, at least in Canada, long pre-dates the relatively recent reductions in corporate income tax rates (which didnt really start until the late 1990s).

http://www.camagazine.com/archives/print-edition/2011/june-july/regulars/camagazine49727.aspxand

CIT rates were at 52% in the 1970's and dropped in the 1980's under Mulroney. Over to you, Bob.

Why focus on health benefits? Surely a more interesting question is: do unionised workers have more complex compensation packages than non-unionised workers (extending beyond health benefits)? For if that is so, then we have identified a union interest in compensation complexity.

It is not hard to see why unions would favour complexity. They are intermediaries between workers and employers. The more complex the compensation arrangements, the more workers "need" their intermediary to manage/negotiate for them. (Employer organisations have a similar interest in compensation complexity, if they are negotiation intermediaries for firms.)

There is also a justification narrative issue. Each element in the package then becomes a marker of union achievement. A sort of in-place talking-point for the rhetoric of unionisation.

There was a very revealing case in Australia some years ago. The mining companies in the Kimberleys offered to "cash out" very holiday leave-loading entitlements for higher wages, since miners had become largely more based around 2-weeks on, 2-weeks off arrangements (I am going from memory on the weeks on, weeks off). The unions resisted. The workers loved it, and there was a dramatic drop in union membership. It was a straight tussle between unions committed to a compensation package being more complex, with markers of union achievement built into it, and (highly paid) workers who want even higher-cash payment simplicity. (So highly paid that just-hiring-a-lawyer had become a live alternative to union-as-intermediary.)

(This is a more general version of Avon Barksdale's principal-agent point, but focused on the general role of unions, rather than particular union leadership.)

"on net it is the employer on whom the incidence of income tax actually falls. An increase in income tax from the other level of government leads to a demand for a salary increase for public servants. "

Got any evidence to support that remarkable proposition? I could see a case being made that employers bear the incidence of employee income tax in the case of highly mobile and irreplaceable workers - think sports stars or CEOs (i.e., people who do get premiums for agreeing to work in high tax jurisdictions). I think it's fair to say that "highly mobile and irreplaceable" is not a reasonable description of most civil servants (or, indeed, most employees in any sector in Canada). Moreover, it's not clear why higher taxes would result in increased demands for a salary increase. Presumably employees (public or otherwise) always want higher incomes if they can negotiate them. The effect of higher taxes is to depress after-tax income accross industries, so it's not clear why government employees are in a better bargaining position to negotiate a pay increase after a tax increase than they were before. You certainly don't see public sector wage demands go down when taxes are cut.

"CIT rates were at 52% in the 1970's and dropped in the 1980's under Mulroney. Over to you, Bob"

I stand corrected. I stand by my original point, though, that the impact of the tax system on the decline of the DB pension plan, if any, is immaterial.

Indeed, two points are telling. First, the downward trend in private sector defined benefit pension plan (and the upward trend in DC plans) doesn't seem to have any relationship with changes in tax rates. Indeed, the biggest corporate tax cut in the last 4 decades took place in the 1970s, when federal tax rates on manufacturing and processing corporations (i.e., those industries which traditionally offered generous DB pension plans and which have been disappearing since the 1970s) were slashed by 25%. Curiously, the decline in DB coverage doesn't start until a decade later (although before the next round of corporate tax cuts in the late 1980s. The subsequent trend is fairly consistent notwithstanding sharp tax cuts thereafter.

(see: http://www.parl.gc.ca/content/LOP/ResearchPublications/prb0706-e.htm#overviewdomestic and http://www.caledoninst.org/publications/pdf/625eng.pdf at page 21)

Second, the fact that the same trend (albeit less pronouced) appears in the public sector (where employers are exempt from taxes) also suggests that something else is at play here.

The decline in DB pension plan coverage isn't an employer side tax story. It's partly an employee side tax story (the 1988 changes took away their tax advantage of DC pans and RRSPs), but I think its principally a demographic and industry change story (not for nothing, but the decline in DB pension plans since the mid-1980's looks an awful lot like the decline in manufacturing employment over the same period). You've made this point before, DB pension plans are glorified annuities, but most businesses aren't insurers. No wonder they want out of that business.

"It is not hard to see why unions would favour complexity. They are intermediaries between workers and employers. The more complex the compensation arrangements, the more workers "need" their intermediary to manage/negotiate for them. (Employer organisations have a similar interest in compensation complexity, if they are negotiation intermediaries for firms.)"

At the risk of defending unions (not my usual role), could part of this story be less Avon's principal-agent problem (although I think there's likely something to that), and more of a story of unions playing a paternalistic role in protecting their members from themselves?

