The Impact of Automation: How I narrowly escaped task dispacement and what automation means for the US
The impact of automation and artificial intelligence (AI) on the future of employment is a widely debated topic. There are some disagreements about the future net employment effects of these technologies, but most economists broadly agree that nearly all occupations will eventually be affected. In today's post, I will take a look at a potential instance where automation may have affected one of my previous firms.
The first full-time job I had post-undergrad was at a boutique healthcare consulting firm (small-sized consulting firm). I applied through a job posting. I was grateful for the opportunity - as it was during some of the lowest points in my life. The job afforded me the opportunity to recover from personal challenges and provided me with the knowledge of the basic foundations of the healthcare market.
It was a tough year. In 2017, I didn't feel like myself. I was grieving from the loss of two close family members, healing from a broken heart, and coping with a chronic illness diagnosis. It's true when it rains, it pours. Every day was a battle. I sat with a sadness that I carried for a long time. Reflecting on the days I couldn't get out of bed, I owe thanks to a couple of roommates of mine that made my heart a little less heavy.
They would check on me and ask me to do cycling, running, or yoga. Whether an intense workout activity or a girl's therapy session - I ended up feeling a little bit lighter afterward. If not for them, then I would not have survived that period in my life. The biggest piece of advice that I learned from that time in my life is to not be afraid to lean on your friends and family a little. Call that friend who is dealing with a lot or speak to them about their troubles. You never know how the tiny gesture will have a ripple effect.
Though I was dealing with these personal challenges in 2017, I had to keep carrying on. I had to adjust to the new chapter in my life - adulting. With my newfound independence, I was creating boundaries personally and professionally. At work, I was getting task after task. My coworkers didn't know what I was going through. I had to keep my personal problems to myself at work because otherwise, those personal details would seem like a hindrance to productivity. So, I tried to maintain professionalism and gain new skills.
It wasn't easy nor was it perfect but it was a necessary period in my life to get me to where I am today. Professionally, I learned how to define the geographic reach of a hospital market. And parts of this framework could be applied to other industries too. Healthcare in the United States is only another type of business. I also learned how to determine physician demand and supply. A part of the task was the manual process of calling physician offices and manually defining markets. Almost every time I was given an assignment in this job, I knew there had to be a better way to complete the tasks more efficiently. The tasks were tedious and redundant. There was a lot of room for human error. I thought maybe if I could program I would have the tools to do the tasks more efficiently and accurately. However, at the time the small firm didn't have the tools and I didn't have the training.
My father would incessantly lecture me to keep up with technological advancements and development whether through education or credentialing to avoid task displacement (See Glossary). I knew after working at this boutique consulting firm for about half a year I needed to add more value to the workplace by automating processes. I didn't want to be laid off in 5-10 years when my tasks would get displaced or my job would become automated. Typically, the news talks about the manufacturing sector when discussing automation; however, other sectors can be affected by automation as well (See Glossary).
So, I did my research and invested in myself. I searched for programs in my industry that would add value to my skill set. I had a few options but with my credentials, I only applied to a Health Economics program instead of a data engineering. When searching for programs, I looked at courses specifically. I saw that there were programming and modeling seminars. If I could go through the process of applying to Master's programs again, I would have talked to current students through LinkedIn to understand the benefits of the program as well as the placement opportunities post-graduation.
I'm grateful I moved on and strived to stay current in my industry. Because I found out later that the firm laid off all the consultants. One coworker put it as, "The firm was bleeding money."
Looking back, I am curious to see what factors may have influenced a massive layoff at the previous firm I worked at.
- Could automation be one of the reasons my previous firm could not compete efficiently?
- What other industries are affected by automation?
- How does automation affect unemployment?
- And what is the relationship between inequality and technology?
- What will automation mean for wages and income inequality?
Let's Talk Social Science
Automation and Wage Inequality
In June 2021, two researchers published a working paper in the National Bureau of Economic Research. Drs. Daron Acemoglu & Pascual Restrepo from MIT and BU, respectively, studied where tasks across various industries are allocated to different types of labor and capital. For example, in the car manufacturing industry tasks such as assembly to painting are being displaced by advanced robotics.
The researchers found that between 50% and 70% of changes in the US wage structure over the last four decades are accounted for by the relative wage declines of worker groups specialized in routine tasks in industries experiencing rapid automation. In other words, 50-70% of the changes in the wage structure are due to the wage declines of those car assemblers or painters versus the wage increase of the engineer innovating the robotic equipment for the car manufacturers.
Dr. Acemoglu and Restrepo revealed that there was considerable variation in task displacement across industries, with the largest levels of task displacement seen in mining, chemical products, petroleum, car manufacturing, and computers and electronics. Most notably, the researchers highlighted that task displacement leads to sizable increases in wage inequality, but only small productivity gains.
Wage Structure
Wage earnings are the main source of income for many workers, and wage increases act as a lever for raising living standards. There have been reports of stagnant median wages (see glossary). These reports have caused concern over economic growth overall has not been distributed in gains across all worker groups over the last several decades.
