Study Debunks the ITAA

Date: Thursday, August 01, 2002 9:33 AM

*** H-1B NEWSLETTER ***

Get the Facts on H-1B at
www.ZaZona.com

Robert Rivers recently did a study that shows why the ITAA is wrong. Here is
a short bio followed by the study. Even for the people on this newsletter
that are not in technical fields this study is interesting because Robert
Rivers shows how psuedo-scientific statistics can be used to manufacture
shortage propaganda. Harris Miller, the president of ITAA, uses the shortage
propaganda in order to justify H-1B.

Mr. Rivers is graduate of MIT with a SB degree in Electrical Engineering.
He is a registered Professional Engineer in New Hampshire and a Fellow of
the Institute of Electrical and Electronic Engineering, (IEEE).

He was elected to the IEEE Board of Directors (1975-1976) where he protected
and promoted the professional activities including an Engineers Needs policy
statement that was passed by the Board of Directors. He was the point man
that organized the Engineering Joint Societies Employment Practices
Guidelines that was originally designed as an IEEE employment guidelines but
was quickly expanded to apply to all engineers. He ran for President of IEEE
for 1977 but did not make it in a three way race. He subsequently served on
or led committees or task forces on Manpower with some contributions on
forecasting unemployment specifically for the 1991-1992 recession. He
continued to be involved in manpower activities during the 90’s. He
attempted to force the IEEE-USA Manpower Committee to respond to the
Information Technology Association of America (ITAA) promotion of expansion
of quotas for temporary work visas for foreigners. He was shortly removed
from committee membership. Having seen the inaction at the IEEE, he
organized the American Engineering Association (AEA) Manpower Committee to
respond to the ITAA.

ANALYSIS OF MAY 31, 2002 ITAA “REPORT”

The Information Technology Association of America (ITAA) is a trade
association of about 500 member firms. Their recently released survey
report, “Bouncing Back: Jobs, Skills and the Continuing Demand for IT
Workers,” was supported by Microsoft, Cisco Systems, Intel Corporation, ITT
Technical Institute, American Association of Community Colleges,
Brainbench, The Chubb Institute, Dice Inc., Prosoft Training and SRA
International. Project advisor was the National Workforce Center for
Emerging Technologies (NWCET) (1) which has developed IT skill standards
with funding from the National Science Foundation, and has devised the eight
IT job categories used to classify responses.

When evaluating a report based on a survey it is important to know why it
was done and who financed it. In this case, ITAA is doing a service for its
members by supporting and lobbying for increased availability of temporary
foreign workers through the H-1B visa program. In order for ITAA to support
and publish survey results, the survey must “prove” the need for increased
temporary worker admissions. By selective survey techniques and
non-disclosure of the weakness of the connection between the sample and the
total population, the survey has
delivered the desired conclusions for ITAA. What has happened is that ITAA
has issued yet another misleading “survey” whose conclusions are
predetermined (and favorable to their large corporate members, who pay for
the surveys).

One of the techniques employed in the latest of a long string of “surveys”
that attempt to justify an expansion of the controversial H-1B visa program
is to “stuff the ballot box” by choosing a survey technique that
exaggerates the demand for information technology workers. By surveying
hiring managers, they collect exaggerated data because each hiring manager
optimistically believes that his or her company will take more of the
available business. Their own data will show that the demand is
exaggerated. The survey data shows that their own forecast is optimistic by
9.64%.

Regardless of claims of random sampling that would permit the conclusions of
their sample to be extended to the universe of IT workers, their own data
showing composition of the surveyed hiring managers discloses that the
sample is highly biased toward surveying of IT firms.

In an attempt to justify the continuation and expansion of the H-1B

temporary worker permit, ITAA bases its conclusion on a survey of 532 hiring
managers in IT and non-IT companies conducted by a market research firm of
Market Decisions Corporation of Portland Oregon. While there is a mountain
of detail in the report consisting of cross-sectional analysis of the data,
the fundamental finding was a number called the GAP. The GAP is the
“shortfall” -- the difference between the number of IT personnel they say
they would like to hire and the number they say they expect to hire. Both
numbers are opinions of the surveyed hiring managers.