I'm reminded of an old "All in the Family" episode ("Archie's Raise") where Archie and his buddies go on strike, but ultimately settle for a raise over an escalator clause that would protect against inflation (over Meathead's objections) - this in the 1970s. In your example, I could see someone telling a story that refusing to give up hard-won entitlements for cash payouts is in the long-run interest of union members (including future union members - keep in mind that Union leaders are likely to have a longer time horizon than individual members) even if its not what current members would choose if left to their own devices.

You see an aspect of this a lot of employment negoations. Often, employers are willing to make concessions if the union agrees that, going forward, new employees will be covered, say, a less generous health plan, or a less generous pension plan. In theory, the current employees should be more or less indifferent to that change - it doesn't affect them and may even benefit them (to the extent that they get something for that concession - but there is no union leader alive who accepts that proposal without a bloody fight.

Certainly union leaders are quite vocal about what they think the government should impose on the rest of us, without regard to our choices (damned Ontario pension plan), it isn't hard to believe that they'd have strong views as to what's in the best interests of their members.

The decline in DB pension plan coverage isn't an employer side tax story. It's partly an employee side tax story (the 1988 changes took away their tax advantage of DC pans and RRSPs), but I think its principally a demographic and industry change story (not for nothing, but the decline in DB pension plans since the mid-1980's looks an awful lot like the decline in manufacturing employment over the same period). You've made this point before, DB pension plans are glorified annuities, but most businesses aren't insurers. No wonder they want out of that business.

First, the DC pension figure in the public sector would be dominated by Saskatchewan's provincial public service, which switched to DC in the late 1970's. That and a few universities in British Columbia.

And that feeds right back into my original point. When the choice is between paying a DB pension or paying CIT, the pension often wins, at least you get happy employees. And a corporation can afford to be indifferent to much of its pension costs when they are actually small, after-tax.

High CIT rates with pension contribution deductibility allows the sponsor to use the tax system as a "buffer" for excess contributions. LIkewise, you could in theory hide corporate profits by overcontributing to a pension plan in certain years. This is why the CRA has limited pension funding to 110% for decades.

When CIT fell it shifted ever more of the burden onto corporate income statements. The choice changed from pension or taxes to pension or after-tax corporate income. And after-tax corporate income won. As CIT fell, companies had much more of a personal stake in their pension plans and became much more cost-conscious about them. And no wonder, the feedback mechanism had changed dramatically!

Of course I stand by "glorified annuities" argument, but it was high CIT rates that allowed the DB pension system to flourish (to make companies even venture into the annuities field in the first place) and the cuts in CIT rates that undermined it.

Lorenzo "do unionised workers have more complex compensation packages than non-unionised workers "

Part of the issue is that union and management almost never start from scratch - collective agreements get modified and modified and modified, with each subsequent modification adding complexity.

The other thing is that complexity can be away of masking the meaningless of concessions. I have a very specific example in mind here - which is causing me a great deal of irritation at the moment because union and management agreed to a cumbersome and mindbogglingly inefficient process that has to be implemented by yours truly.

Complexity is also a product of the dynamics of negotiations - the need to put in stuff that is redundant/useless/trivial/meaningless so that you have stuff to concede - as Jim Sentence was describing earlier.

" In theory, the current employees should be more or less indifferent to that change - it doesn't affect them and may even benefit them (to the extent that they get something for that concession - but there is no union leader alive who accepts that proposal without a bloody fight."
Because any union leader,such as your humble servant, know that once you open the concession game on the young ones, the old one will be next.

"First, the DC pension figure in the public sector would be dominated by Saskatchewan's provincial public service, which switched to DC in the late 1970's. That and a few universities in British Columbia."

Yeah, so what, who's talking about DC pension plans?

"When the choice is between paying a DB pension or paying CIT, the pension often wins, at least you get happy employees."

But that's not the choice. The choice is between paying employees cash or paying them in benefits (be it pensions or health insurance or whatever). In either case, the payments are deductible to the employer, so it reduces their CIT (and their income). There's nothing about that analysis that favours DB pension plans.

"When CIT fell it shifted ever more of the burden onto corporate income statements. The choice changed from pension or taxes to pension or after-tax corporate income. And after-tax corporate income won. As CIT fell, companies had much more of a personal stake in their pension plans and became much more cost-conscious about them."

