A firm wanting to invest in a worker, to train her and give her valuable skills, runs a risk: what if she leaves, and takes her valuable human capital elsewhere?
A firm can induce its employees to stay through deferred compensation. It pays new employees less than their productivity merits, and senior employees more. The junior workers work hard and invest in themselves and the company, because they know they will get the big pay-off eventually. The senior workers are happy because they are well-remunerated. The firm is happy because it has a productive and highly trained workforce.
A deferred compensation contract must have a fixed termination date, otherwise the firm risks being burdened with a large number of unproductive older workers (yes, at some point, productivity begins to drop off). But employees also need some guarantee of job security, otherwise they would not accept deferred compensation in the first place - they know the firm would fire them the minute their salary rose above their marginal productivity.
A standard retirement age - a pre-determined time at which employer and employee agree to part ways - is one example of a fixed contract termination date. Current university pay structures, at least in Ontario, were structured with a retirement age of 65 in mind.
Canada, for the most part, no longer has a standard retirement age. But if us geezers hold on until we are forceably removed from our offices, average productivity will decrease, average pay will increase (given existing salary structures), and firms' financial viability will be jeopardized. Sure, some superstars are still highly productive at 65, 70 or 75. They're typically the ones who retire from Firm A and start a new job at Firm B so that they can collect a pension and a salary simultaneously.
When I was ranting about this subject in class the other day, a student put up her hand and said "but what about people who can't afford to retire?" That got me thinking: what about people who would retire if they were only concerned with their own financial situation, but who keep working because they want to help their children get established in life?
It's Catch 22: the geezers can't retire, because they're paying their kids' mortgages; the kids need help paying the mortgage because there are no good jobs out there. Why? Because the geezers have them.
Let's assume that, over their lifecycle, children are about as productive as their parents. That's plausible enough - with a bit of training, my kids could be as good profs as I am. Also assume that productivity first rises, then falls, with age. These two assumptions together imply that, when a past-her-prime prof steps down, and passes her position onto her child, average productivity increases.
It's a second best solution: it would be better to replace the the past-her-prime prof with the best candidate for the job. But how can the past-her-prime prof be persuaded to go? Recall that deferred compensation schemes come with an element of job security. Employment protection does not (or should not) protect the grossly incompetent, but someone past their prime might still be performing with some minimal level of competence, and thus be hard to fire.
Nepotism encourages retirement through two, separate channels. First, when a person's children have good jobs and are financially self-sufficient, a person has less need for income, so can afford to retire. Second, the possibility of creating work for one's children gives a person a long-term interest in the future of the firm - if the firm goes bankrupt, there will be no jobs for the kids. Hence a person is induced to retire - if that is what it takes to keep the firm solvent - in order to protect his children's prospects.
Nepotism - promoting one's children - is an ancient tradition with a long history. Any institution that has survived so long must have something going for it.
(Nick Rowe: if you're reading this and thinking it's time for you to go, this is what students are saying about you on ratemyprofessors "without him is like no snow for winter.keep it up nick!")
The Roman Empire developed a caste/feudal system for occupations starting in the late 300's. For instance baker's sons had to be bakers. But this was more a response to Rome's failing economy.
That's all. I am really not the person to talk to about nepotism, favouritism or patronage right now as I have just filed an appeal with the Public Service Commission over a selection process gone freaky.
Posted by: Determinant | January 24, 2012 at 11:34 PM
On historical patterns, if transport and communication costs are high, nepotism is further advantaged, given the range of potentially more productive folk is much smaller.
Nepotism can also have useful social externalities, in promoting commitment to a stable political order. Hereditary knightly landownership promoted social stability in medieval Europe, and made the Christian crusader states more stable than the neighbouring Muslim rulerships (which were more "meritocratic" in their warrior tax-farmer arrangements).
Posted by: Lorenzo from Oz | January 25, 2012 at 12:51 AM
"A deferred compensation contract must have a fixed termination date, otherwise the firm risks being burdened with a large number of unproductive older workers (yes, at some point, productivity begins to drop off). ..."
