Introduction: The Employment Effects of Technological Change
In October 2025, Amazon announced the elimination of 14,000 corporate jobs, a decision that marked the beginning of a substantial impact of artificial intelligence technology on white-collar employment. The company statement indicated that this organizational restructuring aimed to optimize operational efficiency and reallocate resources to strategic areas such as generative artificial intelligence. This case reveals the intrinsic link between technological progress and adjustments in the employment market structure, sparking a new round of discussion about technological unemployment.
The concept of technological unemployment was first proposed by Keynes in 1930, defined as a decrease in the demand for labor due to technological innovation. Historical data shows that this phenomenon has obvious cyclical characteristics.
According to bibliometric analysis, the frequency of the term "technology unemployment" showed three significant peaks in the 1920s-1930s, the 1960s, and after 2010, corresponding to the technology diffusion periods of the Second Industrial Revolution, the wave of automation, and the artificial intelligence revolution, respectively. Currently, although the US unemployment rate remains relatively stable at 4.3%, the structural adjustment of white-collar jobs has attracted widespread attention. This paper will explore the impact mechanism of artificial intelligence on the job market through historical comparative analysis, assess potential risks, and propose corresponding policy recommendations. Historical Comparative Perspective: The experience of the Industrial Revolution shows that the impact of technological progress on employment has obvious structural characteristics. In the early 20th century, the average annual growth rate of US manufacturing productivity exceeded 5%, but this growth was accompanied by a 20% decline in agricultural employment. Between 1929 and 1933, the unemployment rate rose from 3% to 25%, demonstrating that technological change could exacerbate employment pressures during economic downturns. The automation wave of the 1960s further confirmed this structural impact. Studies at the time showed that automation technology had a significant substitution effect on manufacturing employment, but the overall job market remained relatively stable due to the expansion of service sector employment and the special demands brought about by the Vietnam War. During this period, the US government established a special committee to study the impact of automation on employment, providing important references for subsequent policy-making. In the long term, the employment effect of technological progress depends on the dynamic balance between substitution and compensation effects. The substitution effect is reflected in the replacement of existing jobs by technological progress, while the compensatory effect manifests in the creation of new jobs and the increased demand brought about by the reduction in production costs. Historical experience shows that this balance requires appropriate policy intervention and a favorable market environment. At the macro level, artificial intelligence (AI) technology is becoming an important driver of economic growth. Between 2023 and 2025, AI-related investments are expected to contribute nearly one percentage point to US GDP growth. Corporate profit margins rose from 6.5% in 2003 to 10.69% in the second quarter of 2025, demonstrating the role of AI technology in improving productivity. Industry-level data shows that the impact of AI exhibits significant heterogeneity. In the banking industry, artificial intelligence technology has increased fraud detection accuracy to 95%; in the insurance industry, claims error rates have decreased by 20%; the energy sector has reduced equipment downtime by 30% through predictive maintenance; the retail industry has achieved a 15% increase in sales through personalized recommendations; and the healthcare sector has improved medical efficiency by 25% through assisted diagnosis. Behind these efficiency improvements lies a profound adjustment in the employment structure. Amazon's layoffs demonstrate that white-collar positions such as management and data analysts are facing direct impact. The company plans to improve the efficiency of middle management by 30%-50% through organizational flattening. This shift foreshadows a fundamental transformation in the traditional model of knowledge-based work. The current employment market transformation exhibits several significant characteristics: First, the skill structure of affected positions is changing. Traditionally, automation technology primarily impacts procedural production jobs, while artificial intelligence technology can replace some non-procedural cognitive tasks. This means that traditionally high-skilled fields such as education, finance, and healthcare are also beginning to face automation risks. Secondly, the pace of job turnover is accelerating. Deloitte predicts that by 2026, 92 million jobs globally will be replaced by artificial intelligence, while 170 million new jobs will be created. This rapid turnover places higher demands on workers' skills. Thirdly, the income distribution pattern may change. The application of artificial intelligence technology may further widen the gap between capital income and labor income, particularly impacting medium-skilled workers. This trend may exacerbate existing income inequality. Regional Economic Warning Signals
Regional Economic Warning Signals
Texas' economic data provides important warning signals. In October 2025, the state's service sector income index fell to -6.4, the lowest level since July 2020. The employment index was -5.8, and the business activity index was -9.4, both showing a clear contraction trend.
