Summary: For the past half-century, much of the developed world has experienced a puzzling slowdown in productivity growth—the rate at which workers and businesses become more efficient over time. While digital technologies have advanced at a remarkable pace, innovation in the physical world has slowed considerably. The problem is not a lack of scientific breakthroughs or a shortage of good ideas. Rather, it is a failure to translate discoveries into products, infrastructure, and services that improve everyday life. This slowdown is largely the result of policy choices. By reforming outdated permitting systems, using innovation incentives such as R&D prizes and Advance Market Commitments, and reducing barriers created by protected local monopolies, we can accelerate the spread of new technologies and usher in a new era of prosperity.

---

In a previous exploration of the housing affordability crisis, I observed a sobering reality: artificial scarcity is often a policy choice. We have placed arbitrary limits—mostly through local governments—on our ability to build homes, driving up costs and restricting opportunity. But this pattern of self-imposed constraint does not stop at the edges of our neighborhoods. It extends into the institutions and policies that shape economic growth. It is one of the primary reasons why, despite living in an age of extraordinary digital innovation, we remain stuck in a decades-long productivity slump.

Economists often measure technological progress using a concept called Total Factor Productivity (TFP). In simple terms, TFP measures how efficiently an economy turns labor, land, and capital into goods and services. When TFP rises, society discovers better ways to produce more with the same resources.

From the 1920s through the early 1970s, TFP in the United States and much of the developed world grew at more than 2 percent per year. This was the era that gave us commercial aviation, widespread electrification, antibiotics, and the Apollo program. The physical world was transformed in a single generation.

Since the early 1970s, however, productivity growth has slowed dramatically to less than 1 percent in most years. As investor Peter Thiel famously quipped, “We wanted flying cars; instead, we got 140 characters.” Digital technologies have advanced rapidly, while progress in energy, transportation, infrastructure, and advanced manufacturing has been far slower. We can send vast amounts of information across the globe in milliseconds, yet we often struggle to build major infrastructure projects on time or on budget.

A 2020 paper by Nicholas Bloom and co-authors argues that good ideas are getting harder to find – that is, more investment in research and development has become necessary for each new patentable idea. However, more recent research by Teresa Fort and co-authors (currently in working paper form) suggests that this is not the case. The Bloom et al. result may, in fact, be an artifact of focusing on manufacturing firms, which were dominant from about 1970 to 1990. Fort and her co-authors show that patenting and innovation have shifted in recent decades, becoming dominated by firms in information, management, and professional services.

Because manufacturing is a physical process, it is much more likely to be subject to, for example, environmental regulations, whereas an IT firm operates in a much less regulated sector. So, our relative stagnation may not be the result of a scientific drought after all. Universities and research laboratories continue to produce remarkable discoveries. We are not failing at invention; we are failing at diffusion, the process of turning new discoveries into widely used products and services.

The Diffusion Deficit and the Permitting Veto

Innovation does not benefit society until it escapes the laboratory and enters the marketplace. The journey from a peer-reviewed paper to a consumer-ready product is long, expensive, and uncertain. Over time, policymakers have added layer upon layer of regulatory complexity to that journey.

Physical innovation requires physical construction. New technologies need testing facilities, advanced laboratories, semiconductor fabrication plants, energy infrastructure, and transportation networks. Yet building almost anything of significance in the modern West often requires navigating years of environmental reviews, public-comment periods, and multi-agency approvals.

Laws such as the National Environmental Policy Act (NEPA) and state-level counterparts such as the California Environmental Quality Act (CEQA) were originally intended to prevent environmental harm. Over time, however, they have increasingly become tools for the delay of progress. Because these laws frequently allow opponents to challenge projects on procedural grounds, they have contributed to what political scientist Francis Fukuyama calls a “vetocracy”—a system in which many actors can block decisions but few can make them. Average NEPA environmental impact statements now take almost four years to complete, with many extending far beyond a decade. Thankfully, the median is a bit shorter, but still about 2.5 years.

Consider the recent push to reshore semiconductor manufacturing. While the government has allocated billions of dollars in subsidies to build these vital factories, the physical construction is bottlenecked by years of permitting and environmental reviews. A state-of-the-art fabrication plant (commonly called a “fab”) that takes 18 months to build in Taiwan or South Korea can take three to five years just to obtain a permit in the United States.

