Wednesday, July 06, 2005
ISM Non-manufacturing Index Up, Even Its Employment Component
Confusing the experts again, the Institute for Supply Management's Non-Manufacturing Index is up to a robust 62.2, and the employment component jumped to 57.4, up 4 points.
As manufacturers downsize, they outsource to service businesses for a wide range of tasks that they used to do internally. These also play havoc with the way the government tracks manufacturing. Got a cafeteria? Bring in an outside firm to run it. Those cafeteria workers used to be manufacturing employees. Now that they work for a company like Aramark, they're now service employees, even though they go to work every day to the same building and do the same exact things. Got a payroll department? Close it and use ADP of PayChex. Those payroll workers were manufacturing employees. When ADP or PayChex hires the same workers they're counted as service employees. Have mechanics on staff for your equipment? Let them go. They used to be manufacturing workers, and now that they work for a equipment maintenance company, they're service employees. The examples are endless, but I think it's clear. Some of it comes from regulations, such as pension laws that are incredibly complex resulted in companies hiring pension management firms, or benefit management firms, for example. Most of it comes from information specialization. Companies like PayChex have the intricacies of payroll laws and payroll information programmed in their systems, and offer lower cost and better service than a payroll department can.
That's just one reason why the ISM non-manufacturing data is so strong. Construction has been a hot sector for a while, and people forget that's not a manufacturing business, and neither are mining (the oil business is not part of manufacturing either!), and the usual suspects of finance and banking are all examples of high-paying service industries. This is an indicator that Friday's employment figures should be good, except for the usual warning that payroll data have missed the strength of the service businesses, which are often too small to have impact on payroll data, and are best reflected in the household survey.
http://www.ism.ws/ISMReport/NMROB072005.cfm
As manufacturers downsize, they outsource to service businesses for a wide range of tasks that they used to do internally. These also play havoc with the way the government tracks manufacturing. Got a cafeteria? Bring in an outside firm to run it. Those cafeteria workers used to be manufacturing employees. Now that they work for a company like Aramark, they're now service employees, even though they go to work every day to the same building and do the same exact things. Got a payroll department? Close it and use ADP of PayChex. Those payroll workers were manufacturing employees. When ADP or PayChex hires the same workers they're counted as service employees. Have mechanics on staff for your equipment? Let them go. They used to be manufacturing workers, and now that they work for a equipment maintenance company, they're service employees. The examples are endless, but I think it's clear. Some of it comes from regulations, such as pension laws that are incredibly complex resulted in companies hiring pension management firms, or benefit management firms, for example. Most of it comes from information specialization. Companies like PayChex have the intricacies of payroll laws and payroll information programmed in their systems, and offer lower cost and better service than a payroll department can.
That's just one reason why the ISM non-manufacturing data is so strong. Construction has been a hot sector for a while, and people forget that's not a manufacturing business, and neither are mining (the oil business is not part of manufacturing either!), and the usual suspects of finance and banking are all examples of high-paying service industries. This is an indicator that Friday's employment figures should be good, except for the usual warning that payroll data have missed the strength of the service businesses, which are often too small to have impact on payroll data, and are best reflected in the household survey.
http://www.ism.ws/ISMReport/NMROB072005.cfm