The report extrapolates the 532 hiring manager surveys to the universe of IT
and non-IT companies based on what is claimed to be a chain of random
sampling. Because the report itself shows the inability of the opinion of
hiring managers to forecast next years’ conditions, the GAP cannot be used
to make any policy decision. In addition, the composition of the sample as
shown in the report does not support the contention of random selection that
would be necessary to extrapolate to the universe.

A proportional random sample (532) of the universe would produce 27.6 hiring
managers to survey at IT firms. In fact, they surveyed 155, almost 6 times
the number they should have interviewed.

Market Decisions Corporation (MDC) states in the report that it used a basic
organizational population of 19,443 IT organizations and 354,686 Non-IT
Companies. In statements to this author, a random sample of 22,626 were
drawn from the universe by Survey Sampling Inc. and supplied to MDC. A
sample selection program was then run on the 22,626 by MDC until they had
what they considered to be a sufficient size sample to produce a +/- 3.6%
sampling variability at the 90% level. If the process truly produced a
random sample, one would expect 19,443/374,129*532=27.6 to appear in the IT
sample and 354,686/374,129*532=504.4 to appear in the Non-IT sample more or
less. There were 155 hiring managers selected in the IT companies, not
27.6. There were 377 selected in the Non-IT organizations, not 504.4 as
would be expected from random selection. IT companies were grossly
over-represented in the surveyed sample and Non-IT companies were
significantly under-represented. The conclusion must be reached that the
sample selection was not random. Further, dividing the sample into smaller
subcategories negates the confidence level claimed for the overall sample,
in providing separate analyses for IT firms and for non-IT firms, for four
geographic regions of the United States, and for eight IT job categories
(plus “other”). In addition, a claim of stratified samples is not supported
by any data other than what is presumed to be the +/- 3.6% maximum sampling
variability at the 90% confidence level of the whole sample. The data
supplied is insufficient to independently verify validity of the survey.
Extrapolation of the survey findings to the universe is unacceptable.

The data for the year 2001 in the report indicates a surveyed demand of
901,589 positions and a so called gap of 425,358 anticipated unfilled
openings. The difference of 476,231 is the net number of new positions
expected to be filled for a total employment of 10,900,644. The 2002 Survey
shows starting employment of 9,895,916 . Those surveyed indicated there
would be a 4.57% increase in the workforce while the actual surveyed
employment in the beginning 2002 indicated a 5.07% decrease in employment.
The spread between what was forecast and what actually happened was 9.64%.

This is what is characterized as the optimistic bias of the technique of
surveying hiring managers. For the year 2002, the expected expansion of the
work force according to the survey is 5.76% (569,928). Applying the
correction factor derived from the 2001 data to 2002 would lead to the
conclusion that employment will decline by 438,979. The raw numbers produced
by the employer hiring manager survey technique cannot be used without
adjustment for bias inherent in the technique.

Any attempt to use the GAP numbers as a proxy for shortages cannot be
accepted. The GAP is completely under the control of the employer having
the ability to adjust wages, working conditions, benefits, training and
skills needs to name a few. Increasing the offers will increase the supply,
as in any free market.

Summarizing, the survey data cannot be extrapolated to the universe of IT
employees for failing a test of random selection. The survey technique
yields a positive bias of overall demand and if used at all should be
corrected to eliminate the bias. The use of the GAP as an indication of
shortages should be rejected.

Surveyed
Starting IT Demand Forecast Net% Yr-nd Forecast

Year Employment Increase Jobs Filled Gap Base Hired/Riffed IT Employment

2000 10,771,364 1,608,499 765,171 843,328 3.30 11,536,535
2001 10,424,413 901,589 476,231 425,358 4.57 10,900,644
2002 9,895,916 1,148,639 569,928 578,711 4.44 10,465,844
2003 9,456,937

* Adjusting for the cumulative error in 2000 and 2001 forecasts when
compared to
actual employment, by reducing the forecast year-end employment by 9.64%.
This suggests that 2002 year-end employment will be lower by 438,979 for the
year.
Both the 2000 and 2003 Starting IT Employment are estimates.

Robert A. Rivers, Life Fellow IEEE, American Engineering Association
Manpower
Chair
Orange, MA 07/23/02

1) Access the annual report for NWCET at www.nwcet.org

My thanks go out to many that assisted, contributed to and reviewed this
report.

Help to Keep ZaZona.com Online

Donate to the Cause at
http://www.zazona.com/Donations.htm

Back to archives