Political dramas like “The West Wing” and “House of Cards” are popular for their depictions of politicians and their staff as people motivated by some combination of idealism and Machiavellian ambition who develop and carry out elaborate schemes to serve those ends. While this makes for entertaining television, a running joke among those who have witnessed Washington in action is that reality is far closer to the comedy “Veep,” where events play out more in a fly-by-the-seat-of-your-pants kind of way.
In a similar vein, popular depictions of automation point to dystopian futures populated with robots that hunt humans, or artificial intelligence-driven computers that bring about the destruction of humanity. The reality is probably closer to the episode of “I Love Lucy” where Lucy and Ethel were tasked with wrapping chocolates coming down a conveyor belt. While the work began simply and efficiently at a leisurely pace, the chocolates began to arrive at ever-increasing speed, hilariously outpacing the characters’ abilities and altering their relationship with the automation.
The post-COVID economy has seen several instances where trends broke from established assumptions, deviating from historical trajectories. Following an initial drop during the opening months of 2020, personal expenditures on durable goods moved to a higher plane of growth compared to the growth trend exhibited between the end of the financial crisis and the beginning of 2020. This elevated demand from consumers, along with elevated business investment and government spending on military and other manufactured goods, caused a fundamental shift in the manufacturing economy. This article will explore how demand for automation responded to these historic shifts, the impact on manufacturing jobs, and what impacts these shifting trends have had on productivity.
Measuring Automation
Defining automation can be more art than science. Because automation can be broadly defined as any improved or added technology that increases the efficiency of a manufacturing operation, its growth can be somewhat difficult to track over time because what was once considered “added” automation may now be considered a standard or common part of a machine. This ubiquity can be seen in the adoption of CNC controls, automatic tool changers, bar feeders, or even simultaneous 5-axis machining. The next steps down the automation path can be seen in technologies that have become more common in recent times, such as pallet pools, predictive maintenance, and even AI-enabled machining. It is easy to equate automation with robotics, but robotics represents only a portion of the technologies available that can add efficiencies to a manufacturing operation.
Given the broad scope of automation, its adoption can’t be measured in just the number of robots added to shop floors each year. A more accurate measure would consider all added features and quality improvements of particular machines. We can approximate the adoption of automation by combining two different data series: the average value of orders as reported by the U.S. Manufacturing Technology Orders (USMTO) report, published by AMT – The Association For Manufacturing Technology; and the producer price index (PPI) for metal cutting machinery, published by the U.S. Bureau of Labor Statistics, which can be thought of as the rate of inflation for machine tools.
The difference between the change in inflation and the change in average order value over time indicates the additional automation added to the base value of each machine order. One drawback to this methodology is that as once-novel features become more common, the prices to build and implement those features tend to fall as economies of scale begin to take hold. This makes the method less useful when comparing demand for automation across longer time spans, but it is still helpful in comparing shorter-term demand.
While it’s easy to think of automation as discrete components of an order, an integrated robot, a bar feeder, or some other added feature, it is important to remember that this methodology also captures the effects of demand shifting to higher-value machines with added features or complexities. Machines that perform more complex operations may not be traditionally considered automation, but they should be, as they reduce the time to complete a part that would otherwise require refixturing or being moved between two or more machines.
By indexing both the PPI and the average order value to January 2019, we initially see a dip in the average value of orders below the PPI. It is important to remember that orders of manufacturing technology, measured by USMTO, experienced a larger decline between 2018 and 2019 than from 2019 to 2020. That environment of declining orders in 2019-2020 meant customers were generally ordering additional machinery only when capacity needs required it, a buying pattern typically associated with periods of declining average order values.
In May 2020, average order value hit bottom and began a rapid ascent, surpassing the PPI by June 2021. Like inflation in the wider economy, prices of machine tools, measured by the PPI, continued to accelerate before tapering off at an elevated level. At the same time, average order values accelerated, averaging 6.7% per month above the rate of inflation, showing a sustained level of demand for automation. This methodology accounts for nearly $370 million in automation sold with machinery in 2023, a 73% increase from 2022. Through the first eight months of 2024, $230 million of order value can be attributed to automation, a 17% increase over the first eight months of 2023.
Labor Markets
Similar to how automation is popularly misunderstood, the effects of automation on labor demand in manufacturing is generally perceived as having an inverse correlation. Research has quantified the effects of globalization and technology changes in the 1990s on the availability of jobs in manufacturing. Manufacturing employment in the United States peaked in the summer of 1979 with nearly 20 million people employed in the sector. Since then, following each recession, manufacturing employment has failed to return to its previous peak as the economy recovered. That changed following the COVID recession in early 2020.
Manufacturing employment declined rapidly like many other sectors during the recession, but it rebounded just as quickly as the service sector afterwards, surpassing the pre-pandemic peak by July 2022. Manufacturing employment continued to grow, eventually averaging out to about 110,000 jobs above the high prior to the COVID recession.
