True Unemployment Numbers

The current official unemployment figure (9.5% for June) should be replaced (or at least supplemented) by a better measurement of actual unemployment. Unemployment should be measured in terms of hours worked by all employees, full-time and part-time. Counting full- and part-time workers equally is giving misleading statistics. Hours per employee, which are changing over time, are not properly recognized in unemployment as currently reported.

Part-time employment for economic reasons - U.S. Dept. of Labor (DOL) terminology - is not directly recognized in the official unemployment rate (U-3). If a person who is working half-time wants to work full-time, but cannot find full-time work, this is the equivalent of half a job lost. The same is true for an employee being forced by an employer to cut half his hours.

The official DOL unemployment rate counts either individual equally with full-time employees in measuring the level of employment. This was discussed in recent articles (here and here) and will be analyzed in detail in this article. John Mauldin has also discussed this problem is some detail recently (here).

Involuntary part-time employment increases significantly in economic downturns, whereas part-time by choice (“part-time for non-economic reasons” in DOL terminology) is much less changed from its trend roughly proportional to the increase in the workforce with time. This is shown in the following two graphs (from DOL BLS Table A-5).

Note that there is a big jump in the non-economic graph from December, 1993 to January, 2004, presumably from a change in the measurement rules. There is a drop of the same magnitude for employed part-time for economic reasons at the same time. Apparently, at that time, some of the people who formerly would have been counted as involuntary part-time were transferred to part-time by choice.

There are a number of areas that must be examined to try to determine how involuntary part-time employment should be included in what I have chosen to call “implied total unemployment”.

Among these are:

1. the definition of full-time employment;
2. determination of the average work hours in a part-time week;
3. calculation of the implied total unemployment when data is adjusted so that the full-time equivalent number of employed is used; and
4. possible changes in the length of the work week over the past decades. These will be discussed in the following sections.

What Is Full-Time Employment?

The DOL does not define the length of a full-time work week. They state that is the domain of the employer. It is clear that a full work week must be at least 35 hours for the DOL statistics because all employment less than 35 hours is classified as part-time. The traditional full-time work week used to be considered to be 40 hours, but that may not be the actual practice today when averaged over all employers. The DOL determination of the average work week since 1964 (including part-timers) is shown in the following graph.

There was a dramatic drop in the average work week from 1964 to 1984. The decline for the past 28 years was less severe, but minima reached during recessions never recovered to pre-recession levels during this period.

The reason for this drop in average weekly hours worked may be because

1. the 40 hour standard was degraded (say from 8:00 – 5:00 with two 15 minute coffee breaks and a 30 minutes lunch to maybe an extra 30 minutes for lunch, a 37.5 hour week); or
2. the average part-time weekly hours was decreased (say, for example, from an average of 3 days per week - 24 hours per week - to an average of 2 1/2 days per week – 20 hours per week).

Of course, there may have some of both of these factors operative. Possible interplay between these two factors will be examined in the next section.

Average Hours in a Part-Time Week

The average weekly hours in a part-time week are calculated by the equations shown in Appendix 1 at the end of the article.

The following table shows the relationship between assumed values for average hours in a full-time week and calculated values for average hours in a part-time week. The DOL employment data for May, 2009 was used.

It is clear that the average full-time work week in May, 2009 is not 40 hours, or even 39 because of the calculated values for part-time work week hours. I say this because an average of 8.4 hours or less for part-time work seems unlikely, but I’ll reserve final judgment on that for later. For now let’s assume it is likely that the current average part-time work week is probably between 16 and 24 hours, or the equivalent of 2 to 3 days per week. Those part-time numbers can be valid only for average weekly hours for full-time employees between 35 and 37. When the economy is operating on all cylinders, the average full-time work week may be much closer to 40 hours than it is now.

What Is the Current Implied Total Unemployment?

The current (May) official unemployment rate is 9.4%, in which each part-timer is counted the same as a fully employed person. This is the so-called U-3 rate. The DOL also reports a broader unemployment rate, named the U-6 rate, which was 16.4% for May. U-6 differs from U-3 by using an expanded definition of the labor force to include the so-called marginally attached, as well as counting the marginally attached and part-time for economic reasons as unemployed.

The relationships used by the DOL for calculating U-6 are given in Appendix 2 at the end of the article.

