Originally posted at The Daily Economy. 

​For much of the last half-century, California benefited from a powerful first-mover advantage. Dense networks of talent, capital, and research institutions allowed the state to absorb policy mistakes that would have crippled competitors. High spending and taxes, restrictive housing rules, and regulatory complexity were treated as nuisances rather than binding constraints, because growth could outstrip their costs.

That margin of error has narrowed dramatically.

What California is now experiencing is not a cyclical tech downturn or a post-pandemic anomaly. It is a measurable, policy-driven decline in relative competitiveness. The most important evidence is not that tech employment has fallen in absolute terms, but that California’s share of national tech employment has been shrinking, while other states gain ground.

Markets are responding to incentives exactly as economic theory predicts.

Employment Share, Not Headlines, Tells the Story

According to Bureau of Labor Statistics Current Employment Statistics data, California’s technology employment growth has underperformed national trends for several years, including during periods when tech hiring stabilized or rebounded elsewhere, and recently has been declining. California’s share of US tech jobs is falling from roughly 19 percent pre-2020 to closer to 16 percent in recent years, a nontrivial shift for an industry this large.

This is a classic example of relative decline. California still employs more tech workers than any other state, but it is no longer where the marginal job is being created.

Commercial real estate data corroborate the employment figures. Office vacancy rates across Silicon Valley remain elevated well beyond what remote work alone would explain. Bay Area office markets have not recovered in the way peer regions have. Persistent vacancies signal not just a shift to hybrid work, but geographic reallocation of firms and labor.

Migration as a Labor Market Signal

Labor mobility reinforces the same conclusion. US Census state-to-state migration data show continued net domestic outmigration from California, particularly among working-age adults. While international immigration partially offsets population losses, domestic migration is more relevant for employer location decisions, especially in high-skill sectors.

Economic theory predicts that firms follow labor when relocation costs are low and regulatory frictions are high. California now faces both: high regulatory frictions at home and increasingly credible substitutes elsewhere.

Founding Versus Scaling: A Crucial Distinction

California still dominates early-stage venture capital totals, as shown in venture investment data. This is often cited as evidence that concerns about the state’s competitiveness are overstated. That interpretation conflates firm formation with firm expansion.

Founding activity reflects legacy advantages such as universities, networks, and capital concentration. Scaling decisions reflect marginal costs. Increasingly, firms are choosing to incorporate or raise seed funding in California while expanding headcount in lower-cost, lower-regulation states.

From an economic standpoint, this is predictable. Scaling in California exposes firms to the nation’s highest marginal income tax rates, comparatively punitive capital gains taxation, rigid labor mandates, slow permitting processes, and volatile regulatory expectations. These costs rise nonlinearly as firms grow.

AI Regulation as a Binding Constraint

Artificial intelligence policy may become the clearest illustration of California’s regulatory overreach.

A recent CalMatters analysis documents how California lawmakers have pursued some of the most expansive state-level AI regulations in the country. These proposals extend liability, mandate preemptive risk assessments, and impose compliance obligations before alleged harms are empirically demonstrated or even defined.

From an economic perspective, this approach treats innovation as a presumptive externality rather than a productivity-enhancing input.

AI is widely understood as a general-purpose technology. Research shows that such technologies generate broad, economy-wide productivity gains, not sector-specific benefits. Overregulating AI therefore depresses expected returns not only in software, but across healthcare, logistics, manufacturing, finance, and education.

California’s AI regulatory framework has drawn federal scrutiny, which is instructive. As noted in CalMatters, state-level AI mandates were referenced in Trump’s recent presidential executive order, citing concerns over fragmented and inconsistent state regulation. Regardless of political framing, the economic concern is straightforward: regulatory fragmentation raises fixed costs and discourages upscaling.

Regulation, Market Structure, and Incumbency

California’s regulatory posture also has implications for market structure. Extensive empirical literature shows that high fixed compliance costs reduce entry and increase concentration. The OECD’s work on regulation and competition consistently finds that heavier regulatory burdens favor large incumbents at the expense of startups and challengers.

This dynamic undermines the very competition that drives innovation. Europe’s experience with digital (over)regulation offers a cautionary parallel, acknowledged even in European Commission competitiveness reports. California risks reproducing that outcome domestically, exporting innovation to other states rather than other continents.

Costs Complete the Incentive Structure

AI regulation is best understood as the marginal constraint layered atop an already expensive environment. California has the highest top marginal income tax rate in the United States, and taxes capital gains as income. Housing scarcity, documented extensively by UC Berkeley’s Terner Center, raises labor costs without increasing real purchasing power. Energy prices remain among the nation’s highest, as shown by EIA electricity price data.