That makes no sense. First, the obvious point is that the pension liabiities that show up on corpoate financial statement are pre-tax. So nothing would have changed as a result of tax changes. Second, your point is that a $100 pension benefit only costs a company $50 in after-tax income when tax rates are 50%, but $75 in after-tax income when tax rates are 25%. Ok, but that's true of all forms of employee compensation. What makes DB pension plans different from any other form of compensation? Step back for a second, on your theory, companies will increase compensation when taxes are high, but decrease them when taxes fall. Apart from being inconsistnet with common sense(typically, it's easier for unions to negotiate concecessions when companies have high after-tax profits, and it would be easier for a corporation to fund its pension liabilities when it has high after-tax profits), this is inconsistent with the consensus view that, in the long term, corporate income tax is laregely borne by employees (i.e, as taxes rise, they get passed onto employees through lower wages, or lower wage growth).

I think you're right that part of the story of decline DB plans is that companies started paying closer attention to them in the 1980s. But that's not a tax story, that's a story of companies realizing either that their plans were underfunded (as people were living longer than expected) and/or that the Japanese/Koreans/whatever were eating their lunch in the market, and the pension plans that were manageable in the 1960s (when domestic corporations faced more limited competitive pressures) were now deadweights.

Take GM (although one could look at any of hundreds of Canadian companies who have eliminated or pared back their DB pension plans over the last two decades, they'd all have a similar story). It didn't pare back its pension plan because of changes in the tax rate, it pared back its pension plan because it had to compete with Japanese and Korean manufacturers without the same legacy pension liabilities who could sell their cars for less than GM's costs.

"Because any union leader,such as your humble servant, know that once you open the concession game on the young ones, the old one will be next."

Perhaps, but I think this supports the paternalist argument, that the union negotiators may have a broader perspective than their members as to what's in their interest. After all, that's their job.

part from being inconsistnet with common sense(typically, it's easier for unions to negotiate concecessions when companies have high after-tax profits, and it would be easier for a corporation to fund its pension liabilities when it has high after-tax profits), this is inconsistent with the consensus view that, in the long term, corporate income tax is laregely borne by employees (i.e, as taxes rise, they get passed onto employees through lower wages, or lower wage growth).

First, I don't agree with the consensus view on corporate taxes and want no part of it.

Second, as I said pensions, DB pensions especially, were favoured by the Canadian tax system until 1988 as a conscious policy choice. Additionally, to sweeten the pot in Pensions favour, pension contributions were and are exempted from payroll levies, EI charges, CPP, provincial payroll taxes, etc. The same strategy has been implemented with PRPP's. It is a 20% inducement to pay through pensions. The same goes for health plans, actually.

Third, your "common-sense example" is simply nothing of the kind. Pensions and corporate taxes have the exact same dynamic as RRSP's and PIT deductibility. They are both heavily sold on that basis. It is easier to extract concessions when companies are highly profitable on a pre-tax basis and have a tax-efficient way of delivering those concessions. The cost is borne by the government through avoided taxes.

I wrote: "large Health Savings Accounts as well as a wide (and confusing) set of options that seem to demand that I know what health problems I am likely to experience in the coming calendar year."
Frances wrote "The Canadian health insurance landscape is so different it's hard for me to compare/comment - the big things covered by private insurance are dental, pharmaceuticals, and then (to varying extents) things like therapy, vision, physio. Would you say this is an example of people being made worse off by having too many choices? Does the wide set of options nudge people in particular ways?"

This is Canada. We have a certain amount of "money" and get to allocate it between levels of coverage for the things you list. Do I want 100% dental and no physio or do I want full prescription coverage and no optical care? I suspect it does make us worse off, but to be honest I get so confused I would find it difficult to say.

To speculate for a moment: is this heading to the end of insurance? At the end result of a large menu of choices and the elimination of adverse selection problems through monitoring and incentive programs (track your gym activity and get a discount etc) would it be rational to refuse any policy that an insurance company would offer me?

I used to sell insurance; yes those plans exist, and I don't like them much.

The Disability Income insurance is worth it; it's worth it for any worker. Because those policies are actually insurance and priced as such. The prescription drug coverage is worth it for a person like me (a diabetic) but we're right back to adverse selection.

Several of these are plausible.

Another possibility, which overlaps with 4, 6 and 7:

Benefits are information-intensive. It is easy for workers to assess the quantity they are being paid. It is hard to assess the quality of health coverage. Even if there are efficiency gains for group purchase of insurance, they won't be realized if workers can't tell whether what kind of plan they are being offered. (It's a kind of lemon problem.) So the efficiency gains from group purchase of insurance won't be realized unless workers have a trusted third party to verify the quality of the plan on offer.

