Great. I’ll compile a report examining the historical trends in labor’s share of U.S. GDP from 1900 to the present, and project forward to 2070 with a focus on the quantitative impact of AI on labor’s income share. I’ll also include a section on the rise and decline of the U.S. horse industry and its economic contribution over time.

I’ll let you know when the full report is ready to review.

The U.S. Labor Share of GDP (1900–2025) and Projections to 2070, and the Rise and Decline of the U.S. Horse Industry

Labor’s Share of GDP in the United States (1900–2025)

Historical Overview: Long-Run Stability and Shifts (1900–2025)

Economists have long observed that labor’s share of national income – the portion of GDP paid out as wages, salaries, and other worker compensation – remained remarkably stable over most of modern history () (). From the early 1900s through the mid-20th century, roughly two-thirds of U.S. national income went to labor, with only modest fluctuations around this level (). This empirical regularity was so robust that it became one of “Kaldor’s stylized facts” of economic growth, suggesting a balanced rise of both labor and capital incomes over time. Indeed, even as the United States transitioned from an agrarian society in 1900 into an industrial and then post-industrial economy, labor’s share held within a relatively narrow band for decades ().

However, beneath this long-run stability, shorter-term economic shifts did cause labor’s share to vary cyclically. For example, during the Great Depression of the 1930s, corporate profits collapsed while wages were somewhat stickier, causing labor’s share to spike temporarily as the economy contracted. Conversely, in the wartime boom of the 1940s, profits surged and wage controls were in place, keeping labor’s share contained. By the post-World War II era (1950s–1960s), the U.S. saw a thriving middle class and strong unions, and labor’s share was at a historically high level – by some estimates around the high-60% range (). In fact, there was a “sharp increase in the labor share” in the 1960s, reflecting robust wage growth and collective bargaining outcomes, followed by a long-run decline after the 1970s (). Over the last quarter of the 20th century, the labor share in the U.S. steadily eroded, essentially falling back to its 1960 level by the 2010s ().

Key measurements underscore this downward trend. For instance, across advanced economies, labor’s share fell from about 54% around 1980 to ~50% by 2014, reflecting a broad international pattern (A new look at the declining labor share of income in the United States). In the U.S., measures of the aggregate labor share (adjusted for self-employment) likewise declined markedly after 1980. By the 2000s, U.S. labor’s share hovered near its lowest levels on record in postwar data, roughly in the lower 50-percent range (depending on the exact definition) (A new look at the declining labor share of income in the United States) (). Notably, this decline paused or even reversed briefly around recessions (labor’s share often ticks up in downturns as profits fall faster than wages), but the overall trajectory has been downward since the late 20th century (Class War and Labor's Declining Share - Monthly Review) (). Today, as of the mid-2020s, labor’s share of U.S. GDP remains historically low – on the order of low-60s% or even high-50s%, compared to mid-60s% decades earlier ().

Factors Behind the Decline in Labor Share

Multiple factors have contributed to the post-1980 decline in labor’s share of income. Globalization and trade integration have put pressure on manufacturing wages (industries most exposed to import competition saw larger labor share declines) (). The decline of labor unions and shifts in bargaining power – for example, policy changes in the 1980s that weakened organized labor – also reduced labor’s ability to claim a stable share of output (). But perhaps the most-studied driver is technological change and automation. Recent research finds that a significant part of the U.S. labor share decline is linked to the adoption of new technologies, from information technology to robotics. For instance, the rise of “hyper-productive” firms that deploy automation extensively has led to a reallocation of income toward capital and away from wages ([PDF] Why Is the Labor Share Declining?). One study notes that much of the decline in U.S. labor share over recent decades can be explained by the spread of software and robots that substitute for routine human labor (). Consistent with this, industries that rapidly adopted automation experienced larger labor share drops (). In short, technological progress in the late 20th and early 21st centuries has not always been “labor-augmenting” – instead, it has often capital-augmented production, tilting income toward owners of capital.