In combination, these policies alter the expected return on investment at the margin. States like Texas and Florida offer credible alternatives: no personal income tax, faster permitting, lower housing costs, and a lighter regulatory touch. 
Firms do not need ideological motivation to relocate. The incentive structure does the work.

Opportunity Costs and Distributional Effects

The economic cost of tech job relocation extends beyond headline employment figures. When tech employment relocates, these spillovers disappear as well. The distributional consequences are regressive. High-skill workers are mobile. Lower-income workers tied to local economies are much less so. Policies that suppress growth (even under the banner of equity) often hurt the poor most.

A Predictable Outcome

Unless California changes course, the trajectory is clear. AI firms will incorporate elsewhere. Venture capital will follow labor. Scaling will increasingly occur in states that treat innovation as an asset rather than a liability.

California will remain an important source of ideas. It will be a diminishing source of jobs. Markets are not ideological. They respond to incentives. On that front, the verdict is already in.

Originally published on Substack.

Every serious conversation about AI and jobs eventually runs into the same fear: creative destruction.

The phrase sounds ominous, which is exactly why it’s so often misunderstood. Destruction gets all the attention. Creation does the real work.
​
The latest analysis from the Federal Reserve Bank of Dallas shows modest employment declines among young workers in occupations with high AI exposure.

Predictably, this could reignite calls for intervention, protection, and preemptive regulation. But zoom out, and the real story is far more optimistic—and far more important for America’s economic future.

Creative Destruction Is How Progress Happens
​
AI is not unique. It fits squarely into a long history of productivity-enhancing innovations that initially disrupt specific tasks and roles before expanding opportunity across the economy.

What gets lost in today’s debate is that creative destruction is inseparable from gains from trade.

When AI allows workers and firms to specialize in what they do best, resources shift toward higher-value uses. That transition can feel uncomfortable at the margins, especially for young workers entering the labor force.

But the payoff is higher productivity, lower prices, better products, and—over time—higher real wages.

That’s not theory. It’s economic history.

When ATMs spread, bank teller jobs didn’t vanish. They evolved. When spreadsheets replaced paper ledgers, accounting didn’t disappear. It expanded. When the internet automated information retrieval, entire industries emerged that no regulator had the foresight to design.

AI is doing the same thing—faster, yes—but not fundamentally differently.

What the Dallas Fed Data Actually Tell Us

The Dallas Fed’s findings reinforce this point. There’s no surge in layoffs. Job-finding rates among unemployed workers look broadly similar across AI exposure levels. The main change is fewer young workers moving directly into certain AI-heavy occupations from outside the labor force.

That’s a classic adjustment pattern.

Young workers are exploring alternatives, building skills, and reallocating their time toward new opportunities. That is exactly how a dynamic labor market should respond to technological change.

The aggregate effects remain small, and the authors themselves acknowledge the uncertainty around causality.

In other words: this is early, subtle, and far from a crisis.

Innovation, Competition, and Abundance at Home

Where this matters most is competitiveness.

AI-driven productivity gains don’t just raise domestic living standards. They strengthen America’s ability to compete globally--especially with China.

Slowing innovation at home through precautionary regulation doesn’t protect workers. It weakens them.

Open, competitive markets encourage diffusion of AI across firms, not just among a handful of incumbents. That diffusion lowers costs, increases output, and expands abundance for consumers. It also forces firms to compete on quality and efficiency rather than political connections.

The irony is that many policies sold as “worker protection” end up entrenching market power, reducing entry, and narrowing opportunity—particularly for young workers and startups.

The Policy Crossroads We’re Facing

To its credit, the Trump administration has been surprisingly constructive on AI in key respects. A lighter-touch federal posture has helped avoid a fragmented state-level patchwork and allowed innovation to proceed with more certainty than many expected.

But there’s a real tension brewing.

Aggressive antitrust postures—especially those that resemble the Joe Biden era’s reflexive crackdowns on “big business”—risk undoing that progress. Size alone is not evidence of harm.

Market power that emerges from innovation is fundamentally different from power granted or protected by regulation.
The real pro-competition agenda is not breaking up successful firms for political satisfaction. It’s removing barriers that prevent new firms from challenging them.

That means fewer regulatory obstacles, clearer rules, and a bias toward entry and experimentation rather than enforcement theater.

