[lbo-talk] Productivity in Services

[Christian Gregory]

I wrote to BLS this morning after thinking about the productivity thing.
Phyllis Otto confirms a thought I had then: that outsourcing does affect
labor productivity because the imported service is still cheaper than the
one produced in house. If I understand her correctly, when Dell USA pays for
Dell India's services, and then charges a customer account, there are two
output transactions. The payment to Dell India reduces output, but the sale
of computer support to the US customer increases it. Hours also fall. So,
though she says it's hard to measure, it clear that the direction of the
mismeasurement would be positive, since output will fall less than hours,
probably.

Christian

----- Original Message -----
From: Otto, Phyllis - BLS <Otto.Phyllis at bls.gov>
To: 'Christian Gregory' <christian11 at mindspring.com>
Sent: Tuesday, December 02, 2003 2:30 PM
Subject: RE: Productivity in Services Question


> Christian,
>
> I'm going to answer your inquiry for two different types of measures
because
> I'm not exactly sure which series you're interested in.
>
> For the measures of productivity in the business and nonfarm business
> sectors, we get our measure of output from the data used to construct the
> gross domestic product (GDP).  The employment and hours series are derived
> from the BLS establishment and household surveys.  In the event that a
> service previously performed in-house by an establishment is now purchased
> from a foreign establishment, the exact effect on the productivity
measures
> for these sectors cannot be measured.  Both output and hours will fall.
For
> output, the domestic product sold (whether a good or service) will
continue
> to be included in GDP.  To the extent that the product is now less
expensive
> (because the imported service is presumably cheaper), the cost of the
> product would fall.  However, this will be picked up in the price index
and
> "real" output would remain the same.  However, the imported service will
be
> subtracted from output, which will now be lower than before.  Whether
> productivity rises or falls depends on whether hours or real output drops
> more.
>
> In our manufacturing measures, labor productivity will rise when a service
> is shifted from being handled in the same establishment to being done
> elsewhere.  This will happen whether the service is now purchased from
India
> or from across the street in Indiana.  The BLS does produce measures of
> multifactor productivity (MFP) for manufacturing industries and one of the
> inputs for these measures is purchased services.  (The others are labor,
> capital, purchased materials, and energy.)  In the MFP measures, it's
> possible to observe how businesses make shifts between inputs--we can see
if
> labor is being replaced by capital or purchased business services--in an
> effort to use their resources most effectively.  However the MFP measures
> cannot be produced in as timely a manner as the quarterly labor
productivity
> measures; they lag by several years.
>
> Phyllis Otto
> Major Sector Productivity Division
> Bureau of Labor Statistics
> 202-691-5604
>
>
>
> -----Original Message-----
> From: Christian Gregory [mailto:christian11 at mindspring.com]
> Sent: Tuesday, December 02, 2003 8:20 AM
> To: Otto.Phyllis at bls.gov
> Subject: Productivity in Services Question
>
>
> Hi Phyllis,
>
> I have a question about how productivity in services is measured. In
> particular, I'm interested in how outsourced labor services are accounted
> for in BLS stats. If , for example, an investment bank is serving clients
in
> the US, but some of their staff is located in India,  how do the BLS stats
> compensate for the fact that the output measure records the dollar output
of
> workers both here and abroad but only counts the hours of workers here?
Even
> deducting the factor payments overseas from output, wouldn't this
overstate
> productivity?
>
> Thanks in advance.
>
> Christian Gregory
>