In this sense, certain employers can benefit from unionization. Offering certain kinds of insurance may be complementary to purchasing labor on (formal or informal) long-term contracts. In the case of defined benefit pensions -- another benefit often negotiated by unions -- deferring compensation may allow the firm to borrow (from the workers, in effect) and the workers to save on more favorable terms than either could get from a third party, given various financial frictions. But precisely because the loan from worker to employer embodied in a pension takes place outside the financial system, it requires a trustworthy guarantor.

Tom - o.k., I've not ever had any experience with that kind of plan.

" the elimination of adverse selection problems through monitoring and incentive programs "

Interesting thought - also the management of adverse selection through much much better data on benefit take-up and risk factors, making it possible to price plans appropriately.

Given that this is the Canadian context, though, perhaps adverse selection is less of a problem? It's not like a Canadian health insurance plan is going to be asked to pay for hundreds of thousands of dollars with of brain surgery. The big risk is for super-expensive prescription drug coverage. But basically private plans here are much less likely to be landed with catastrophic medical expenses than US plans are.

Determinant: "Second, as I said pensions, DB pensions especially, were favoured by the Canadian tax system until 1988 as a conscious policy choice. Additionally, to sweeten the pot in Pensions favour, pension contributions were and are exempted from payroll levies, EI charges, CPP, provincial payroll taxes, etc. The same strategy has been implemented with PRPP's. It is a 20% inducement to pay through pensions. The same goes for health plans, actually.

I don't disagree with any of that, and have already said as much. But that has nothing to do with the corporate tax system. The preferred treatment of DB pensions was on the individual side (in that DB pension plans were treated more favourably than RSPPs), that means DB plans were preferred by employees.

Determinant: "Third, your "common-sense example" is simply nothing of the kind. Pensions and corporate taxes have the exact same dynamic as RRSP's and PIT deductibility. They are both heavily sold on that basis. It is easier to extract concessions when companies are highly profitable on a pre-tax basis and have a tax-efficient way of delivering those concessions. The cost is borne by the government through avoided taxes.

Actually, your comparison with the RRSP/PIT system proves you're wrong. If you save in an RRSP you get a deduction, if you save outside an RRSP or consume, you don't. There's a discontintuity between the taxation of RRSP savings and the taxation of other forms of saving or consumpiton, hence the tax incentive for RRSP savings. But for a corporation, whether you pay your employees with pension benefits or you pay them higher wages or give them other perks (with some exceptions), you get a deduction. The treatment of pension benefits is no more "tax-efficient" a concession than giving cash (indeed's it's not accurate to say it's tax efficient - beig able to claim a deduciton for your expenses allows you to accurately measure your income, it's not advantageous tax treatment). There is are very few forms of compensation that are not tax deductible for corporations (the big one that isn't is employee stock options - so you might be able to tell a story about how tax cuts have increased the willingness of corporations to grant stock options, though I doubt you could tell a story about stock options replaced pensions for most employees). Without a discontinuity between the tax treatment of various forms of benefits, there's no reason to expect changes in the tax rate to affect one form of compensation over others.

Wrong, Bob, and I am getting tired that you aren't reading my posts fully. As I stated, pensions were made a deduction from corporate tax and exempted from payroll tax to make them tax efficient. The decision, in a high tax environment, is between pensions/employee pay and taxes, if a business does not give extra pay to the employees the taxes kick in; either way not much cost or benefit makes it to the bottom line.

It is very clear that in a 52% CIT environment, with a 20% additional tax advantage it is the tax system, not the employer bears 70% of the costs of a pension plan, and when this reduces to 45% at present rates, the sponsor has been stuck with an increase in pension costs to its bottom line by a third. And that changes a sponsor's perspective on cost and risk management.

As CIT decreases, the bottom line absorbs more of the pension cost directly and affect profits more.

Third, you cited in your previous post that there was a small trend to DC pensions in the public sector, I said that was mainly SK and you said it was irrelevant. Please follow my posts if you wish to argue.

Lest you think that deductibility is a given, let us turn our eyes to Disability Insurance. There most plans are not in fact deductible to the employer in and of themselves. A Wage-Loss Replacement Plan (deductible premiums) is, a Grouped Accident & Sickness Plan is not, and the benefits from policies with taxable premiums are received tax-free. Because DI premiums are relatively stable, they are usually insured outright. DI solves the problem of what to do with a disabled employee (financially) and allows them to be terminated with minimal fuss. Tax incentives were not required.

DI did not come around until the 1970's and it did not receive the same tax treatment as prescription drugs and pensions. Those items involve ongoing, variable costs to an employer, and it was seen as beneficial to give them favourable tax treatment to encourage wide adoptions. As I said, a pseudo-social programme in lieu of outright government spending.