Why Markets Outperform Central Planning—Again

No regulator can predict which AI applications will succeed, which jobs will evolve, or which new roles will emerge. Markets discover those answers through decentralized decisions made by millions of workers and firms responding to prices, profits, and consumer demand.

Trying to steer that process from the top down doesn’t reduce disruption. It amplifies it—by slowing adaptation and locking resources into declining uses.

The Dallas Fed study doesn’t justify panic. It reinforces restraint.

Bottom Line

AI will destroy some tasks and create many others. Young workers will feel the adjustment first, as they always do. The gains from trade and innovation will follow, as they always have.

The choice facing policymakers is simple: trust the market process that has delivered generations of rising prosperity—or substitute it with rules that protect the past at the expense of the future.

If we want abundance at home and strength abroad, especially in competition with China, the answer isn’t more control. It’s more competition.

Originally published on Substack. 

​The push to ban social media for minors is back in Texas, and it’s being sold as a moral imperative.

In a recent Dallas Morning News op-ed, State Rep. Jared Patterson (R) argues that Texas should follow Australia’s lead by restricting minors’ access to social media, framing the issue as a matter of urgency in response to rising youth mental-health challenges. His column, “What Australia’s social media ban means for Texans”, is thoughtful in tone and sincere in concern—but deeply flawed in its assumptions and conclusions.

Parents’ concerns are real. Patterson is right about that. Online predators exist. Some platforms may be addictive by design. Screens can crowd out sleep, exercise, and real-world relationships. These are legitimate issues.

But acknowledging real risks does not justify handing broad new powers to the state. The leap from “this is hard for parents” to “government must ban or condition speech” is neither conservative nor constitutional.

Patterson’s argument relies heavily on two pillars: the work of Jonathan Haidt and the example set by Australia. Both deserve closer scrutiny.
​
Start with Haidt. His work treats social media as a primary driver of youth anxiety and depression. But his evidence shows correlation, not causation. Mental health outcomes are influenced by many things like family stability, academic pressure, economic stress, sleep deprivation, social isolation, and lingering effects of pandemic policies. Social media may exacerbate some problems, but elevating it to the dominant cause oversimplifies reality and invites blunt policy responses.

​More troubling, Haidt’s framework subtly shifts responsibility away from parents and toward government. Kids are cast as helpless. Parents are portrayed as overwhelmed. The state becomes the fixer. That worldview is incompatible with a society built on personal responsibility and limited government.

Then there’s Australia, which Patterson presents as a model worth emulating. That comparison should immediately raise red flags. Australia does not recognize free speech as a fundamental constitutional right in the way the United States does, or the even stronger protections found in Article I, Section 8 of the Texas Constitution. Policies that may be legally permissible there face serious constitutional barriers here.

More importantly, Australia’s recent history shows how readily its government defaults to control. The same country now praised for “protecting kids online” imposed some of the longest and most severe COVID lockdowns in the democratic world, enforced through surveillance, travel bans, and coercive policing. Texans who value liberty should be deeply skeptical of importing policy ideas from a system that so casually set freedom aside.

In Texas, this debate played out during the 2025 regular session through House Bill 186, which proposed a categorical ban on social media for minors regardless of parental consent, maturity, or purpose. Thankfully HB 186 died, and that outcome matters. It reflected legitimate concerns about constitutionality, parental authority, and the dangerous precedent of state control over lawful speech.

Yet the push didn’t stop. Senate Bill 2420 passed using a different but equally troubling route by regulating app stores and forcing age verification on everyone, not just minors. That means adults, too, would have to hand over personal or transactional data simply to access lawful speech. This is how liberty erodes—not always through outright bans, but through “safety” systems that normalize surveillance.

Courts are beginning to recognize this danger. A federal judge recently blocked Louisiana’s social media age-verification law, finding that conditioning access to lawful speech on identity verification likely violates the First Amendment. Similar laws across the country are being rightly halted for constitutional violations. Good intentions do not override constitutional limits.

What Patterson’s column also overlooks is the opportunity cost imposed on kids. Like it or not, today’s world is digital. Social media is not merely entertainment. It’s how young people build communication skills, form networks, explore interests, organize civically, and prepare for modern careers.

Blanket restrictions don’t eliminate risk; they restrict access to opportunities that matter for future success.

This is especially inconsistent in Texas, where lawmakers rightly argue that parents—not bureaucrats—should control education. If parents can be trusted to choose schools, curricula, and values, they can be trusted to manage screens.

There are economic and privacy costs as well. Mandatory verification systems impose compliance burdens that smaller platforms struggle to absorb, reducing competition and innovation. At the same time, they expand data collection and increase the risk of breaches and identity theft.