I propose that a better measurement of employment would be with respect to a full work week equaling one equivalent employed person. This methodology could count a part-timer working reduced hours by choice in either of two ways:

1. equal to full-time employment or
2. equal to a fraction of full employment.

We will compare the impact on the magnitude of the implied unemployment rate using each definition. In all cases, the impact of an involuntary part-time worker will be a contribution to employment only as the ratio of the average part-time hours to the full-time hours.

The two different ways of calculating the equivalent full-time employment will be called Case A and Case B. Case A considers all part-timers as fractionally employed; e.g., 10 hours per week constitutes 0.25 of an employee, 20 hours per week is 0.50 employee and 30 hours per week is 0.75 of one employee. These fractions apply, of course, for a 40 hour full-time work week. The fractions would increase proportionately for shorter full-time work weeks.

Case B considers all voluntary part-time work the equivalent of full-time individuals in counting employees. In this case, only the part-time for economic reasons (involuntary) part-timers are counted fractionally.

The numbers for the various assumptions for full-time work and the two cases are shown in the following table.

The corresponding unemployment rates, compared to the official U-3 rate are shown in the following table.

Next, the unemployment rates calculated when the marginally attached are added to the labor force are shown in the following table. In this situation, the appropriate DOL rate to compare is U-6.

Finally, if we include all the individuals not in the labor force who are classified as “want work now”, the following table shows the calculated implied unemployment levels. The marginally attached (2,176 in May) are included in the group classified by the as “want work now” (6,454 in May). In this case also, we provide the U-6 rate for comparison.

I may be old fashioned, but I think the official work week should be defined as 40 hours. This means that, using Case B data, the implied total unemployment corresponding to the official U-3 rate is 14.3% (blue table) compared to the DOL rate for May of 9.4%.

The agreement between the corresponding implied total unemployment and U-6 is considerably better for the two expanded labor force definitions, 15.5% when the marginally attached are included (green table) and 17.8% when all who “want work now” are included (pink table). These two extended labor force calculations bracket the DOL U-6 value (16.4%).

Why not use Case A? I think it is a possible measure of potential total slack in the labor force, but is not a good measure of currently available slack. After all, it is reasonable to assume that most of the people working part-time for non-economic reasons are working as much as they are available or want to work.

However, saying that they are not looking for full-time work is not the same as saying they would not accept more hours, so there may be potential slack in the labor force from people working part-time voluntarily. Perhaps the DOL interview procedures would have to be changed before we would get some information about this potential slack in the labor force.

Historical Perspectives

What would have been the behavior of implied total unemployment during the past 40 years? Before we look at that, let’s review the three widely recognized unemployment rates. The following graph from ShadowStats.com shows how the three measurements have functioned since 1994. Note that there is little difference in the movement of the three metrics; they maintain very closely correlated movement throughout the business cycle.

In the next graph, the implied total unemployment and the official U-3 rates are plotted. The two curves show generally correlated movements, but you may detect variations if you look closely.

Implied Total Unemployment, Blue; Official Unemployment (U-3), Red

The difference between the value of implied total unemployment and U-3 is plotted in the following graph. Lower correlations show up as steeply sloped regions. Since 1980, most of these steeply upward sloping areas have been associated with recessions. This can be attributed to the better accounting of increased involuntary part-time employment by implied total unemployment during economic contractions.

Another factor evident in the above graph is the establishment of a new equilibrium difference between implied total unemployment and U-3 of about 3.5% during periods between recessions. The difference dipped temporarily below 3% during the internet bubble in the late 1990’s. The equilibrium difference moved up during the 1970’s from a level around 1.0% during the 1960’s.

Thus implied total unemployment is showing the impact on work during recessions (shorter work hours, whether from increased involuntary part-time employment or from reduced hours for individuals still classified as full-time). It is also showing the effects of the change in employment practices during the rest of the business cycle. Even during expansions, employers have been using increased numbers of reduced hour workers (involuntarily reduced from the employee point of view) since 1980 compared to the 1960’s.

All of these observations can be demonstrated by doing detailed analysis of hours worked and distribution of the employed among full-time and part-time employment. The advantage of implied total unemployment is that the effect of varying employment practices is shown in a single metric in an intuitively obvious way. This is discussed further in the next section.

Why Is Implied Total Unemployment Important?

Implied total unemployment is a better measure of slack in the labor force than the generally accepted U-3 number. In an economic contraction, employers not only furlough and fire employees, they also cut back on hours for employees retained. This is not reflected in U-3.