As for your tax-cut narrative, I believe it conflicts with figure 2 in this document. There were large corporate tax cuts in the early 1970's, 1989-1995 and in 2005 and later.

http://www4.agr.gc.ca/resources/prod/doc/pol/pub/itdat60-05/pdf/tax_e.pdf

"As I stated, pensions were made a deduction from corporate tax and exempted from payroll tax to make them tax efficient."

Well, as I noted, since there's no difference between the treatment of pensions and most other compensation under the corporate tax system, I don't know why you keep refering to their deductibility under the corporate tax system making them tax efficient. If you think its the exemption from CPP/EI which makes pensions tax efficient, I agree. But since EI and CPP rates have INCREASED since the 1970s, on your theory that should increased the offering of pension benefits.

"Lest you think that deductibility is a given, let us turn our eyes to Disability Insurance.There most plans are not in fact deductible to the employer in and of themselves."

Determinant, are we talking past one another here, are you sure you understand what it means for an expense to be deductible to the employer? With few exceptions all costs incurred in compensating employees are deductible under section 9 of the Income Tax Act. Disability insurance is treated differently then pension beneits (and certain other benefits)in the sense that it's treated as a taxable employee benefit to the employee, but that has nothing to do with deductibility for CIT purposes. That's solely a benefit for the employees. I agree that the differential treatment on the employee side would make DB plans more valuable than forms of taxable compensation to the employee, but that's not a function of the CIT rates.

Indeed, if you wanted to make that argument, you might suggest that declines in the PIT rate since the 1970s made DB benefits (and other non-taxable benefit) less valuable to employees (since it would reduce the value of the tax benefit), resulting in reduced demand for them. In theory that makes sense, though empirically I think many employees value DB pensions regardless of their tax treatment (perhaps irrationally), so I wouldn't claim that was a strong factor in the decline of DB benefits.

"Third, you cited in your previous post that there was a small trend to DC pensions in the public sector, I said that was mainly SK and you said it was irrelevant."

It is largely irrelevant, but in any event the claim that it was largely as a result of Saskatchewan's shift in the 1970s is belied by the chart I posted, which shows the trend of shifting from DB to DC plans starting largely in the 1980's (at the same time as for the private sector) and continuing over the next two decades. Chalking that up to Saskatchewan and "a few universities in British Columbia" doesn't cut it.

"As for your tax-cut narrative, I believe it conflicts with figure 2 in this document. There were large corporate tax cuts in the early 1970's, 1989-1995 and in 2005 and later. "

I'm sorry, but how does that conflict with my tax-cut narrative? I observed that there were significant cut cuts in the mid-1970s - as you chart shows - but that the decline in DB pension plans didn't chart until nearly a decade later (albeit before the second round of CIT cuts in the late 1980s), and then declined steadily, with seemingly little regard to variations in CIT rates (i.e., it doesn't decline faster when CIT rates are cut). That the rate of decline in DB pension plan coverage seems to be more or less independent of CIT rate changes, at first glance, seems fatal to your proposition that declining CIT rates drove the decline in DB coverage.

"Those items involve ongoing, variable costs to an employer, and it was seen as beneficial to give them favourable tax treatment to encourage wide adoptions"

Sigh, I'm going to say this again. Yes, they get favourable tax treatment, but not to the employer, to the employee. The employer is indifferent (at least under the CIT system) between providing taxable or non-taxable benefits, either way it can claim a deduction. The employee is not indifferent, under the PIT system, in receiving those benefits. The favoured tax treatment is that employees get benefits that aren't taxed. That has nothing to do with the taxation of the employer.

Enlightening responses, thank you.

I have a very specific example in mind here - which is causing me a great deal of irritation at the moment because union and management agreed to a cumbersome and mindbogglingly inefficient process that has to be implemented by yours truly.
Perhaps part of the dynamics of negotiations is they are not done by administrators?

I'm late to the party, but nobody has mentioned what seems like the most obvious answer to me: Unions are predominantly run by their more senior members, and health benefits disproportionately flow to that demographic. If you're 55 years old, paying $2000/year for the latest statins, and you're on the bargaining committee, why would you not give up $1000/year of wages in order to get health benefits? (And in addition to the age / medical costs correlation, employees with greater seniority are more likely to be married and have kids -- no surprise that they'd negotiate to receive coverage for their entire families, subsidized by predominantly younger and single junior workers.)

Fun thing that my union has three different drug plan with various coverages and different level of payment for original and generic drugs and various coverages for singles and families. All negociated by us old geezers...

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad