I've been trying to do some cursory research on labor productivity. In the process, I've noticed that the readily available data in the US, as accessed through either BLS or FRED (yes, I know it's the same underlying data), is all in terms of year-to-year change, or level relative to a fixed year. Is there any theoretical or technical reason for this? It seems like an absolute level of labor productivity should be well defined: GDP divided by total hours worked. Am I neglecting something that would prevent reliable calculation of an actual dollars-per-hour number? If not, where can I find this figure?
The data you ask for is in the OECD stats website under
Productivity > Productivity and ULC – Annual, Total Economy > Level of GDP per capita and productivity
This is the graph (in 2010 PPP US dollars):
Aternatively, you can compute it manually:
Select a series of constant GDP of your interest.
Divide it by number of hours for a group of interest.
You can find both series in the above website. GDP is under
National Accounts > Annual National Accounts > Main Aggregates > Gross domestic product
Hours are under
Labour > Labour Force Statistics > Hours worked
To answer your first question, a level refers to the value of a certain indicator at a given point in time. A level of productivity doesn't have much reporting value unless it is used in some intermediate step of a calculation/formula.
When it comes to percentage change this value tells the change from one period to the next and is what is usually used productivity analysis. A growth rate is used when an economist wants to know how fast an indicator has risen (or declined) over a certain period. For example, the labor productivity rate from 2009 to 2010 or 2009Q1 to 2009Q2. This estimate is more valuable for reporting purposes and to researchers and policy makers.
the Bureau of Labor Statistics often reports the productivity change increasing or decreasing over quarter to quarter or year over year in order to answer the question are we more productive this year or not. Thereby given policy makers and researchers more tools to help them with their decisions.
For the second question, You can make your own calculation of productivity level by dividing GDP by total hours of all persons. However, realize that total hours of all persons measures doesn’t include some of the sectors that GDP covers. For example, total hours of all persons doesn’t include General government, nonprofit institutions, paid employees of private households, and the rental value of owner-occupied dwellings (https://www.bls.gov/lpc/lpcover.htm).
When constructing your own productivity measure make sure that each measure that you are using (GDP and hours in this case) has the same coverage in their respect data if not then you will have a biased estimate.
Here is the OECD manual on productivity measurements that can help answer other questions you may have: http://www.oecd.org/sdd/productivity-stats/2352458.pdf