This looks like a contract design problem to me. If the deferred pay was written as such in the contract (conditioned on the worker still being employed by the firm), there would be no need for fixed retirement; just let the bonus pay expire and scale the wage back to coincide with the worker's actual productivity. (The firm can also pre-commit to providing employment protection as long as bonus pay is in the picture.) This provides both workers and firms with well-aligned incentives, and ensures that firms are financially viable and can employ more workers.
Your nepotism proposal can actually be improved in a fairly straightforward way; the firm can offer a severance payment to the worker in order to induce him to quit; the worker transfers the severance payment to his children, so that they may improve their job match (e.g. by asking for slightly lower wages, or extending job search).
Posted by: anon | January 25, 2012 at 01:48 AM
anon
Yes. But I think firms wouldn't do it. They want the option of sacking the worker in a downturn and the option of bringing in an outsider with newer skills from outside. What you described WAS the implicit contract situtation in the 1970s. Then starting sometime in the 80s it broke down from both sides.
Posted by: reason | January 25, 2012 at 04:50 AM
Frances,
can't you make a similar argument for seniority?
Posted by: reason | January 25, 2012 at 05:07 AM
anon: "just let the bonus pay expire and scale the wage back to coincide with the worker's actual productivity"
In a sense, that's what a standard retirement age does - people can still keep on working, but at the spot rate, i.e. they get what they could earn in the market, not the contract rate.
From a theoretical point of view (and this is a largely theoretical exercise), the question with bonus pay is: what would you tie the bonus to? Deferred compensation tends to arise in jobs where productivity is hard to measure - teams, particularly ones producing intangible goods like "knowledge." If the bonus goes to everyone on the team - which it would if everyone makes those firm specific investments - does it really make sense to call it a bonus any more?
reason - "can't you make a similar argument for seniority?"
Perhaps. Another argument for seniority is that it lessens insiders' resistance to change. E.g. about 15 or so years ago Carleton replaced its (well paid, unionized) maintenance staff with contract workers. But they didn't fire the unionized staff. Instead they just didn't replace them. Gradually, as the unionized workforce shrunk, one building after another was transferred over to the contract employees. The unionized workers didn't like it, but there wasn't, say, violent opposition, as there might have been if people's own jobs were directly on the line.
Of course you can't discount the position that this is just insider rent-seeking with absolutely no good economic justification.
Posted by: Frances Woolley | January 25, 2012 at 08:15 AM
I think you meant remuneration. It's odd that most people type renumeration, despite it being a rather archaic word.
Posted by: Andrew F | January 25, 2012 at 08:51 AM
Andrew F - I had to read your comment about three times before I could even see your point, I just kept on reading remuneration as renumeration.
Thank you.
Posted by: Frances Woolley | January 25, 2012 at 08:59 AM
As a practical matter, nepotism isn't possible in most firms for most employees. Where I work it'd get you canned. Of course, all that does is put it out of reach of the mere mortals. The executives and board members just trade in it. e.g. "Nice put Fred. By the way, my CFO tells me your boy is working out just great. Hey, my daughter is graduating in May with a psychology degree and she'd really like to get into an HR role. How about you hook her up with your HR people?"
Posted by: Patrick | January 25, 2012 at 11:14 AM
Imagine an insider-outsider model in a firm with high fixed capital that creates a "hold-up" problem. The firm is scared to invest, because the insiders will capture all the rents, and the firm slowly declines to zero just as the last insider retires and the last machine wears out. If the insiders lived forever, they would want to make sure the owners of the firm had an incentive to invest in new machines. Nepotism could solve that problem, by making the insiders live forever. Think the Big Three car firms maybe?
Posted by: Nick Rowe | January 25, 2012 at 12:02 PM
I should have apologized for the nit-pick in advance. I make that particular mistake too, so out of habit I check the spelling carefully.