Other structural forces include the shift from labor-intensive manufacturing to capital-intensive digital services, and even demographic or firm dynamics. Some analyses point out that an increasing share of national income has been accruing to profits and to very highly paid executives (blurred in labor share measures), rather than to rank-and-file wages (Economic Impact of Horse Industry in United States - News - The Northwest Horse Source) (Economic Impact of Horse Industry in United States - News - The Northwest Horse Source). Despite these nuances, the consensus is that automation and global market forces are primary reasons labor’s slice of the pie has shrunk () (). This shift has raised concerns because a falling labor share can contribute to inequality – if a smaller portion of GDP goes to workers, the gains from growth accrue more to owners of capital.

(Data visualization of U.S. labor share over time can be inserted here, showing the relative stability through mid-century and the decline after ~1980. For instance, a line chart could illustrate labor’s share (%) on the y-axis and years 1900–2025 on the x-axis, with annotations for major events like 1929, 1970, etc. All source data for such a figure would come from historical national accounts and research by Piketty et al. and the U.S. Bureau of Labor Statistics, as cited above.)

(The labor share remained around ~65% through much of 1900–1970, then trended downward to around 57–60% by 2020 (). This visualization would reinforce the narrative described.)

Key Economic Turning Points in Labor Share

Several key economic shifts over the last century help explain the pattern of U.S. labor share:

  • The Great Depression (1930s): During 1929–1933, U.S. GDP shrank nearly 30% and unemployment spiked to 25% (Great Depression Economic Impact: How Bad Was It? | St. Louis Fed). Corporate profits were wiped out in many sectors, while many workers still retained (diminished) wages. As a result, labor’s share of a much-smaller GDP likely rose sharply in the early 1930s, even as absolute incomes fell. New Deal policies later bolstered labor (e.g. strengthening unions and setting wage floors), keeping labor’s share elevated.

  • World War II and Postwar (1940s–1960s): War production and postwar rebuilding led to high demand for labor. By the 1950s, strong industrial unions and steady productivity gains translated into broad-based wage growth. Labor’s share was unusually high by the late 1960s, reflecting what some call the “golden age” of the American worker. As one analysis notes, the share “increased sharply in the 1960s” (). During this era, a combination of high union membership, minimum wage increases, and a booming industrial sector ensured workers captured a large portion of economic gains.

  • Globalization and Oil Shocks (1970s–1980s): The 1970s brought twin oil crises and stagflation, squeezing corporate profits, and initially kept labor’s share high. But by the 1980s, major shifts occurred: President Reagan’s policies (e.g. the breaking of the PATCO strike in 1981) signaled a decline in union power (), and globalization accelerated (with manufacturing jobs moving offshore). Labor’s share began a persistent decline. In many industries, outsourcing and foreign competition curbed wage growth, and capital began claiming a larger share of income.

  • Tech Boom and Automation (1990s–2000s): The digital revolution further altered income dynamics. Highly profitable tech firms (with relatively few workers) emerged, and automation started replacing routine jobs (from factory assembly lines to office bookkeeping). The late 1990s saw a brief uptick in labor’s share as a tight labor market raised wages, but thereafter the trend was downwards. By the early 2000s, three-fourths of the labor share decrease since 1947 had occurred post-2000 ([PDF] A new look at the declining labor share of income in the United States), indicating an acceleration of the decline. The Great Recession of 2008–2009 initially caused labor’s share to jump (as profits plunged), but this was temporary – in the recovery, profits recovered faster than wages, pushing labor’s share down again.

  • 2010s–2020s: In the last decade, U.S. labor’s share has remained low by historical standards. There was a notable rise in 2020 during the COVID-19 recession (as government support and wage maintenance contrasted with faltering corporate earnings), but as the economy normalized, labor’s share settled roughly back to its pre-pandemic pattern. The structural forces of tech and globalization continue to keep labor’s share subdued relative to mid-20th-century levels. Policymakers and economists are increasingly concerned with this trend, arguing that restoring a higher labor share may be important for reducing inequality (Class War and Labor's Declining Share - Monthly Review).

