Re: AltInst: a "work spreading tax"

From: George L. O'Brien <>
Date: Sun May 31 1998 - 11:44:56 PDT

Tom Walker's discussion of the costing of "training" is interesting, but
I remain rather skeptical. The studies I've read suggest that if anything,
training costs are generally underestimated due to the poor tracking of
curve costs. In a great number of situations, by the time the new hire
fully productive, the person either leaves or gets transferred to a new job.

This turns training costs into a probability problem. What is the
likelyhood that
this person will work long enough at full capacity at this job to repay the
costs of the trainer, added supervision costs, the below optimal
productivity, and higher defect rate of the employee prior to coming up to
speed? If we assume the learning curve is a function of the number of
hours worked on the related tasks, if someone works only 20 hours a week,
then we should assume it will take roughly twice as long to come up to
speed while doubling the probability that the person will leave prior to
reaching payback on the investment.

By contrast, the cost of working existing employees overtime is easy to
calculate as long as it does not result in increased turnover.

George L. O'Brien

At 12:53 PM 5/30/98 -0700, you wrote:
>George L. O'Brien (Certified Management Accountant)
>>Currently the fixed costs of hiring an employee in the US are heavily
>>weighted by the way various payroll taxes work. A number of taxes such
>>unemployement and state disability have relatively low cutoff points, so
>>extra hours are not included. More importantly, most medical insurance
>>plans are the same once the employee qualifies (usually over 20 hours a
>At the risk of repeating myself, I couldn't agree more with your analysis.
>That's why I have campaigned for several years for "levelling the playing
>field" of fixed cost statutory contributions, especially by raising the
>ceilings for contributions to social insurance plans.
>But a funny thing happens when you chart the regressive, fixed payroll costs
>together with progressive income taxes, using annual income as the scale on
>the x axis: the trend lines of the two roughly cancel each other out so that
>the per hour total contribution is roughly constant. I haven't done the
>spreadsheet calculations for U.S. rates, but the major differences would be
>in higher private insurance health premiums (adding regressivity) and in the
>much higher annual contributions ceiling on social security (adding
>progressivity). U.S. income tax rates are substantially lower than Canadian
>rates, but I would imagine that the progressivity is comparable, perhaps
>with some degree of unevenness introduced by the earned income tax credits.
>Given those substantially flat per hour costs, the escalating costs of pay
>increments for years of service create a situation where a more senior work
>force (on average) is more expensive than a more junior work force. One
>might object that the higher cost of more senior workers is justified by
>their higher productivity. I would respond that such "higher productivity"
>depends entirely on how well their skills and experience are being utilized
>by the division of labour.
>For example, think of an engineering firm that decides to cut costs by
>laying off the receptionist. The savings could easily be overwhelmed by the
>cost of the highly paid engineers having to answer their unscreened
>telephone calls and direct traffic for people dropping into the office.
>I've also noticed a tendency in the literature to systematically overstate
>non-payroll fixed costs, such as hiring and training costs, by ignoring
>economies of scale and productivity gains. If it costs $10,000 to hire one
>new employee, does that mean it costs $50,000 to hire five? Not likely.
>The way that I've seen training costs accounted for is also highly suspect.
>Here's the way it's done in a pulp industry report on employment costs: say
>a new employee requires 100 hours of on the job training, this will require
>a senior employee to spend 100 hours training the new recruit. If the cost
>per hour for the senior worker is $40, the total training cost is estimated
>to be $4,000.
>But let's say that senior worker spends two hours out of an eight hour day
>training the new recruit. What happens to the senior worker's productivity
>during the *remaining six hours*? It's entirely possible that part of the
>hypothetical training "cost" is absorbed by slack time rather than
>productive time, so it doesn't really cost the employer anything.
>Furthermore, what about the training effect on the senior worker? The best
>way to learn something new about your specialty is to teach it to someone
>To treat training time as a charge against production, with no side
>benefits, is to describe a worst-case scenario. Worst-case scenarios invite
>hoarding, which leads to shortages, bottle necks and dead losses. But why be
>verbose? There's an old maxim, 'penny wise and pound foolish' that describes
>the miser's dilemma.
>It would be ironic if economists thought they could fight fallacy with
>fallacy. But I think that is *exactly* what has happened. The evidence for
>the charge of 'lump of labour fallacy' -- routinely levelled against any
>proposal to redistribute work time -- is itself replete with unexamined
>assumptions, anachronisms and worst-case scenarios. It has become a 'lump of
>cost fallacy' that, for some inexplicable reason, is exempt from criticism.
>Tom Walker
>#408 1035 Pacific St.
>Vancouver, B.C.
>V6E 4G7
>(604) 669-3286
>The TimeWork Web:
>[To drop AltInst, tell: to: unsubscribe altinst]

[To drop AltInst, tell: to: unsubscribe altinst]
Received on Sun May 31 12:03:00 1998

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:49:12 PST