Economic productivity is important. Do you actually know what it means?
With labor productivity growth slowing down, let’s go back to the basics
Whether you ascribe to the idea that we’re in the Second Machine Age or entering the Fourth Industrial Revolution, you cannot deny that the world around us is changing at a dizzying pace. Just looking at the products demoed at last week’s Consumer Electronics Show (CES), you’ll learn about AI enabled mirrors or wearable brainwave devices that can improve your focus. So with all this new technology, why is economic productivity actually slowing down?
Before we can begin to answer that question, we need to abandon all of our assumptions around the concept of productivity. By learning more about what productivity actually means and how it is measured, we might have a chance of starting to understand the larger social questions.
What is productivity?
The most basic definition of productivity is the ratio of how much we produce to how much we use to produce it. You can see this illustrated in the simple graphic below. Productivity in the bluejeans industry is the output (typically measured by revenue and finished goods) divided by the inputs (typically measured by the factors of production: labor, land, and capital).
This is distinct from efficiency, which is concerned with increasing yield by reducing waste. You can think of efficient production as being an ingredient in productivity. They are certainly related concepts, but productivity is the one we are more concerned about when trying to understand how an economy is performing over time.
How we measure productivity
Every measurement is really just our best approximation of a phenomenon using the data that we can access, and finding a good way to measure productivity has been complicated. In 1961 economist George Stigler jeered that,
productivity measures of important economic magnitudes arose in the face of a theoretical tradition which denied them any relevance to economic structure or policy. (47)
So even if we have made progress since Sigler wrote those words, we must remember that the variables we select and assumptions we make will color our results, and therefore should be considered before we draw any conclusions around our findings.
Productivity measurement methods are either partial, meaning they look at one input factor, or multifactor (MFP), meaning they look at an aggregate of several factors. The differences are well outlined in Measuring Productivity, the manual created by OECD, which is used as the established standards across nations.
Within each productivity measure, analysts might be using either gross or value added methods. Paul Schreyer, the author of Measuring Productivity differentiates the two by noting:
When purchases of intermediate inputs are deducted from gross output, one obtains a measure of value added. In this sense, value added is a net measure. (24)
In other words, gross measures includes intermediate inputs (things like materials and capital) while value added subtracts them out. I highlight this because studies have found that the choice of method can result in a notable difference in the results at the industry level.
Understanding two common measures
Now we’ll go a bit deeper into the commonly used partial measure — labor productivity, and commonly used multifactor measure — total factor productivity.
- Average Labor Productivity (ALP)
Labor productivity is our attempt to understand the output of each person in the workforce. It is an incredibly valuable measurement because it is associated with rising in living standards. As workers produce more, this can increase their earnings, and the increased availability of goods can also lead to more affordable prices.
Can you believe we’ve gotten this far without mentioning gross domestic product? Well, a country’s GDP is the market value of everything it produces during a given period of time. Look back to our blue jeans example, and you might notice that GDP can correspond to output in the productivity equation. In one way of measuring labor productivity, we take GDP and divide by the number of hours worked. Occasionally the number of persons employed is used instead.
We can quickly note two limitations with this method: 1) GDP doesn’t do a very good job of capturing output in service and knowledge industries, as well as capturing the quality of output and 2) hours worked does not necessarily capture the quality of labor or skill composition of the workforce.
2. Total Factor Productivity (TFP)
Next we’ll look at Total Factor Productivity, which economist Moses Abramovitz referred to as, “the measure of our ignorance.” In the simple macroeconomic model below, you use the measured output (Y) and the function of inputs like capital (K) and labor (L), to get residual A, your TFP.
Also referred to as the Solow residual, after Nobel Laureate Robert Solow, this metric is said to capture the influence of technology and efficient production that cannot be explained by the other inputs. In other words, if you can use the same amount of capital and labor as another country, but have higher output, then we interpret this as having a more efficient and innovative way of combining those inputs.
Two notable limitations with this method are 1) R&D and organizational capital are important inputs, but as noted by the OECD Future of Productivity report, accurate measurement of these factors is, “still a work in progress” 2) measurement error with capital or labor will end up embodied in the TFP metric.
Exploring productivity issues
There is no simple answer to the productivity paradox question that started this article. But there is a lot to be unpacked when it comes to the issues with how we’re measuring output in a service and knowledge economy, as well as the timing of when we expect to reap the benefits of the digital infrastructure that we have been putting in place.
There’s much more to learn when it comes to issues of economic productivity, far beyond the scope of this article. But with this newfound appreciation for the definition of productivity and the limitations in measurement, I invite you to continue exploring the following relevant topics:
The productivity paradox
- Beyond the Productivity Paradox: Computers are the Catalyst for Bigger Changes (Brynjolfsson and Hitt, 1998)
- The productivity paradox: why we’re getting more innovation but less growth (Vox, 2016)
- The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox (David, 2010)
Decoupling of productivity and wages
- The Great Decoupling: An Interview with Erik Brynjolfsson and Andrew McAfee (Harvard Business Review, 2015)
- Is Automation Labor Share–Displacing? Productivity Growth, Employment, and the Labor Share (Autor and Salomons, 2018)
- The Productivity–Pay Gap (The Economic Policy Institute, 2019)
Measurement issues with productivity
- How Should Capital be Represented in Studies of Total Factor Productivity (Blades and Meyer-zu-Schlochtern, 1997)
- The Productivity Slowdown, Measurement Issues, and The Explosion of Computer Power (Baily and Gordon, 1988)
- Productivity: What Is It, and Why Do We Care About It? (Steindel and Stiroh, 2001)
As the world around us continues to change, we must be prudent about understanding the way that we are measuring that change. Going back to the basics helps us build a more critical analysis, so I hope this article has helped you move forward with your learning with a more nuanced lens.
Looking for more? Subscribe to the Workable newsletter — a monthly roundup of stories and research about the way that work is changing around the world.