Here’s the bottom line: we need more responsible parents, not more irresponsible government, in kids’ lives. Legislation like HB 186 and SB 2420 moves us in the wrong direction. It weakens parental responsibility by substituting state control, all while expanding government power in ways that should concern Texans across the political spectrum.

Closing

Rep. Patterson is right to care about kids. Where his argument goes wrong is in assuming that state mandates, speech restrictions, and surveillance-based enforcement are the answer. Texas does not need to follow Australia. It needs to uphold its own constitutional traditions—trust families, respect liberty, and resist the urge to regulate away complex social problems with blunt government force.

Call to Action

If you found this valuable, please subscribe, like, share, and leave a comment. These conversations matter, and the only way to protect liberty is to defend it—clearly, consistently, and without apology.

Originally published on Substack. 

Artificial intelligence isn’t an existential threat. It’s not magic, and it’s not going to swallow society whole. AI is simply advanced computing, the next logical extension of tools humans have been building for thousands of years.

Yet politicians—especially here in Texas—are treating AI as if it’s a radioactive substance that must be contained through sweeping regulation. State Senator Angela Paxton and a few colleagues recently urged Congress to reject a federal “moratorium” on state AI laws, claiming Texas has taken “important steps” to protect children and consumers. Their letter and justification can be seen in her tweet.

I understand the desire to protect kids. I’m a dad of three. But overregulating technology out of fear—and surrendering parental authority to capitols—only creates bigger problems. The Texas AI laws passed last session weren’t “responsible.” They were a case study in correlation-over-causation thinking, moral panic, and political control dressed up as child protection.

This is why a federal pause, though not ideal, is needed.

Texas Isn’t Protecting Families. It’s Expanding Government.
​
Texas’s AI bills created vague liability standards, broad mandates, and bureaucratic authority with almost zero grounding in sound economics or constitutional limits. These laws shift decision-making away from parents and toward politicians and regulators—the opposite of what a free society demands.

Illegal acts using AI—fraud, exploitation, child endangerment—are already illegal under decades of existing statutes. We punish outcomes, not inputs. We don’t ban cars because someone might speed. We don’t ban pencils because someone wrote something awful.

Yet Texas is regulating AI as if every computing tool is a pre-crime device.

This is not classical liberalism or the limited-government Texas Model I defend in my work at Ginn Economic Consulting and my research on economic freedom. It is paternalism that undermines parents, weakens competition, and slows innovation.

The Precautionary Principle Isn’t Prudence—It’s Paralysis

Texas lawmakers are leaning heavily on the flawed precautionary principle: regulate now “just in case.” But public policy driven by hypothetical harm consistently produces:

• fewer choices
• slower innovation
• higher costs
• more concentrated power
• and worse outcomes for the very people policymakers claim to protect

If applied historically, this principle would have stopped electricity, airplanes, automobiles, calculators, and the internet. Every transformative technology looked scary before it became essential. AI is no different—unless government freezes it in place.

Freedom—not fear—is the best safeguard.

AI Is a Complement, Not a Substitute, If Politicians Let It Be

Much of the panic ignores basic economics. Automation enhances human productivity. It creates new jobs, new opportunities, and new industries—as I’ve explained repeatedly in my work on economic growth and technology.

AI can empower:

• teachers with better tools
• doctors with sharper diagnostics
• parents with improved monitoring and safeguards
• small businesses with capabilities once limited to Fortune 500 firms
• entrepreneurs with lower startup costs

Politicians frame AI as a threat to children and jobs. In reality, political overreach is the threat—not the technology.

Why a Federal Moratorium Is Necessary—for Now

I’m no fan of Washington micromanagement. But when states begin passing contradictory, constitutionally questionable laws that strangle interstate commerce and innovation, a temporary federal pause becomes the lesser evil.

Otherwise, the U.S. will repeat the disaster of California’s CAFE standards, where one state distorted the entire nation’s auto market and made cares more expensive for everyone. A patchwork of 50 incompatible AI regimes would be even worse.

AI isn’t a local zoning matter. It’s core to:

• national security
• interstate commerce
• global economic leadership
• the future of work
• innovation and entrepreneurship

A short-term federal moratorium is not about controlling AI—it’s about preventing states from crippling innovation before the country fully understands the technology.

Meanwhile, China Isn’t Slowing Down

While Texas and other states push fear-based restrictions, China continues racing ahead in AI and robotics. But as the Wall Street Journal recently noted in a recent piece on China’s robot boom, the appearance of progress masks deep structural problems. Their surge in automation reflects demographic decline, state coercion, and misallocated capital.