Projections Through 2070: The AI Era and Future Labor Share

Looking ahead to the coming decades, how might labor’s share of national income evolve through 2070? Many forecasts hinge on the impact of artificial intelligence (AI) and advanced automation on the economy. AI is often compared to past General Purpose Technologies (GPTs) like the steam engine or electricity in terms of potential impact, but it also raises unique concerns. Notably, experts warn that AI could drive an unprecedented shift in labor’s share of income, perhaps far larger than past technological revolutions (Generative AI: A Turning Point for Labor's Share?) ().

In historical tech revolutions, new machines automated some tasks but also spurred new industries and productivity gains that kept labor broadly employed and maintaining its income share. During the First and Second Industrial Revolutions, labor’s share did fluctuate – research suggests labor’s share of income in industrializing economies fell by about 5–15 percentage points in the 19th century during the spread of steam power and factory production (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School). However, this decline was largely transient: ultimately, new jobs and industries (and perhaps labor-friendly policies) restored labor’s share, preserving the long-run two-thirds proportion () (). Indeed, by the mid-20th century, labor’s share in countries like the U.S. and U.K. was not drastically different from pre-industrial levels, confirming the historical resilience of labor’s share ().

Will AI be different? A growing number of economists fear that it could be. AI today has the ability to not just mechanize routine manual work (as past machines did) but also to perform cognitive tasks, decision-making, and potentially a broad swath of white-collar jobs. A recent Federal Reserve analysis cautions that AI may “undermine labor’s share of national income” in the long run, potentially permanently reducing the role of human labor in the economy (). In other words, unlike previous technologies which eventually increased demand for human labor elsewhere, AI could automate such a wide range of tasks that the balance between labor and capital income might fundamentally shift. According to researchers Drozd and Tavares (2024), the advent of generative AI and related technologies poses a “once-in-a-lifetime” risk of a structural decline in labor’s share (Generative AI: A Turning Point for Labor's Share?). Their economic models suggest that if AI automates tasks faster than it creates new ones, the share of income going to workers could fall significantly below its historic norm in coming decades.

Quantitatively, empirical forecasts are now emerging. One study focusing on big data and AI in the financial sector found that adoption of these technologies led to about a 5 percentage-point decline in labor’s share of income in that industry (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School) (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School). Extrapolated economy-wide, a 5% drop in labor’s share due to AI could materially lower workers’ slice of GDP. For context, a drop of that magnitude (e.g. from ~60% of GDP to ~55%) is comparable to the entire decline experienced in many countries from 1980 to 2010 (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School). Some analysts note that the Industrial Revolution’s major technological upheavals reduced labor share by 5–15% temporarily (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School); AI’s impact could be on the high end of that range if its deployment is pervasive. The U.S. Congressional Budget Office (CBO) and other forecasters incorporate gradual automation gains in long-term projections, generally implying modest further erosion of labor’s share over the next 50 years absent policy changes. For instance, if current trends continue, by 2070 labor’s share in the U.S. could drift toward the low-50s% (as a share of GDP), especially if AI adoption accelerates in the 2030s and 2040s.

That said, there is considerable uncertainty. Optimistic scenarios envision AI dramatically boosting productivity such that even a smaller share going to labor yields higher real incomes for all. In such scenarios, labor’s share might decline, but total labor income still grows robustly (a smaller slice of a much larger pie) (Generative AI: A Turning Point for Labor's Share?) (). Additionally, new categories of jobs could emerge (just as the computer industry created new occupations) that sustain demand for human labor. Pessimistic scenarios, however, imagine AI and machines capturing most new value, leading to a decoupling of productivity and wages. In extreme cases – if AI reaches human-level capabilities across most fields – one could imagine labor’s share falling to very low levels by late-century, as capital owners (those who own the AI and robots) accrue the majority of income. While such extremes are speculative, mainstream forecasts do anticipate a noticeable downward pressure on labor’s share. The Federal Reserve Bank of Philadelphia warns that to keep labor income growing, AI’s productivity benefits must be very large; otherwise, labor could become significantly less important in the national income distribution () ().