When an economy starts to recover, an employer will increase the number of man-hours worked as demand develops for his products and services. If he has part-timers, the employer is likely to add hours to their schedules before adding additional employees to the payroll. If there is any uncertainty, he will even add overtime before new hires. This has been discussed recently in a good article by Jeffrey Fraenkel (here). Again, U-3 totally misses these labor market changes.

Why not replace U-3 with a better measurement of labor force activity? This process could be started by adding another measurement classification to the analysis currently reported. Maybe call it U-7? Usage of the DOL data in the future will prove whether or not my assertion of value for U-7 is valid or not. If useful in future economic analysis, I will change my nomenclature from “implied total unemployment” to “real total unemployment”.

Appendices

In Appendix 1, basic definitions are given for variables in this analysis. For reference, how the DOL calculates U-6 is described in Appendix 2. In Appendix 3 and 4, the variables and the equations used for calculating U-7, as well as U-7(A) are defined. The U-7 and U-7(A) numbers are in the blue table in the article.

I have also included definitions for two expanded labor force unemployment measures, U-8 and U-9, corresponding to the green and pink tables in the article.

Do not look for U-7, U-8 and U-9 in the DOL data bases. These are my inventions and may never be implemented in the official records of the DOL.

Appendix 1. Basic definitions

Appendix 2. DOL methodology for calculating U-6.

Appendix 3. Data and variables for calculating implied total unemployment.

Appendix 4. Equations for calculating implied total unemployment. Variables are defined in Appendix 3.

Comments

[Steven Hansen:]

John, this analysis blows me away.

i hope people at the dol and bls are passing around this article right now.

Jul 06 03:11 AM

[Moon Kil Woong:]

It is funny to read this. Your analysis is much more detailed and logical than the DOL or other government bureau releases about unemployment or the even more important labor market slack (weven though everyone knows this is fundamentally important to consider since a lack of labor market slack contributes heavily to wage inflation). It is also more reasoned than the hours of testimony the Fed and the treasury feed to Congress regularly. Of course, elected officials tend not to ask such questions anyway (they may upset the boat).

It is fundamentally important to know how many labor hours are worked and how many labor hours there are available to work. The fact DOL and other reports onscure this most important fact is annoying. I just hope they do this reasearch and choose to keep it secret. Otherwise, our economic leaders are as blind as a bat.

Jul 06 03:43 AM

[nobby73:]

This is excellent analysis which is consistent with the data on payrolls, tax receipts and welfare payments, suggesting it is the better quality, full-time jobs that are being lost.

Napoleon famously decried Britain as a nation of shopkeepers, maybe today he would say America is a nation of fast food workers.

Jul 06 04:07 AM

[markfl:]

"When an economy starts to recover, an employer will increase the number of man-hours worked as demand develops for his products and services. If he has part-timers, the employer is likely to add hours to their schedules before adding additional employees to the payroll. If there is any uncertainty, he will even add overtime before new hires. This has been discussed recently in a good article by Jeffrey Fraenkel (here). Again, U-3 totally misses these labor market changes."

I'm sure you'll be please to know the Cleveland Fed essentially agrees with you.
www.clevelandfed.org/r...

The number of non-farm part-timers due to slack work (at their existing "full-time") job has increased since Feb. 2009. That, combined with a decrease in avg. weekly hours to 33.0 suggests GDP forecasts may still be too optimistic. "The data 'strongly suggest that consensus forecast for a second half recovery is overly optimistic,' said Steve Ricchiuto, chief economist at Mizuho Securities in New York."
www.standard.net/live/...

Jul 06 05:02 AM

[DONE_SONZ:]

-involuntary part-time employment- I guess that wouldn't matter if we only had to pay 50% of our bills if all we can get is part time.Not the case ofcourse.People that are getting refused unemployment,or who have used up their amount, and people that have given up on looking for work are not factored in either.The % of unemployed would alot higher if all these numbers were factored in.Even counting part time as half.

Jul 06 05:39 AM

[Dave Wrixon:]

Yes, but don't you think it is Government policy to make things as opaque as possible? What possible motive would they have to make it clearly understood by the masses?