Posted by: Andrew F | January 25, 2012 at 03:40 PM
Nick, that's a really excellent point - it's what I was getting at in terms of the worker having an interest in the long-term future of the firm, but more precisely expressed.
The thing is, of course, in the real world, I'm not going to be replaced by a full-time professor, related to me or otherwise, I'm going to be replaced with a video-link and a bunch of terminals with standardized exam questions (see, e.g. Larry Summers' recent opinion piece on the future of universities in the NY Times, talking about the amazing potential of video-links to give students access to the best professors).
And knowing that, I would still oppose technological change, because there isn't a way of getting my kids in.
Which is why the argument made in this post is just a house of cards - everything is very finely balanced and looks highly impressive, but just one cold hard blast of reality, and it comes tumbling down.
I'm slightly surprised that everyone has been too polite to say so. But then this is a famously civil blog.
Andrew F - no need to apologize, I'd much rather be corrected than have my poor spelling displayed on the internet for all time. And if I want to hide the evidence, I can always delete the incriminating comments.
Posted by: Frances Woolley | January 25, 2012 at 04:50 PM
Does this analysis include pensions? Specifically defined-benefit pensions? If so then it is leading straight into the "deferred wages" theory of pensions, which is FLAT WRONG as a theory. DB pensions are life annuities. In Canada all provincial pension benefits acts state that benefits to current beneficiaries cannot be reduced and the sponsor is on the hook to make up the difference (except if you are in a Multi-Employer Plan).
You cannot turn a series of deferred wage payments (a stock) into a flow of income for life without significant extra steps to transform the stock into a flow.
Your student was entirely right Frances though not for the reasons you suppose. Most people don't have DB pensions. Especially if they are employed in the private sector. Micro economists really, really have to get away from the DB model of pensions if they are going to analyze micro behaviour correctly.
Posted by: Determinant | January 25, 2012 at 09:05 PM
Determinant: "Most people don't have DB pensions. Especially if they are employed in the private sector."
Ironically, it's the people with the least income who have the highest replacement rate under the current system of CPP (QPP)/OAS/GIS. Even if they continue to work part time - as some do - hitting 65 is great financially.
Posted by: Frances Woolley | January 25, 2012 at 09:46 PM
Nick, shouldn't "insiders" get ownership of the firm? If the insiders have a stake in the firm, the value of their stake will depend on the firm's long-term viability, so incentives are correctly aligned again. (Alternately, the parties might pre-commit to avoiding hold-up problems as part of the contract. But this might be unfeasible due to imperfect enforcement.)
Posted by: anon | January 26, 2012 at 12:03 AM
Anon - yes fully agree. That is the logical solution.
Posted by: reason | January 26, 2012 at 03:39 AM
Nick, you are ignoring that there are two situations for nepotism - the 'nepotizer' owns the firm, or doesn't. In the first case, the nepotizer is choosing to divert part of the money flow to a relative. In the second case, the nepotizer is (in a minor or major way) looting the firm, taking something from the owners.
IMHO, you approach this from a viewpoint where looting (major or minor) is not an obvious explanation for things.
Posted by: Barry | January 26, 2012 at 08:22 AM
Barry, you're responding to Nick's comment, right, not the post in general, which you can't blame him for?
Posted by: Frances Woolley | January 26, 2012 at 08:53 AM
Barry: "you approach this from a viewpoint where looting (major or minor) is not an obvious explanation for things."
I don't think of it so much as looting as "feeding from trough". The old pig understandably doesn't want to give up his or her place at the trough, so is induced to do so by the opportunity to pass the place onto his or her offspring.
Hereditary looting rights mean that old and inefficient looters are replaced by young and vigorous looters. Which might possibly be a good thing.
Posted by: Frances Woolley | January 26, 2012 at 10:32 AM
"...there are no good jobs out there. Why? Because the geezers have them."
Because you assume that younger people are in general more productive for a given wage rate, there is a profit opportunity for a firm to hire them all up. Sure, existing firms might be constrained because they are paying for all the old workers, but new firms could just start the cycle from scratch. It's a little "lump of labor"y.