The retail sector performed even worse, with the sales index falling to -23.5 and the employment index falling to -5.3. These data are consistent with the overall US economic trend. In August, national retail sales increased by 0.6% month-on-month, but core sales growth was only 1.5%, reflecting insufficient consumer spending.
Labor market indicators also show signs of pressure. The consumer confidence index fell to 94.6, while the labor market differential index rose to 9.4%. These changes are temporally correlated with the widespread application of artificial intelligence technology, suggesting that technological change may be impacting the job market through multiple channels. From a macroeconomic perspective, the employment risks brought about by artificial intelligence are mainly reflected in the following aspects: In the capital market, the median price-to-earnings ratio of the top 10 artificial intelligence companies in the S&P 500 reached 32 times, significantly higher than the market average. This valuation discrepancy may reflect an overly optimistic market expectation of the benefits of artificial intelligence. If actual benefits fall short of expectations, it could trigger a market correction. The relationship between productivity and employment also warrants attention. In the second quarter of 2025, US non-farm productivity grew by 3.3%, but unit labor costs rose by only 1.0%. If this gap continues to widen, it may mean that the benefits of increased productivity have not been fully translated into workers' income, thus affecting aggregate demand. Historically, the current situation bears some resemblance to the 1930s. Technological advancements at that time also led to significant increases in productivity, but due to insufficient demand and income inequality, they ultimately exacerbated employment pressures. This historical experience reminds us that we need to comprehensively assess the employment effects of artificial intelligence. Based on historical experience and current situation analysis, effective policy responses should include the following elements: Education system reform is a long-term foundation. Emphasis should be placed on strengthening data literacy, analytical skills, and innovative thinking, and establishing a curriculum system and vocational training system adapted to the era of artificial intelligence. Special attention should be paid to building a lifelong learning system to help workers cope with frequent skills updates. The improvement of the social security system is crucial. This includes expanding unemployment insurance coverage, establishing career transition assistance programs, and exploring social security systems adapted to new employment patterns. During the technological transition period, a robust social safety net can effectively reduce the costs of transformation. Industrial policies need to play a guiding role. Deep integration of artificial intelligence with traditional industries should be encouraged, and the development of emerging industries should be supported to compensate for job losses caused by technological substitution by creating new employment opportunities. At the same time, attention should be paid to coordinated regional development to avoid excessive regional concentration of employment shocks. Conclusions and Outlook Artificial intelligence technology is triggering a new round of employment restructuring. Historical experience shows that technological unemployment has cyclical and structural characteristics, and its depth and duration depend on the speed of technological progress, labor market flexibility, and the effectiveness of policy intervention. Amazon's layoffs reflect the company's adaptive adjustments to technological change. From a macro perspective, this adjustment is a necessary process for improving resource allocation efficiency, but it also brings friction to the employment market. Successful transformation requires collaborative efforts from governments, businesses, and educational institutions to reduce transformation costs through institutional innovation and achieve social sharing of technological dividends. Future research should focus on the heterogeneity of the impact of artificial intelligence on different skill groups and the adaptability of regional labor markets. Simultaneously, a more comprehensive data monitoring system needs to be established to promptly assess the employment effects of technological change and provide a scientific basis for policy formulation. Ultimately, employment issues in the age of artificial intelligence are not only related to economic development but also to social stability and people's well-being. Only through systematic policy design and the joint efforts of the whole society can we achieve coordinated development of technological progress and employment stability, and promote society towards a more inclusive and sustainable direction.
Catherine
Kikyo
Anais
Weatherly