On Jul 06 03:43 AM Moon Kil Woong wrote:

> It is funny to read this. Your analysis is much more detailed and
> logical than the DOL or other government bureau releases about unemployment
> or the even more important labor market slack (weven though everyone
> knows this is fundamentally important to consider since a lack of
> labor market slack contributes heavily to wage inflation). It is
> also more reasoned than the hours of testimony the Fed and the treasury
> feed to Congress regularly. Of course, elected officials tend not
> to ask such questions anyway (they may upset the boat).
>
> It is fundamentally important to know how many labor hours are worked
> and how many labor hours there are available to work. The fact DOL
> and other reports onscure this most important fact is annoying. I
> just hope they do this reasearch and choose to keep it secret. Otherwise,
> our economic leaders are as blind as a bat.

Jul 06 06:06 AM

[Dan in mpls:]

While The information is clear and makes a compeling arguement for rethinking unemployment, using the tradition (and hopeully unchanging) formulas for common consumption has the huge benefit of allowing us to roughly compare to current figures with those, say, of the 1930's. The author mentioned a change in the way the numbers are figured in 93-94. I have been curious as to how that reworking of the numbers have changed the comparisons we use now and say, the 80's.

Jul 06 07:23 AM

[BlueOkie:]

More and more people are doing 1099 work or contract work. Gov't regulation is making it harder for people to find work. What effect did the recent minimum wage increases have on unemployment? We reward people for not working with unemployment benefits, welfare etc. The gov't cannot find 11 million illegals, how can good can these numbers be?

Jul 06 07:38 AM

[FI:]

great!

Jul 06 07:43 AM

[John Galt:]

All of the statistics and analysis remind me of baseball.

You will usually hear somebody ask " what's that players batting average", as people don't usually ask what a players on base percentage is but it's really a better stat rigth? One counts hits, one counts hits + walks to see how often a player gets on base. Well isn't a walk really as good as a single? Often times a walk is even better because it forces the pitcher to throw at least 4 pitches ( usually more), where as a single might result in only 1 pitch. More pitches thrown = tired pitchers and tired pitchers aren't as good. There is a whole "sabermetrics" revolution in baseball.

When times were good most people didn't dig as deep as Mr. Lounsbury did in this article but are we to the point where the DOL can start reporting more accurate statistics? Do they even want to try and best report the truth?

Jul 06 08:53 AM

[kruser53:]

Wow! A real original research paper right here on SA. This goes way beyond the category of analysis, which many SA authors do so excellently. This is original thinking and provides a thoroughly solid foundation for future analysis.

Imagine! Defining the level of employment (and unemployment) by counting the man-hours worked instead of counting the number of people leaning on shovel handles. It doesn't seem an original idea, but since it appears to be a new definition maybe it is (an original idea). I believe that many have focussed on the shorter average work week and the increasing numbers of part-timers, but the author has put it all into one simple package. And the package makes perfect sense.

Jul 06 09:24 AM

[sleepless_o...:]

This article displays some great thinking outside the box. This shows what can be done with the data that is collected by Dept. of Labor if some original thinking is done.

Jul 06 10:06 AM

[robert.b.fe...:]

Team Obama has no reason to level with the public. That would make reallity clear for everyone to see. Things are not improving as they would have you believe. I have heard the argument that unemployment is a lagging indicator and that Clintons retooling of the calculation methods for economic metrics was an improvement. However there is a point at which unemployment is a current indicator especialy when the numbers go down for several months at a consistant rate as they have this quarter. There are no green shoots unless you believe that cap and trade actually will create green jobs. Good lick with that concept. Excellent piece of work Team Obama would be well served to read and heed this. Even with the improved Clinton calculations the real deal is bound to emerge sooner or later. Lets hope it's sooner.

Jul 06 10:49 AM

[ain't no fo...:]

That's why the government will never consider incorporating it into its "doctrine".


On Jul 06 09:24 AM kruser53 wrote:

> Wow! A real original research paper right here on SA. This goes
> way beyond the category of analysis, which many SA authors do so
> excellently. This is original thinking and provides a thoroughly
> solid foundation for future analysis.
>
> Imagine! Defining the level of employment (and unemployment) by
> counting the man-hours worked instead of counting the number of people
> leaning on shovel handles. It doesn't seem an original idea, but
> since it appears to be a new definition maybe it is (an original
> idea). I believe that many have focussed on the shorter average
> work week and the increasing numbers of part-timers, but the author
> has put it all into one simple package. And the package makes perfect
> sense.