In summary, the baseline expectation through 2070 is that labor’s share in the U.S. will likely continue its gentle decline, potentially reaching levels substantially below the 20th-century average. Many forecasts peg U.S. labor’s share to drop several more points by mid-century (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School). For example, if labor’s share is around ~60% today, it might ease toward ~55% or even lower by 2050–2070, assuming AI and automation proceed unchecked. This trend would mark a historic break – a departure from the roughly two-thirds ratio that held for over a century (). Policymakers are thus closely watching AI’s rollout. Depending on how things unfold, we may see responses (such as new labor protections, retraining programs, or redistribution policies) aimed at ensuring the gains from AI are shared. The stakes are high: the balance between labor and capital in the economy will influence everything from inequality and consumer demand to political dynamics. As one Federal Reserve analysis concludes, “thanks to artificial intelligence, labor’s share of national income may no longer hold steady” – a cautionary note that the coming decades could rewrite one of the fundamental relationships in economics () ().

The Rise and Decline of the U.S. Horse Industry (1850–Present)

Rise of the Horse Economy (1850–1910)

In the mid-19th century, horses were a pillar of the American economy. They were the engines of agriculture, transportation, and even early industrial power (via horse-driven machinery). In 1850, the United States had on the order of only ~4 to 5 million horses and mules (Horse and Mule Population Statistics - Energy History), but this number would grow explosively as the country expanded westward and industrialized. By 1900, the U.S. equine population had reached 21.5 million (horses and mules combined) (SOA_IV_2007_Inside). And by the 1910s, horses were near their historical peak: the 1910 census counted over 24 million horses and mules (SOA_IV_2007_Inside), and the equine population continued growing to an all-time high around 1915, when roughly 26.5 million horses and mules were in the United States (SOA_IV_2007_Inside). This was a sixfold increase from the horse population in 1840 (which was 4.3 million) (Horse and Mule Population Statistics - Energy History), illustrating the tremendous rise of horse power alongside America’s economic growth.

The early 20th century was the zenith of the horse-based economy. In 1910, virtually every farm used horses for plowing and harvesting, every city depended on horse-drawn wagons for goods delivery, and personal travel often meant a horse and buggy. The result was that horses had a huge economic footprint. Millions of acres of cropland were dedicated just to feeding horses. In fact, around the 1910s, an estimated 27% of all U.S. farmland (about 93 million acres) was required to grow feed for horses and mules (Microsoft Word - Tractor for NBER.doc). This astonishing figure – over a quarter of cropland output devoted to supporting draft animals – indicates how central horses were to production. Large sectors of the economy revolved around horses: feed growers, saddlery and wagon manufacturers, blacksmiths, stable owners, and veterinarians. One analysis notes that circa 1910, farm horses and mules consumed the output of ~22% of harvested cropland (and draft animals in cities another ~5%), demonstrating the heavy burden of feeding working horses (Microsoft Word - Tractor for NBER.doc).

In terms of GDP contribution, the horse industry in the late 19th and early 20th century included both the direct output of breeding and trading horses and the indirect output enabled by horses. While it’s hard to isolate “horse GDP” in that era, we can grasp its scale through related metrics. For example, in 1900, horse labor effectively powered most of the $4.8 billion farm sector (which was a large share of GDP at the time), as well as urban transport and trade (Microsoft Word - Tractor for NBER.doc) (Microsoft Word - Tractor for NBER.doc). The capital value of horses was significant: they were an essential asset for farmers – often the second most valuable asset after land. Cities struggled under literal mountains of horse manure, a byproduct of their ubiquity, illustrating the environmental and logistical significance of equine transport (Automobiles Freed Us from the Tyranny of Horses | Mises Institute). It’s fair to say that from the late 1800s to about 1910, a sizable percentage of the nation’s economic activity was horse-dependent, whether via farm output, freight movement, or the maintenance of horses themselves (When horses powered our lives | historyonthefox) (When horses powered our lives | historyonthefox). One historian remarked that a “fairly large percentage of the nation’s economy was based on horses” during this period (When horses powered our lives | historyonthefox) – likely on the order of 10% or more of GDP when considering all linked activities (direct and indirect).