Since achieving that heightened level, job growth in the manufacturing sector has been stagnant, averaging a monthly loss of just 810 jobs since January 2023, an incredibly steady rate of change for any economic data series. Some may point to this stagnation – along with the increased adoption of automation – as evidence that technology is continuing to replace people on production lines. This assumption ignores the fact that manufacturing job openings are up 73.5% in 2024 compared to the period between the end of the financial crisis in June 2009 and December 2019. The demand for labor remains strong, but the supply is hampered by demographic, skill, and perception issues rather than the “rise of the robots” as some pundits proclaim.
The United States' population is aging, and this has serious consequences for the availability of labor for the manufacturing sector (as well as the wider economy), which should come as no surprise to readers who have attended AMT’s MFG Meeting or MTForecast conference. Combine this with a gap in the skills needed to fill current manufacturing job openings, and the pool of available workers that manufacturers are able to draw from becomes smaller.
The final workforce challenge is the longstanding perception that manufacturing is not a desirable profession. The truth is, thanks to the addition of technologies like automation – not despite it – manufacturing positions are cleaner, safer, and more engaging than commonly perceived.
Overcoming these obstacles to achieve a fully trained manufacturing workforce requires additional investments in two areas: in automation to improve the productivity of the current workforce; and in developing the necessary educational and training programs for the next generations of manufacturing workers to not only match the skillsets of those who are retiring but surpass them in using the emerging technologies needed to meet the increased demand for manufactured goods.
Productivity
After exploring the increased adoption of automation and the rebound in manufacturing employment, the next natural step is to investigate the effects these trends have had on productivity in the manufacturing space.
From the late 1980s to the beginning of the 2008 financial crisis, labor productivity in manufacturing increased steadily. The end of the financial crisis saw a mild uptick in labor productivity, but by the first quarter of 2011, it entered a shallow but steady decline that lasted until the end of the COVID recession in 2020. After an initial upward move following the COVID recession, productivity again declined, mirroring its path following the previous recession. In a curious turn of events, productivity found a bottom in the first quarter of 2023 and began an upward trajectory that lasts through today.
Writing in the summer of 2024, Danial Lashkari and Jeremy Pearce, two economists with the Federal Reserve Bank of New York, examined declining productivity in the manufacturing sector since the 2008 financial crisis. While stopping short of offering an explanation, they go into great detail to show that this trend is present across industries and company sizes.
In order to project if the current upswing in productivity can be sustained, it is important to identify some possible reasons for the observed decline in productivity following the last two recessions. Thinking back to the economic climate of the early 2010s, which was characterized by near-zero interest rates and persistently below-trend growth, two possible explanations emerge.
First, low rates may have allowed companies to survive by operating less efficiently than they would be able to do profitably in times of tighter money. While this is not to argue that monetary policy should have been more restrictive in this environment of below-trend growth and below-target inflation, efforts should be made to understand the potential loss of productivity during periods of prolonged monetary accommodation should this theory prove correct.
A second and possibly compounding explanation is that the below-trend growth, coupled with a downturn in manufacturing, meant manufacturers experienced a lot of idle capacity for quite some time. This long-term underutilization could have introduced systemic inefficiencies into the production process. While these theories need to be further examined and tested, they should be kept in mind as we analyze the current uptick in productivity.
The current upward trend in productivity began in the first quarter of 2023. This coincides with both a massive uptick in the demand for automation, as measured by the average value of USMTO orders, and overall manufacturing employment reaching its post-COVID plateau. Although orders of manufacturing technology and estimated automation were elevated prior to this time, lags between orders and delivery, as well as other supply bottlenecks, may have prevented earlier productivity gains. Furthermore, heightened interest rates caused by the Federal Reserve’s tight monetary policy have forced manufacturers to find operational efficiencies after no longer being able to pass price increases onto customers, and historically elevated capacity utilization rates in the manufacturing sector have forced shops to improve processes, invest in additional technology, and expand the capabilities of their workforce. These conditions point to the likely continuation of productivity gains as well as a continued need for manufacturing technology through the remainder of 2024 and beyond. Setting the Stage for Tomorrow
As the chocolates came down the conveyor faster and faster in that famous episode of “I Love Lucy,” Lucy and Ethel struggled to keep pace. While this factory configuration may have been more efficient at producing laughs than wrapped chocolates, the two characters’ frenetic pace is emblematic of the way manufacturers have had to face a barrage of events amid the rapid pace of technological change in recent years. Ensuring continued productivity gains requires manufacturers to embrace further advances in automation as well as to foster the workforce of tomorrow that will utilize that automation.
To read the rest of the Industry Outlook Issue of MT Magazine, click here.