Jul 06 11:29 AM

[2nynbak:]

John, you epitomize what is best about the Internet. Original thinking with real analysis at a price we can all afford! It's too bad the old media refuses to do the quality of work you are doing for the readers of Seeking Alpha.

Jul 06 11:59 AM

[Gregman2:]

Reminds me of Jeremy Rifkin's book, "The End of Work." I think many are doing uncompensated work as well. This is not properly measured.

Jul 06 12:23 PM

[Alan von Al...:]

Fine work, John. Now all we need to do is minus out Fed, state, local government direct employment and government contractors, because their paychecks and pensions are deducted from private sector taxpayers/investors/wo...

Hmm. 20% labor slack by your formula + 5% small business closed and self-employed 1099 terminated + 25% govt workers/contractors = 50% of US workforce producing nothing?

Jul 06 02:03 PM

[PainfullyAware:]

Big Props John Lounsbury - Excellent Research And Presentation !!!

As a corollary the Tax Revenues are plummeting because of the reduction in "Total Hours Worked". Income and Spending are being assailed. The unfortunate thing is that I hear very little about reducing the "Welfare/Warfare State" to match the new low income paradigm.

The Disconnect With DC Grows Ever More Apparent.

An interesting metric is the shift from "Employees" to "Sub-Contractors". I have been watching this trend for some time and I will say that it is Totally Driven By The Desire To Reduce The Bureaucratic Cost Mandates. "Rent Rather Than Own" is becoming the new business model when it comes to Labor.

Jul 06 02:12 PM

[Ferdinand E...:]

This author really knows how to write a paper. Makes me wonder if I could match his competence if I dealt with the same subject? Or if he could match mine if he were interested in energy economics?

Jul 06 02:45 PM

[Joel87:]

Great Analysis. Send a copy to the DOL, and hopefully after reading they will adopt this method.

Jul 06 02:58 PM

[Joel87:]

Great Analysis. Send a copy to the DOL, and hopefully after reading they will adopt this method.

Jul 06 02:58 PM

[Larrysyr:]

John,

Your excellent analysis raises many important questions.

However, I have one bone to pick.

You want to define a "full-time" job as 40 hours per week. This is not only old-fashioned, it is out of touch with current practice.

In my 30 years of full-time work experience, I have never had a 40-hour work schedule. My normal schedules have ranged from 35 to 37.5 hours. This slack allows employers to add some overtime while still paying straight-time wages.

Economic measurements do need to reflect reality. What constitutes full-time work these days is working enough hours to qualify for health insurance and "retirement" benefits. That threshold is less than 40 hours.

Jul 06 03:08 PM

[Living4Divi...:]

John - excellent article. keep up the good work!

Jul 06 05:11 PM

[John Lounsbury:]

Larrysyr - - -

Your concern is why I showed the effect of sellecting definitions of full-time at various defined levels, from 35 up to 40. I have made a few inquiries, both corporate and small businesses, and all consider the normal work schedule to be 40 hours. That is definitely anecdotal evidence and therefore does not in any way refute your contention that a shorter work week is quite widespead.

Whatever the definition of full-time work week hours, the process defined gives a better measurement of the actual labor market at times of economic stress.

Jul 06 05:23 PM

[Think-About-It:]

I'm curious about "private sector unemployment/productio...

I've read that total government expenditures now equal roughly 47% - 48% of U.S. GDP up handsomely from around 8% in 1900. Much of this expenditure must be for direct government (local, state & Fed) payrolls which ultimately is paid for by the private sector through taxation or inflation.

During this recession/depression it seams as though government workers are fairing much better on average at maintaining their jobs, hours, pay rates and retirement benefits than private sector workers. Would you care to guess or shed some light as to possibility of reporting unemployment related numbers on each group separately? Or something along these lines?

Jul 06 06:39 PM

[Brian Bober:]

Great analysis, it is dis-heartening how looking beyond the headline numbers continues to reveal a rotten situation.

I've have yet to read a good analysis that accounts for the loss of independent contract jobs like mortgage brokers, construction workers, etc. Any insight there?

Jul 06 06:46 PM

[bigbaboon57:]

Of course, government intentionally under-reports unemployment, both for political reasons and to keep their payouts down. Same reason they under-report inflation. Look for the under-reporting to continue and become even more extreme and the true numbers become worse.

Jul 06 06:57 PM

[Marginaltho...:]

Hi John,

Interesting article. The very first thing that firms do during a belt tightening cycle is to let go of all of the "independent contractors" on their books.