Peak and Early Decline: Transition to Automobiles (1910–1940)

The 1910s mark the turning point. The U.S. horse population peaked around 1915 at roughly 26 million head (SOA_IV_2007_Inside). After this peak, a steep decline set in over the next decades – a direct consequence of the transportation revolution led by the automobile and the tractor. Henry Ford introduced the affordable Model T in 1908, and by the 1910s cars and trucks were rapidly becoming common. Farmers also saw the introduction of reliable gasoline tractors (Fordson tractor mass-produced from 1917), which could plow fields without needing feed or rest. As Americans embraced motor vehicles, the need for horses plummeted.

The numbers tell the story starkly. By 1930, the horse/mule population had fallen to about 18.9 million – a drop of over 7 million from the peak just 15 years earlier (SOA_IV_2007_Inside). By 1940, only 13.9 million remained (SOA_IV_2007_Inside). And the decline continued into mid-century: the 1950 census showed about 7.6 million horses and mules, and by 1960 the U.S. equine population hit a low of just ~3 million (SOA_IV_2007_Inside). In other words, between 1915 and 1960, the U.S. lost almost 90% of its horses. This collapse was directly tied to the rise of automobiles and tractors, which outcompeted horses in nearly every domain – they were faster, could work longer hours, and did not require farmland for feed.

The economic implications of this transition were enormous. Huge swaths of farmland were “freed up” from producing oats and hay for horses. As tractors took over, farmers could convert feed-crop acreage to food crops or other uses. By 1930, only ~65 million acres were needed for horse feed (down from 93 million in 1915), and by 1960 a mere 5 million acres sufficed for the greatly reduced herd (Microsoft Word - Tractor for NBER.doc). One economic history study calculated that the land released from horse feed production by mid-century was equivalent to two-thirds of the total cropland of the entire country in 1920 (Microsoft Word - Tractor for NBER.doc) – an immense gain in agricultural capacity for human food and fiber. This land-use change boosted productivity and economic growth: resources once tied up in supporting draft animals were reallocated to other productive purposes. In terms of labor, the tractor and motor vehicle also saved farmers tremendous effort. The U.S. Department of Agriculture estimated that by the 1940s, farm mechanization (tractors, trucks) was saving about 1.7 billion labor-hours annually in farm work and horse care – roughly equal to 8% of total farm labor requirements, or the labor of 850,000 workers (Microsoft Word - Tractor for NBER.doc) (Microsoft Word - Tractor for NBER.doc). These efficiency gains contributed to the great increases in agricultural output and helped labor migrate to manufacturing and services.

Of course, entire occupations and industries declined. Blacksmiths, carriage makers, feed dealers, and stable operators saw their livelihoods vanish or transform. For example, thousands of blacksmith shops either closed or converted into auto mechanic garages during the 1920s and 1930s. The horse-and-buggy trade gave way to the automobile industry – which itself became a major engine of growth. It’s notable that even as the horse population fell, the total number of vehicles (horse-drawn or motor) per capita kept rising; by 1920 there was roughly 1 car per 13 people (up from 1 car per 200 people in 1910) (Automobiles Freed Us from the Tyranny of Horses | Mises Institute) (Automobiles Freed Us from the Tyranny of Horses | Mises Institute). By 1970, the U.S. had more vehicles than ever (about 1 per 1.8 people, almost all motorized) (Automobiles Freed Us from the Tyranny of Horses | Mises Institute), underscoring how thoroughly automobiles replaced horses.