Posted by: James Oswald | January 26, 2012 at 12:37 PM
That assumption depends on if new firms can get enough capital to start up and thrive as incumbents exercise their market power. This is not a sure thing by any means. Startup capital is hard to come by and 90% of startups fail.
Posted by: Determinant | January 26, 2012 at 03:31 PM
I'm amused by Determinant's defined benefit pension comment. At least in the States, most of us cosseted academics have, and have always had, define-contribution pensions (TIAA-CREF). So it really is a deferred income situation; we accept taking that compensation as (untaxed) current pension contributions rather than as (taxed) current income.
Posted by: Donald A. Coffin | January 27, 2012 at 12:26 AM
"are replaced by young and vigorous looters. Which might possibly be a good thing."
Well at least for the young and vigorous looters.
Posted by: reason | January 27, 2012 at 04:03 AM
Canada is different, Donald.
"Pension" without qualification means Defined Benefit Pension in Canada. Curiously Canadians are notoriously reticent to purchase life annuities themselves or with DC plans but think DB pensions are wonderful.
Speaking of retirement, can we expect a thread on the Government's announcement of raising the OAS age to 67? If not this blog will need to turn in its Maple Leaf.
Posted by: Determinant | January 27, 2012 at 03:26 PM
On the other hand, this study (http://www.bis.org/publ/wgpapers/cgfs27broadbent3.pdf) suggests that participation in DB plans in Canada is decreasing, while participation in DC plans is increasing. It remains the canse that Canada has one of the lowest rates of participationin DC plans among OECD countries.
Posted by: Donald A. Coffin | January 27, 2012 at 03:50 PM
Determinant: "Speaking of retirement, can we expect a thread on the Government's announcement of raising the OAS age to 67? If not this blog will need to turn in its Maple Leaf."
I was actually thinking of doing a blog post on that sort of subject. Demographics - we are now past the peak of the % of population of working age (Stephen's graphs) -- higher elasticity of labour supply of older workers nearing retirement/extensive margin -- it makes sense to have lower marginal tax rates on workers with more elastic labour supply -- wondering what the OAS did to marginal effective tax rates -- burden of the debt -- not just OAS but healthcare liabilities -- etc. But realised it's a big subject on a lot of areas I don't know much about. Googling showed me my ignorance.
Maybe an open post on the subject where we can all chip in bits and pieces?
Posted by: Nick Rowe | January 27, 2012 at 04:23 PM
Determinant: DB pensions are life annuities. You extend credit to your employer by not cashing your wages. The annuity is the mean to deliver the deferred wage.
Posted by: Jacques René Giguère | January 27, 2012 at 04:55 PM
Not under the actual implementation system in Canada.
The annuitization part of DB pensions is what gets companies. Provincial Pension Benefits Acts state that the employer is responsible to guarantee payment of benefits earned. Earned benefits cannot be reduced. So the employer takes longevity and market risk. What amount of funding is available to annuitize is subject to these risks. It is very uncertain. Ignore this paragraph is you are a member of a Multi-Employer Pension Plan. There is no employer guarantee for these plans, at least in Ontario.
Further, and most crucially, DB pension plans do not purchase life annuities from life insurers, even group life annuity policies. They annuitize themselves. This is a very large and little-known problem with our pension system. Group life annuity policies do exist in Canada but they are rarely used. The pension industry views them as expensive even though they serve to cede the risk to life insurer.
It is fundamentally wrong to ask a manufacturing company to also be a life annuity company. Experience has demonstrated that they are very bad at this. Life annuities should be left to life insurance companies. Sponsors make optimistic assumptions about returns and longevity and further run the risk of bankruptcy. Only Ontario guarantees DB pensions (but not MEPP's) through a government pension insurer. Other provinces and the Federal jurisdiction don't. If your employer goes bankrupt, you are out of luck.
Classically the pension industry assumed that corporate sponsors have longer time horizons than employees. That is empirically not true.