These IC's are considered "self-employed" by the Government and do not appear in reporting when their work week is cut to zero. IC's are also not mentioned in your appendix.

What are your thoughts on this?

Jul 06 07:23 PM

[Swashbuckler:]

You sure your PHD is in chemistry (and not economics)?

Jul 06 07:48 PM

[Docsroute:]

Does anyone calculate the amount of overtime hours that have diminished in the last 2 years ? These figures also should be considered. I know from personal experience that an extremely large number of US postal workers who thrived on 10-20 hours of OT a week are now getting almost none as the USPS is somewhat desperately trying to stay afloat by cutting virtually all OT. It would be very interesting to see some stats and graphs on OT rates as well. USPS is in serious difficulty, by the way.

Jul 06 09:27 PM

[John Lounsbury:]

Brian Bober, marginaltho.., and Docsroute - - -

All forms of employment, including sole proprietors and 1099 independent contractors are counted in the population survey (also called the household survey). This survey is the primary data source for determining the civilian labor force, the total number employed and unemployed and the numbers for full-time and part-time (both involuntary and by choice).

Think-about-it - - -

The DOL data bases do provide separate accounting for private and government employment and unemployment. The numbers I use are the combined numbers. I have not done analysis on the breakout between private and public sector. I have read a report from Mike Shedlock (Mish is the website) that gave the ratio of new jobs created in the past ten years as more than 2/1 government/private.

Also, all part-time hours worked, and I believe overtime as well, are factored in to determine the average weekly hours.

Jul 06 11:57 PM

[JWhitling:]

I don't know too many people that work "full time" these days that aren't working 50+ hour weeks, on salary of course. You must allow for that kind of thing in your reporting metric, otherwise we could get further skewed results. Just a thought ..

Jul 07 08:37 AM

[bigdig:]

There are a lot of 1099 subcontractors out of work the at do not qualify for unemployment and therefore I do not believe they count as unemployed. I own a small company that utilizes subs and I have 5 that have not worked since Jan 1. A friend of mine and 26 of his 1099 coworkers were let go last week as well. US Bank got rid of all their contractors. If you added up these in the figures you would hit 20%.

Jul 07 08:51 AM

[value81:]

When looking at average weekly hours data over the past several decades it's important to keep in mind the rising labor force participation rate of women during the 1960s through 1990s. With a higher rate of households containing two people in the labor force, the average workweek shrinks because there are many more people working shorter workweeks. Another way to look at this is to view the labor force participation rate for all adults, which has risen over time.

The way Wal-Mart handles its human resources is another trend pushing average weekly hours downward. The company gives workers shorter schedules in order to keep its total cost per hour of work lower by reducing benefits and overtime expenses. Wal-Mart isn't the only employer doing this, but even just by itself it employs a large and expanding percent of America's workforce.

Overall, I don't think the official unemployment rate needs to be changed in a meaningful way. There are several other BLS measures that track the state of affairs within the US workforce including average weekly hours and earnings, different measures of labor force underutilization (U1 thru U6), labor force participation rates, part time for noneconomic reasons, average length of unemployment, count of discourage workers, payroll employment figures by geography and industry, job openings and labor turnover, and mass layoff statistics.

We can use these measures to portray how the economy is affecting the labor force and determine the extent of this downturn on US workers.

It's clear from the BLS data that this recession is the worst since the 1930s. It's clear that unemployment is high and rising and that payroll employment has dropped by the largest percent of any recession since the 1930s.

Jul 07 12:47 PM

[billddrummer:]

John, great job!

I linked this article to a local blog that suggested the local employment picture wasn't as bleak as you suggest. In this area (northern NV) the construction and real estate industries have been decimated, and esimates for unemployment range from 20%-30% in FIRE, and even higher in the construction trades. Not only that, but as a corollary, retail has been hammered, and commercial real estate is in free fall.

The 'official' unemployment number is about 11%, but anectdotal evidence suggests a much higher rate, perhaps as much as 15-18%.

A brilliant analysis, thanks.

Jul 07 07:37 PM

[Flabergasted:]

Very good write up. I have been keeping it in my browser to really go through it and SOLID. For the guy who hasn't worked more than 35 hours a week in his life... I haven't worked less than 60 in mine. Yes, I am in finance....

Jul 18 11:14 PM