The result is predictable: projects take longer, cost more, and become less attractive to investors. Even when governments provide subsidies for strategic industries such as semiconductor manufacturing, years of permitting can slow implementation. Time is money, and prolonged regulatory uncertainty discourages investment in capital-intensive industries.

The solution is straightforward, even if politically difficult. Critical infrastructure, advanced manufacturing facilities, and research laboratories should face streamlined approval processes. If projects satisfy clearly defined environmental and safety standards, they should be approved in months rather than years.

Pull Mechanisms: R&D Prizes and Commercialization

Reducing regulatory barriers is only part of the solution. We must also rethink how innovation is encouraged and financed.

In addition to corporate financing, most governments try to support innovation through “push” funding. Researchers receive grants to conduct experiments, purchase equipment, and explore new ideas. This model, some economists argue, can be effective for basic science, especially when commercial applications may be years away.

Commercialization presents a different challenge. Many promising technologies fall into what innovators call the “Valley of Death” – the difficult period between a successful laboratory demonstration and a commercially viable product. At this stage, development costs rise sharply while uncertainty remains high.

That is where “pull” mechanisms become valuable. Instead of paying for research inputs, policymakers reward successful outputs. An Advance Market Commitment (AMC), for example, guarantees that a buyer will purchase a product if it is successfully developed. Rather than funding every possible approach, the sponsor commits to paying for results.

Economist Michael Kremer helped pioneer this approach through vaccine development programs. More recently, Operation Warp Speed demonstrated its effectiveness. The government did more than fund vaccine research; it guaranteed large future purchases for successful vaccines. By reducing market risk, policymakers encouraged firms to accelerate development and manufacturing simultaneously. The result was one of the fastest vaccine-development efforts in history.

Consider other approaches. Throughout history, prizes have also stimulated innovation. The Longitude Prize helped solve a critical navigation problem for maritime trade, while the Ansari X Prize helped launch the private spaceflight industry. Pull mechanisms align private incentives with public goals by rewarding success rather than political connections or grant-writing skill.

Breaking Local Monopolies and Regulatory Capture

When people hear the word “monopoly,” they often think of large technology companies. Yet some of the most significant barriers to innovation exist at the local level.

The electric utility sector provides a clear example of how regulatory design shapes technological adoption. Because most utilities operate as regulated monopolies with government-guaranteed rates of return on capital investments, their business model relies on continuous, large-scale infrastructure growth. 

Put simply, utilities make more money the bigger power plants and power lines they build, so they usually prefer huge projects over things like rooftop solar panels that let people generate their own power without the utility having to build as much infrastructure.

Decentralized energy technologies—such as local battery storage, micro-grids, and advanced management software—directly threaten this model by optimizing the existing grid and reducing the need for new capital projects. As a result, studies from the MIT Energy Initiative and industry financial analysts indicate that utilities frequently leverage legacy regulatory processes to delay or block these decentralized innovations from integrating into the wider network.

Similar dynamics exist elsewhere. State dealership franchise laws frequently restrict direct-to-consumer automobile sales, making it more difficult for new manufacturers to enter the market. Occupational licensing requirements now affect roughly one-fifth of American workers and can create barriers to entry that limit competition and labor mobility.

Innovation depends on what economist Joseph Schumpeter called “creative destruction” – the replacement of older, less efficient business models with better ones. When established interests use regulation to shield themselves from competition, they slow technological adoption and reduce future productivity growth. Encouraging competition and reducing regulatory barriers at the state and local level would help accelerate the diffusion of new ideas throughout the economy.

Choosing Abundance

The productivity slowdown is not an immutable law of nature. It is, at least in part, the consequence of policy choices. Human ingenuity remains as powerful as ever. We have more scientists, more capital, and better tools than any previous generation. The challenge is not generating ideas; it is allowing those ideas to spread.

By streamlining permitting processes, expanding the use of R&D prizes and Advance Market Commitments, and reducing barriers created by protected local monopolies, we can accelerate innovation in the physical world.

An additional one or two percentage points of annual productivity growth may sound insignificant. Yet when compounded over decades, the effects are transformative. Higher productivity means higher incomes, better health outcomes, more abundant energy, and greater opportunities for future generations. The ideas already exist. The question is whether we will allow them to flourish.