From a macroeconomic perspective, the GDP contribution of the horse economy sharply diminished in this transition. In 1900, value related to horses (transport services, farm traction, etc.) was still significant. But by 1940, automotive output and petroleum had displaced horses’ economic role. Industries supporting horses shrank, while new industries (auto manufacturing, rubber, oil refining, road construction) grew. This was a classic Schumpeterian dynamic of “creative destruction”: the horse was destroyed as an economic unit, but automobile-related economic activity boomed. The net effect by mid-century was overwhelmingly positive for GDP – productivity and output were higher – but it meant that horses went from being an essential capital input to a nearly irrelevant one in the span of two generations.

(image) Figure: U.S. Horse and Mule Population, 1840–2017. The rise and fall of American horse power is evident. The population peaked in the 1910s (over 26 million horses and mules) and then declined sharply as automobiles and tractors spread. By the 1960s, horse numbers hit a low point, and have since stabilized at a fraction of their former level (SOA_IV_2007_Inside) (7.2 Million Reasons Why Now Is The Time! – American Horse Publications).

The Modern Horse Industry (1960–Present)

After 1960, the horse population stabilized and then modestly rebounded as horses found a new role: recreation, sport, and leisure. Once horses were no longer needed for work, they became animals of hobby and competition. Equestrian sports (horse racing, show jumping, rodeo) and recreational riding gained popularity in the late 20th century, leading to an uptick in horse numbers. From the trough of about 3 million in 1960, the U.S. horse population rose to an estimated ~9 million by 2000 (SOA_IV_2007_Inside). In the early 21st century, economic downturns and rising costs tempered this growth; a 2017 survey estimated about 7.2 million horses in the United States (7.2 Million Reasons Why Now Is The Time! – American Horse Publications). Today (2020s), the horse population remains in that general range (likely between 7 and 8 million), vastly smaller than a century ago, but large in absolute terms for a now-specialized industry.

The economic role of horses today is fundamentally different. Instead of providing transportation or farm labor, horses are part of the service and entertainment economy. The horse industry’s contribution to GDP comes from areas like breeding, training, racing, entertainment, tourism, and recreation. According to the American Horse Council’s detailed economic impact analysis, the **horse industry in 2017 directly contributed about 50 billion to U.S. GDP** in the form of spending on horses, equipment, feed, facilities, and related services ([7.2 Million Reasons Why Now Is The Time! – American Horse Publications](https://www.americanhorsepubs.org/newsgroup/7-2-million-reasons-why-now-is-the-time/#:~:text=7,Is%20The%20Time)). (For scale, 50 billion is only about 0.25% of U.S. GDP today – a tiny share, reflecting how much the economy has grown and diversified away from horses.) This direct impact supported nearly 1 million jobs (988,000 direct jobs) in breeding farms, stables, race tracks, veterinary services, farriery (horseshoeing), feed production, etc. (Economic Impact of Horse Industry in United States - News - The Northwest Horse Source) (Economic Impact of Horse Industry in United States - News - The Northwest Horse Source). When considering indirect and induced effects (e.g. suppliers to the horse industry, and spending by industry employees), the total economic impact was estimated at $122 billion and 1.7 million jobs in 2017 (Economic Impact of Horse Industry in United States - News - The Northwest Horse Source). These figures show that while horses are no longer economically essential, they still form a multi-billion-dollar industry encompassing rural and urban areas – from Kentucky’s Thoroughbred farms to riding stables on the suburban fringe.

Within the modern horse industry, segments include horse racing (and pari-mutuel betting), which is a sizable entertainment business in states like Kentucky, California, and New York; recreational riding and horse ownership (many horses are owned for pleasure riding, youth activities like 4-H, or ranch work on a minority of cattle farms); and show and competition circuits (such as dressage, rodeo, show jumping, etc.). There is also a niche working horse segment that persists: horses used in police units, ranches, Amish farming communities, and for therapy/education programs, but these are relatively small in number. The American Horse Council’s breakdown even identifies a contribution from “traditional work horses,” which in 2023 was noted to add a direct $2.6 billion in value and about 86,000 jobs – showing that even for work purposes, horses aren’t entirely gone (2023 Economic Impact Study - The American Horse Council). Nonetheless, the character of horse-related employment has shifted: far more people today work in equine healthcare, training, tourism, and event management than as blacksmiths or wagon builders.