Lastly life annuities from commerical life insurers are much, much sounder financially than DB pensions. Life Annuities are guaranteed by Assuris, the life insurance counterpart to CDIC. Assuris guarantees 100% payment of life annuity obligations up to $60,000 and you can only take a maximum loss of 15% on income above that. Defaulted policies are not refunded but are transferred to a solvent company. The life insurance industry is based on trust and therefore has a strong commercial interest in maintaining solvency and payment on annuities. Pension sponsors don't have this interest.
The Deferred Wages model utterly fails to capture the complexity of delivering a life annuity and further fails to acknowledge longevity and market risk. It assumed certainty of pension funding based on contributions and further assumes a fixed and known pool of capital at retirement, both of which are clearly untrue based on market and longevity risk. It ignores the very problems that affect DB pensions today.
Posted by: Determinant | January 27, 2012 at 05:22 PM
Spam filter? Topical post on pensions, rebutting Jacques' post?
Posted by: Determinant | January 27, 2012 at 05:29 PM
Donald: "while participation in DC plans is increasing" - I suspect that result would be extremely sensitive to what one considered participation in a DC plan - lots of people have some fairly small amount set aside in an RRSP - $25,000 or $50,000 - amount that are not really meaningful in terms of providing retirement income, i.e. people would be better cashing them out, buying a nice durable asset that will yield a string of consumption benefits or under-the-table-income (e.g. a cottage that can be rented out for cash), and relying on CPP/OAS/GIS instead.
Which takes us to raising the retirement age.
Posted by: Frances Woolley | January 27, 2012 at 10:38 PM
On raising the retirement age - from June, 2010, http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/06/should-recent-immigrants-be-eligible-for-old-age-security.html
"Moreover, the idea that someone, just because he or she has reached the age of 65, is entitled to special income guarantees, is hard to justify now that most provinces have no standard retirement age. The reason that 17 year olds are not entitled to an adequate guaranteed income is that they are expected to be in work, in school, or supported by their parents. If people over 65 are perfectly capable of working - as those in favour of ending mandatory retirement seem to imply - why should seniors automatically be eligible for special income support programs? Yes, many over 65 are disabled and unable to work - but so are many under 65. Why an age test?"
Posted by: Frances Woolley | January 27, 2012 at 10:42 PM
Determinant: I am not arguing the techy details, crucial though they are. For the average Joe, the effect of DBP is getting a guaranteed deferred wage and its good enuf for gunmint work. Once again. sacrificing a black goat would be a good enough explanation of how things works ( Nick borrows apples, I sacrifice goats...)
And no steel or plastics company should ever,ever manage an insurance plan.
All in all the best pension plan should be simply a public pension based on real GDP, something way more stable than profits,interest rates, divident yields and stock and bond prices.
You could require a contribution to be invested in real assets ,a la QPP-Caisse de Depot, Singapore or CPP Investment Board. And by real assets ,I also includes bonds of well functionning governments( anyone to the west side of the 1054 schism...except Argentina)
And I will let Nick start one month of thread on the last paragraph.
Posted by: Jacques René Giguère | January 27, 2012 at 10:48 PM
I had to Google "1054 schism" (the split between the Greek/Orthodox and Latin/Catholic parts of christianity.)
Frances: I was trying to figure out effective marginal tax rates for old folks. They look very high. The clawback on OAS/GIS alone seems to give a marginal tax rate of over 50%. If you add income tax on top of that (and OAS is taxable) old folks with lower incomes must be facing EMTRs close to 100%. (I vaguely remember you telling me something about this at one time). Which doesn't seem to make sense, if those on the margin of retirement should have much more elastic labour supply functions than everyone else.
Hmmm. Maybe Stephen Harper does need to do something here.
Posted by: Nick Rowe | January 28, 2012 at 06:44 AM
Nick - yes, you're right on the high effective marginal tax rates for old folks. It doesn't usually get to 100% because with the various age and pension amounts it's pretty unusual to be in a situation where you're paying income tax and collecting GIS, especially for a couple. But the high MTRs mean that a small RRSP can be a really lousy investment choice - TFSAs are better.