“Ramp AI Index shows business AI adoption crossed 50% for the first time in March, reaching 50.4% of businesses. A year ago, it was 35%. Half of businesses on Ramp now pay for AI.

Anthropic continued its surge, growing from 24.4% to 30.6% of businesses — a 6.3-percentage-point gain, surpassing last month’s record monthly gain.”

From Ramp.

Summary: In just two centuries, humanity has turned light from a rare luxury into one of the most abundant resources on Earth. What once demanded hours of labor now costs a fraction of a second’s work, thanks to relentless innovation and human creativity. From candles to LEDs, the story of light reflects a larger truth: when people are free to invent and exchange ideas, they transform scarcity into abundance and darkness into illumination.

---

Our book Superabundance (2022) was inspired in part by the work of Nobel Prize–winning economist William Nordhaus, who conducted an extensive analysis on the “time price” of light over the span of human history. He called time prices the true prices. Light can be measured in lumens. Comfortable reading light is around 1,000 lumens. Nordhaus reported that in 1830, earning sufficient money to buy the candles necessary for one hour of light at 1,000 lumens required around three hours of labor. A candle generates around 12 lumens; therefore, one would need 83 candles to generate 1,000 lumens.

Innovation replaced candles with kerosene lamps and then with incandescent lighting and then LED lighting. Today, for 75 cents, one can buy a Cree J Series 5050C E Class LED that generates 228 lumens per watt. By increasing the wattage to 4.4 watts one can, therefore, generate 1,000 lumens of light. Electricity prices are currently around 17 cents per 1,000 watt hours, commonly known as kilowatt hours or kWh. One watt hour costs 0.017 cents; thus, the 4.4 watts to power the Cree LED for one hour would cost a mere 0.0745 cents. The average worker earns $36.53 an hour, or slightly more than a penny per second. Working for around 0.0735 seconds, therefore, the average worker earns enough money to buy 1,000 lumens for one hour.

The light that cost 10,800 seconds in 1830 costs only 0.0735 seconds today. The time price has dropped by 99.99932 percent. For the time it took to earn the money to buy 1,000 lumens for one hour in 1830, workers today earn 146,980 hours of light today. That’s a 14,697,900 percent increase. Light abundance has been increasing around 6.3 percent annually on a compound basis, doubling every 12 years.

Calculating Changes in Global Light Resources

Over the last 195 years (1830-2025), the world’s population rose from 1.2 billion to 8.2 billion—a factor of 6.83, or a 583 percent increase. To measure how humanity’s resource base has changed, we calculate the size of the global resource “pie” by multiplying personal resource abundance by population. That reveals how much “total abundance” exists across humanity at a given moment.

As we already saw, during the 195-year period, personal light abundance rose by a factor of 146,980. Assuming for argument’s sake that everyone in the world enjoys American prices of LEDs and energy, combined with the 6.83-fold increase in population, the global light abundance factor would amount to 1,004,360. In other words, the global light pie has grown by 100,435,912 percent—from an index value of 1 in 1830 to 1,004,360 today.

Light abundance would have grown at a compound annual rate of roughly 7.3 percent for almost two centuries, doubling about every 9.8 years. What was once scarce, flickering, and expensive has become nearly boundless—flowing at the speed of electrons and photons across the planet.

Resource Elasticity of Population

In economics, elasticity compares the percentage change in one variable against the percentage change in another. Between 1830 and 2025, global light resource abundance increased by 100,435,912 percent. During same period, the world’s population increased by 583 percent. Dividing 100,435,912 percent by 583 percent gives us 172,176. Every 1 percent increase in population thus corresponds to a 172,176 percent increase in global light abundance.

Let There Be More Light

We have witnessed an exponential efflorescence of light—an illumination not merely of our cities but of the human spirit itself. More people with light has meant more minds, more ideas, and more ventures into the unknown. When free to imagine and innovate, humans transform scarcity into abundance—and ignorance into insight. Over the past two centuries, we have converted the darkness of want into the radiance of wealth, beginning with light itself. From the barbarous glow of whale oil to the humble candle, and from the flicker of gas and kerosene to the steady blaze of electricity and the brilliance of silicon, each technological leap has kindled new horizons of discovery. Every advance has multiplied the possibilities for the next. The ultimate source of growth is not material—it’s the human mind set free.

