Where are we at in our understanding of productivity?
- Productivity is not throughput. Productivity is an efficiency ratio that purports to measure how efficiently we convert inputs into new outputs.
- The ratio doesn’t work, it’s broken. We have a good handle on the inputs but our understanding of the outputs is a misleading dog’s breakfast.
In this article I’m exploring our failure to understand the output side of the productivity efficiency ratio. We’ll look first from a national and then a firm-level perspective.
National Productivity Outputs
The economic concepts used to assess the productivity of a nation or a firm – especially when understanding outputs – have been increasingly outdated since the 1980s.
GDP
At the national level the idea of GDP (Gross Domestic Product) is increasingly seen as a relic of the manufacturing-forward industrial age. It fails to capture great swaths of nuance, yet is used as part of the macro-economic assessment of national productivity in the formulation of GDP-to-hours-worked.
What if those hours worked (or any other input used to generate value) create values (output) other than those captured in GDP (typically: consumption by consumers, investment in the capital required to create more goods and services, government spending, and net exports).
GDP fails to capture everything from unpaid labour to changes in mental and physical health of a population, to basic innovation.
Productivity J-Curve
The Productivity J-Curve (1) describes the phenomenon where productivity appears to dip after the introduction of a new technology, as firms make heavy investments in intangibles that don’t immediately result in measurable output. Over time, however, these investments pay off, leading to significant gains in productivity that exceed initial expectations. See: Brynjolfsson, E., Rock, D., & Syverson, C. (2020). The Productivity J-Curve: How intangibles complement general purpose technologies (2)
In exploring why so few writers talk about the output side of the productivity equation, the Productivity J-Curve provides important insights.
The failures revealed in understanding the Productivity J-Curve are to a degree inherited from our framing of GDP: we are not capturing key outputs, in this case the intangibles of innovation.
But even more importantly it reinforces another issue pointed out in the previous post in this series (Productivity, eh?): the time principle. The inputs (inventories, labour, etc.) and the output (GDP) are both sampled in the same period. But investment is recorded in one period, and the benefit (output) of that investment may not be seen for several years.
Quality
A third area does not rise to the same level of concern about the ‘blunt instrument’ understanding of GDP and the Productivity J-Curve, but I is still worth pointing out: how much is quality accounted for when we talk about the outputs side of productivity?
The answer appears to be “a bit”.
At the level of national GDP calculations one factor sometimes used (I have no idea yet how consistently): Hedonic pricing models, used in economics to adjust the prices of goods according to changes in quality by looking at increases in inflation-adjusted prices in like-for-like products and services.
At the firm level Quality-Adjusted Labor Productivity (QALP) is used to adjust productivity measurements to account for changes in the quality of labor inputs, which can be influenced by education, training, and experience. This reflects the fact that an hour of work today might be more productive than an hour of work in the past.
Also, Total Factor Productivity (TFP) can implicitly include quality changes if these improvements result in higher-quality outputs from the same inputs. TFP is calculated by accounting for all inputs—capital, labor, and materials—and any increase in output not explained by an increase in input is attributed to improvements in technology or efficiency, which may include quality.
There is little reporting on how common or even useful these quality considerations are when calculating productivity, so we’ll leave there for now.
What about Canada?
Does any of this reduce the concerning state of Canadian productivity described in Productivity, eh?
No.
There is little if any evidence that our productivity stats are shaped by the Productivity J-Curve in industries making huge investments and bets on technology and innovation. Investments are being made, but the evidence would suggest that they don’t rise to a level creating a massive J-Curve distortion in our sad numbers. They’re still sad.
Complicating the picture is the issue of concentration and the anti-competitive practices of both industry and government, that we tolerate in Canada. On the one hand we do see that given Canada’s tiny population and massive geography a strong argument can be made that if we value national-scale productivity, the efficiencies of scale found in sectoral concentration may be a good thing.
It is almost like we have to decide which is the bigger problem: market concentration or comparatively weak national productivity stats.
On the other hand the decline of business dynamism that occurs as large firms grow old, fat, satisfied and concentrated may lead to a decrease in productivity.
So which is it? Does M&A-driven market concentration increase or decrease our national productivity?