The Congressional Research Service, a public policy research institute in the United States Congress, prepared a report to shed some light on recent patterns in the wage structure. The report estimates real (inflation-adjusted; see glossary) wage trends at the 10th, 50th (median), and 90th percentiles (see glossary) of the wage distributions for the workforce. The trends overview real wages for the workforce as a whole and several demographic groups. The report also explores changes in educational attainment and occupation for these groups over the 1979 to 2019 period.
Major findings from this report include:
- Real wages at the 90th percentile increased across sex, race, and ethnicity; however, wage growth was much higher for White workers and lower for Black and Hispanic workers. For middle (50th percentile) and bottom (10th percentile), wages rose at lower rates (e.g. women) or fell in real terms (e.g. men).
- The gender wage gap narrowed, but other gaps did not. From 1979 to 2019, the gap between the women’s median wage and men’s median wage became smaller. However, gaps expanded between the median wages for Black and White workers and for Hispanic and non-Hispanic workers over the same period.
- Real wages fell for workers with lower levels of educational attainment and rose for highly educated workers. Wages for workers with a high school diploma or less education declined in real terms at the top, middle, and bottom of the wage distribution. Whereas, wages rose for workers with at least a college degree. The wage value of a college degree (relative to a high school education) increased markedly over 1979-2000. The college wage premium has leveled since that time, but it remains high. High-wage workers, as a group, benefited more from the increased payoff to a college degree because they are the best-educated and had the highest gains in educational attainment over the 1979 to 2019 period.
- Education and occupation patterns appear to be important to wage trends. Worker groups studied in this report were more likely to have earned a bachelor’s or advanced degree in 2019 than workers in 1979, with the gains in college degree attainment being particularly large for workers in the highest wage groups. For some low- and middle-wage worker groups, however, these educational gains were not adequate to raise wages. Workers’ occupational categories appear to matter as well and may help explain the failure of education alone to raise wages.
Wage inequality has risen sharply in the US and other industrialized economies over the last four decades. While the real wages of workers with a post-graduate degree rose, the real wages of low-education workers declined significantly. The real earnings of men without a high-school degree are now 15% lower than they were in 1980. Notably, for low- and middle-income worker groups, education did not lead to a rise in wages.
The Future of the Labor Market
Sree Ramaswamy and Anu Madgavkar discuss automation and its effects on workers. Both are partners at Mckinsey Global Insitute. The conversation aims to also examine the ways smart policy may help prevent employment or salary tensions. In a case study, they discuss how Germany uses a model of work-sharing. Work-sharing is when work is either getting automated or falling off due to temporary demand slowdowns, then whatever work that has to get done gets allocated across a larger number of people. Thereby, people are not out of work at any point in time and keep their skills alive in that context.
Another example, Anu Madgavkar discusses is the apprenticeship system. This system can streamline formal education for young people and give them on-the-job training. It would allow young workers to move quite seamlessly and become part of a workforce, even while they complete their formal education. These pilot programs have worked in other high-income countries. In the European model, the programs have actually worked well because these are almost trilateral partnerships between the government, the firms or the companies, and then the workforce. According to Anu Madgavkar, the future is in trilateral partnerships to bring solutions.
In the 21st-century labor market, the degree to which workers, their families, and their communities adjust to the new technologies will depend on how public policy, private institutions, and businesses evolve to support them. Notably, for now, human cognition is needed for most tasks. However, new evidence signals that workers with routine tasks are at risk of significant disruption and transitions.
Glossary
Task displacement: when technology overtakes a particular task due to efficiency
Automation: the process of applying new techniques and new software combined with capital equipment or new hardware. The byproduct of the process argued by many economists is the removal of the human workforce.
Median: The middle number in an ascending list of numbers.
Inflation: a general increase in prices and the decreasing purchasing power of money. Wages need to be adjusted for inflation, because a 2022 wage may not buy you the same amount of things as a 2018 wage. To be able to compare apples to apples, economists use real wages or inflation-adjusted wages to look at trends.
Percentile: the location of a score in a distribution expressed as the percentage of cases in the data set with scores equal to or below the score in question. For example, the 90th percentile is when 90% of the scores in the distribution are equal to or lower than that score.
Distribution: a function or a listing of all the possible values of a dataset
References
Acemoglu, D., & Restrepo, P. (2021). Tasks, Automation, and the Rise in US Wage Inequality (No. w28920). National Bureau of Economic Research.
Title: Real Wage Trends, 1979 to 2019 Report#: R45090 Author(s): Sarah A. Donovan, David H. Bradley Date: December 28, 2020
McKinsey & Company. (2019, May 11). What will automation mean for wages and income inequality? McKinsey & Company. Retrieved January 31, 2022, from https://www.mckinsey.com/featured-insights/future-of-work/what-will-automation-mean-for-wages-and-income-inequality