The next time you turn on a light switch, please take a moment to appreciate the great work of free and creative people toiling to bring us out of the darkness. Compared to the abundant light of today’s world, our ancestors really did live in the “dark ages.”

Find more of Gale’s work at his Substack, Gale Winds.

Summary: Fearing job losses and loss of control, some politicians are calling to halt technologies like driverless cars and artificial intelligence. Yet history shows that efforts to “protect” people from innovation only delay progress and raise costs. Automation inevitably disrupts, but it also saves lives, boosts efficiency, and expands opportunity.

---

Some US conservatives want to slam the brakes on progress, quite literally. At the recent US National Conservatism conference, Senator Josh Hawley of Missouri declared: “Only humans ought to drive cars and trucks.”

His techno‑scepticism runs deeper than opposing driverless vehicles. “Every so‑called innovation the tech class has delivered in recent decades operates as a power transfer … from us to them,” he warned. “Us” meaning the honest “common man”. “Them” being Silicon Valley transhumanist elites and their job-killing code.

We must take this urge to ban driverless cars and smother AI seriously. Not because Hawley is right, but because his anxieties are commonly held here. A recent survey by the Schwartz Reisman Institute found just 8 per cent of Britons are strongly positive about artificial intelligence, putting us last except for Australia. Only France and the US had higher negative sentiment than the UK too (33 per cent). On driverless cars, just 22 per cent of us say we’d feel safe in one. This is fertile ground for populists looking to appeal to displaced workers in our own politics.

Fears of job loss from new technology aren’t irrational. But trying to resist the tide can smother the benefits of new tech while making any eventual adjustment harder. Still, politicians can’t help themselves. Last year, in opposition, Labour’s Louise Haigh warned that automated vehicles could repeat “the ravages of deindustrialisation”. That’s Labour-speak for “government intervention required”. The tool of choice, of course, is typically “safety” regulation.

This instinct isn’t new. In 1865, parliament passed the Locomotive Act, capping self-propelled vehicle speeds at 2mph in towns and 4mph in the countryside, while requiring each be preceded by a man waving a red flag. Branded as a safety measure, it was backed by the horse-drawn carriage and rail lobbies. Maybe it spared a few pedestrians, but it certainly stunted any early car industry. Innovation was sacrificed to protect incumbents, with safety the excuse.

The US mis-stepped too. For decades, cities required human lift operators even after automatic elevators became safer and cheaper. A few thousand jobs were saved, with higher costs for building owners and slower productivity in sectors based in tall buildings. Like red flag laws, the aim of this safety regulation was really to preserve jobs in aspic.

The Hawley instinct would repeat this error. Trials of autonomous vehicle systems like Waymo in the US show 70–90 per cent reductions in crash rates compared with human drivers. In the UK, where human error is a factor in 88 per cent of collisions, industry modelling suggests autonomous vehicles could save 3,900 lives and prevent 60,000 serious injuries by 2040, with just 20-24 per cent market penetration. Studies predict less congestion, lower fuel use, and cheaper deliveries and logistics. And then there’s time freed. The average Brit spends over 120 hours per year behind the wheel, which is ripe for reclaiming for work or leisure. All these efficiencies would boost downstream sectors.

Yet the public isn’t sold. Polls show most Britons still consider driverless cars unsafe. That gives our own Hawleys an opening. Regulation to slow innovation always attracts both well-meaning safety worriers and affected workers with political allies. Labour has passed enabling legislation for automated vehicles, but seems to be dragging its feet on rollout. Meanwhile, the diffuse benefits of improved safety and efficiency get downplayed and delayed.

As AI spreads through the economy, such political battles will proliferate. Someone will always stand ready to highlight the freak accident, the displaced worker or the imperfect chatbot. And certain politicians will promise to protect workers affected.

That’s why it falls to those not wedded to the status quo to state the obvious: yes, automation is disruptive. But it’s also the path to safer roads, cheaper goods, and higher productivity. Hiding behind a modern red flag is no answer.

Progress that is delayed amounts to progress denied.

This article was originally published by The Times on 9/17/2025.