It’s interesting to note that the horse industry’s contraction and transformation offers parallels for modern technological disruption. Just as horses were rapidly displaced by machines, many tasks performed by humans today face potential displacement by AI and automation. The horse example shows that while society overall benefited hugely from cars and tractors (productivity soared), the transition was painful for those in the old horse-based trades. By mid-20th century, horses had virtually disappeared from economic statistics – yet, a niche survived and then reinvented itself around leisure. Similarly, one might expect that if AI replaces certain human labor in the coming decades, new human-centric niches could emerge (artisanal crafts, experiential services, etc., analogous to equine leisure today), even as the mainstream economic roles change.

From a historical standpoint, the rise and fall of the horse in America is a clear timeline of technological change:

  • 1850: ~4 million horses; horses are primary draft power in a predominantly agrarian economy (Horse and Mule Population Statistics - Energy History).
  • 1900: 21.5 million horses; rapid growth as railroads and cities still rely on horses for “last mile” transport, and farms expand with animal power (SOA_IV_2007_Inside).
  • 1910–1915: Peak ~25–27 million horses; horse-drawn vehicles everywhere from city streets to rural farms (SOA_IV_2007_Inside). Horses integral to GDP (transport, farm output).
  • 1920s: Automobile adoption explodes; horse population begins steep decline (down ~20% by 1925) (SOA_IV_2007_Inside). Fordson tractors cut into farm horse use.
  • 1940: ~14 million horses; half the 1915 number (SOA_IV_2007_Inside). Many urban horses gone; trucks haul freight.
  • 1960: ~3 million horses; the horse is largely absent from work life (SOA_IV_2007_Inside). Only recreational and a few work uses remain.
  • 2000: ~9 million horses; revival driven by recreation and sport (SOA_IV_2007_Inside) (SOA_IV_2007_Inside). Horse racing and equine hobbies contribute to a distinct industry.
  • Present (2020s): ~7–8 million horses; horse industry ~$50 billion direct GDP (7.2 Million Reasons Why Now Is The Time! – American Horse Publications), focused on leisure, sports, and cultural value rather than essential economic utility.

In sum, the U.S. horse industry evolved from a cornerstone of the economy to a specialized niche. Its rise and decline were tied to one of the most significant technological shifts in history – the displacement of animal power by engines. At its height, the horse was as critical to economic output as oil is today; at its nadir around 1970, it was nearly irrelevant economically. Today’s horse sector, while economically modest, remains culturally and historically significant. It serves as a living reminder of how innovation can reshape labor and capital: an instructive parallel as we face the AI revolution. The horse’s story highlights both the immense gains from technological progress (faster transport, higher productivity) and the need to adapt an economy to new realities – lessons that resonate as we contemplate the future of work and the distribution of income in the 21st century.

Sources: Historical labor share data from Piketty, Saez, and Zucman (2018) and U.S. national accounts () (). Labor share trends and causes from Elsby et al. (2013) (), McKinsey Global Institute (A new look at the declining labor share of income in the United States), and Federal Reserve analyses (). AI impact forecasts from Columbia Business School research (Does the Rise of AI Compare to the Industrial Revolution? ‘Almost,’ Research Suggests | Columbia Business School) and Federal Reserve Economic Insights (Drozd & Tavares, 2024) (Generative AI: A Turning Point for Labor's Share?). Horse population data from USDA historical censuses (SOA_IV_2007_Inside) and the American Horse Council (7.2 Million Reasons Why Now Is The Time! – American Horse Publications). Economic implications of the horse-to-automobile transition from Collamer & Moe (NBER) (Microsoft Word - Tractor for NBER.doc) (Microsoft Word - Tractor for NBER.doc) and historical accounts (When horses powered our lives | historyonthefox). All numerical data and quotes are sourced as indicated.