The question is not whether all these things are going to be rolled back, but when. As I noted a while ago, my sister has a theory: every policy initiative is always introduced either too late or too early for her to benefit from it. Which would suggest the serious roll-backs are coming in just before 2025.
Yes, actually, a gradual lifting of the retirement age with a 10 year phase in period is just about what I would guess to be the most likely policy proposal, so that would put the date at 2022.
Posted by: Frances Woolley | January 28, 2012 at 08:37 AM
Um, Pension Income is subject to the Pension Income Tax Credit, which effectively raises the tax-free threshold from $10,000 with the Basic Personal Exemption to $17,000.
Jacques:
No, it's not good enough for gummit work, that's my point. I truly respect you but this model has real problems. Easy ones to explain too.
1) Pensions are insurance.
2) You can't get money from a dead man, e.g. a bankrupt company.
3) You have a real chance of outliving your employer.
3b) Contrary to the pension industry's received wisdom, people have longer time horizons than companies do.
4) Making guarantees about the future is really, really hard. For hard, read expensive. Pensions are generally narrowly balanced, if something changes just slightly, the whole result goes out the window.
The Deferred Wages model should be taken out to the woods and put out of its misery.
Posted by: Determinant | January 28, 2012 at 06:16 PM
I understand your point but we also must kill the idea that a pension is a gift from the employer. Or that teachers and other public sector workers should not get theirs because a lot of taxpayers earn less than they, notwhitstanding that their doctors or the engineers who build their road also earn more. And the political argument must be deferred wages.
I have a dual career : professionnal economist and politician-union leader. I learned how to manage a split personnality and to use contradictory arguments within 5 seconds without skipping a beat.
( A few years ago, my union won a court case for a professor who was found physically unable to teach by his college but still able to do so by the disability employer). The judge found that you could argue simultaneously that you are fit enough to work and unfit enough so that you need to collect insurance.
Salus populi suprema lex esto...do whats needed to do good.
Posted by: Jacques René Giguère | January 29, 2012 at 12:01 AM
Sure, it's called partial disability. It's a very standard definition in the disability insurance industry. Topical to my post, Employer-provided disability policies are written with the employer in mind, not the worker. The standard industry contract provides two years of "regular occupation" coverage where you are disabled if you cannot perform the duties of your last or regular occupation.
After two years the definition switches to a more stringent definition, Any Occupation which means you are disabled if you cannot work at any occupation you are qualified for by reason of education, training or experience. Your friend likely got caught on this part.
The two-definition model is designed with employers in mind to encourage employees to return to work after two years. It also keeps costs low, but it doesn't serve employees well. Remember it is employers who buy group policies. I have sold policies like this.
Most people have a mental image of Regular Occupational coverage until 65. That is widely available on the private market and it is very common to structure a personal policy to supplement group coverage, but you have to know the problem and then go to the market to address it. It is one of the largest latent financial problems in Canada today.
I never said that pensions are a gift. They are savings combined with insurance. But to expect employers to deliver on those savings and to provide that insurance is problematic. Experience has shown that they don't have sufficient economic incentives to actually deliver and can frequently have perverse incentives to renege on their deals. I want the perverse incentives eliminated. I want safety for people who depend on these policies.
And you still can't get money from a dead man, er, a bankrupt company.
Employers need to be eliminated from the equation as the depository and guarantor for savings and provider of insurance. The pension industry needs to become a separate fifth pillar of the financial system with its own protections. Life Insurers which provide annuities have a much greater incentive to honour those annuities. It is a safer, more predictable product for everyone.
If it costs more then that's just realizing that the actual costs of our promises are more than we think they are. The price of life annuities is the real price information point. Frances likes to talk about prices-as-information as a key point of micro.
Posted by: Determinant | January 29, 2012 at 03:54 PM
Determinant is stuck in the spam filter!
Posted by: Nick Rowe | January 30, 2012 at 03:24 AM