An official website of the United States government
Yes. BLS uses the general term "multifactor productivity" instead of “total factor productivity” for the models and measures with multiple inputs generating output.
Multifactor productivity (MFP) refers to productivity statistics that take into account more than one kind of input to production. A full estimate of MFP typically includes measures of the growth of each input, and each factor's influence on output, together with an estimate of residual productivity growth not accounted for by the factor inputs.
BLS publishes several kinds of estimates of multifactor productivity. They incorporate measures of productive capital and labor inputs. Some include other inputs: energy, materials, purchased business services, or the education and experience composition of the workforce.
Total factor productivity (TFP) refers to the same concept as MFP but the word "total" implies that the model includes all relevant inputs. However, estimates of MFP generally do not include measures of economies of scale, organizational capital, labor effort, management quality, capacity utilization, or infrastructure such as highways and computer networks. Furthermore, different studies measure the inputs in different ways. For example, some studies measure labor input using a count of hours worked, while others weigh the hours worked by the education or experience of the work force. No approach achieves a “total” measurement. See #2 below about intangible capital, which is measured differently in different multifactor measures.
Models of multifactor productivity with varying measures of inputs and outputs are active areas of research. Generally, statistical agencies use the abbreviation MFP for these classes of estimates to convey that there are multiple measured inputs, without implying that there is a consensus on the set of factors which account for output growth. Academics and central banks are more likely to use the abbreviation TFP.
Data users should be aware that, whichever term was used, there is no single consensus universal model. The measures presented will vary depending on which inputs were included and how they were constructed or measured. Some researchers use the residual alone as the estimate of MFP, whereas others use the vector of capital and labor estimates along with the residual. Some researchers treat the MFP residual as an indirect estimate of aggregate technological change, although the residual includes many otherwise unmeasured factors listed above.
In the context of productivity statistics, tangible capital refers to structures, equipment, and land owned by organizations that produce output. Intangible capital includes software, databases, brand name recognition, designs, artistic originals, copyrights, patents, trademarks, business processes, and other technologies or institutions which are owned or controlled by productive organizations and are useful inputs to production of output. There is ongoing research as to how to define intangible capital in different contexts and use it in measuring productivity.
Working papers are available on the BLS Working Papers page. Monthly Labor Review articles are available on the MLR website. Many of the published papers are available to subscribers at JSTOR. Unpublished papers can be obtained by contacting the authors.
When hiring, the Division of Productivity Research and Program Development will advertise job opportunities in the Job Openings for Economists. The division is not hiring this year.
Last Modified Date: June 16, 2021