Industry
What does all this mean at the industrial or firm level? Are we doing any better at measuring or understanding outputs there?
Yes. But the picture is still unsatisfactory.
Outputs
Outputs are the products or services generated by a company, industry, or country. For businesses, this can range from the number of Nanaimo Bars baked by a baker to the consulting services provided by a firm. Higher outputs from the same level of inputs indicate greater productivity, a goal for any business or economy. This is the fundamental efficiency ratio described in Productivity is what?
As well as simple volume and revenue metrics, outputs should also be measured as:
- Quality and Efficiency: The quality and efficiency of outputs can create a feedback loop between a company’s reputation and customer satisfaction, which in turn can boost market position and productivity.
- Innovation: Products or services that meet customer needs more effectively.
- Consistency: Outputs that are consistent and reliable help build customer trust.
As with national productivity measurements however, most firms measure their productivity using simplistic throughput and revenue metrics. They continue to confuse productivity and throughput. Even when considering margins and margin compression, there is rarely in my experience a meaningful discussion about productivity understood as inputs (including long-term investments) and outputs that include the three measures describe above.
While that continues to be a problem from a conceptual and assessment/reporting perspective, we can still use what we have learned so far to make progress in improving productivity at the outputs level, in our organizations.
Monday Morning: The Output Edition
Some practical tips for you to apply to your organization, when looking at the question of the ‘outputs of productivity’:
Invest Strategically in Technology and Training: Recognize the impact of a productivity J-Curve in your business investments. When talking about this with clients, we also use the term ‘latency effect’. Just because nothing seems to be happening does not mean nothing is happening. Plan for a period where productivity may not visibly increase after an investment in improvement. We also find that making a much greater investment in employee training to maximize the benefits of new technologies than many leaders are comfortable with, can have a significant impact on the timing and value of the pay-off in technology investments.
Given the delayed benefits of many productivity-enhancing investments, incorporate long-term planning into your business strategy. This will certainly involve adjusting financial forecasts and communication strategies to manage employee, customer, and shareholder expectations during periods of significant investment.
Enhance Quality Measurement Practices: Instead of focusing solely on the quantity of output, integrate quality metrics into your productivity analysis. Use customer satisfaction scores, defect rates, or return rates as indicators of quality. This one is so important. Witness the mindless debate about the Work-from-home/Hybrid/Return-to-office debate of the last few years. Ninety percent of the writing on this subject is about what I call ‘meat factors’ (body in building, bum in seat, hands on equipment). The metrics almost never consider things like culture, mental healthy, product and service quality, innovation, lifetime employee value, etc. The WFH/RTO conversation is a productivity conversation. We must do better here.
Regularly Review and Adjust Productivity Strategies: Make productivity monitoring an ongoing (ideally, continual) process rather than a one-time review. This continuous approach allows you to adapt and refine strategies as you gather more data about the impacts of your investments in innovation, technology and quality improvements. Use continuous customer, employee, and technology-informed feedback about your outputs to drive continuous improvement in this area.
Engage in Long-term Innovation Planning: Given the delayed benefits of many productivity-enhancing investments, incorporate long-term innovation planning into your business strategy. This may involve adjusting financial forecasts and communication strategies to manage shareholder expectations during periods of significant investment.
You can build the plane as you fly it, but output is going to flatten during periods of significant training or technological transformation.
Have own questions about productivity? Concerns about the way I’ve described something in my explorations? Leave a comment and I’ll incorporate your thoughts in a future article.
Previous posts in this series: Productivity is what? and Productivity, eh?
(1) The “Productivity J-Curve” describes the initial decrease in productivity following the adoption of new technologies, before significant improvements become evident. When firms invest in new technologies, they often experience a period where productivity appears to drop as resources are allocated towards learning and integrating these technologies. Over time, as the firm becomes adept at utilizing the new technology, productivity increases, often exceeding the initial levels. This pattern resembles the letter “J”, hence the name “J-Curve”.
(2) Brynjolfsson, E., Rock, D., & Syverson, C. (2020). The Productivity J-Curve: How intangibles complement general purpose technologies (NBER Working Paper No. 25148). National Bureau of Economic Research. https://www.nber.org/papers/w25148