Still, Beijing is moving fast, and America will not win by second-guessing ourselves to death.
As I argued in my recent commentary on China’s misguided industrial push, America’s strength isn’t central planning—it’s free people, free markets, and open competition.

We don’t beat China by copying their control.

We beat China by unleashing our creativity.

My Take

Texas lawmakers mean well, but they’re wrong. Overregulating AI today will harm:

• families
• entrepreneurs
• small businesses
• national competitiveness
• the very children they claim to protect

The federal moratorium is a temporary brake on a runaway train of fear-based policymaking. Texas should use this moment to peel back its flawed AI rules and return to the pro-growth, limited-government principles that once made this state the envy of the world.

AI is part of our future.

Liberty—not regulation—is the best way to shape it.

Let’s get back to trusting people, empowering parents, and letting innovation drive prosperity.
​
That’s how we let people prosper—even in the age of AI.

Originally posted on Substack. 

​A federal court just handed Meta a decisive win—and Washington a badly needed wake-up call.

By dismissing the FTC’s high-profile antitrust case, Judge James Boasberg didn’t simply reject one lawsuit. He exposed a deeper problem: too many in government and across the political spectrum believe competition can be engineered from above, rather than unleashed from below.

This ruling should be a turning point. If policymakers are serious about promoting vibrant, open markets, they must stop trying to centrally plan outcomes and start clearing away the political choke points that smother competition long before any tech company does.

​The FTC’s failed theory—that Meta still holds an illegal monopoly despite facing a swarm of rivals—was built on a narrow, static view of a dynamic market. The agency argued Facebook and Instagram are fundamentally different from TikTok because they revolve around “friends and family.” That argument collapsed the second the evidence hit the courtroom. As the judge noted plainly, TikTok “holds center stage as Meta’s fiercest rival.”

Consumers already made their choice. The government just didn’t like it.

It’s a pattern: where Washington sees monopolies, the real world shows competitive pressure pushing firms to evolve. Meta didn’t maintain dominance by freezing the market—it had to reinvent its platforms to keep up with short-form video, algorithmic feeds, and shifting consumer expectations. That’s why Americans now spend only a fraction of their Facebook time scrolling through updates from their actual friends. The product changed because the competition demanded it.

This is precisely why a healthy economy requires more market entry and less regulatory overreach, not the opposite. When the government defines markets however it chooses, rewrites history, and pursues lawsuits disconnected from consumer realities, the result is fewer new competitors—not more. Investors pull back. Startups get hemmed in. Acquisition pathways shrink. And the entire innovation ecosystem slows.

Some on the right have fallen for this too, using “monopoly” as a catch-all insult rather than a legally meaningful term. But once you anchor antitrust to actual metrics—prices, quality, innovation—Meta doesn’t fit. Its services are no charge. Its ads can be ignored. Its products face constant substitution threats. Its acquisitions (approved originally by the FTC) enhanced consumer value.

The court recognized that reality. And it’s a reminder that the consumer-welfare standard remains the only reliable compass in antitrust law. It measures harm the way economists do—not the way political movements do.

The alternative—the neo-Brandeisian approach championed by Chair Lina Khan—tries to punish companies for being large, effective, or popular. That theory collapses when confronted with actual evidence of competition, as Brian Albrecht pointed out in his analysis. But the danger isn’t just bad lawsuits. It’s the chilling effect on innovation, investment, and the next wave of startups wondering whether success will simply put a target on their backs.
​
If Washington truly wants to support competition, it should start by removing the barriers it created:

• Stop politicizing antitrust. Competition happens bottom-up, not by bureaucratic design.
• End selective favoritism and carveouts that tilt markets. Subsidies, tax breaks, and regulatory privileges distort competition more than any merger does.
• Focus on voluntary exchange and consumer choice. Not on reshaping the economy according to ideological preferences.
• Preserve the consumer-welfare standard. It’s the guardrail that prevents politics from replacing economics.

Meta’s win doesn’t mean the tech sector is perfect. It means the real monopoly threat still comes from government—utilities, water districts, licensing boards, tax-favored entities, and agencies whose incentives align with control, not competition. Breaking that grip would do more for market dynamism than any forced corporate breakup dreamed up in D.C.

The path forward is straightforward: more competition driven by people, fewer decisions dictated by government. This ruling doesn’t end the antitrust debate—but it should reset it. And it’s long past time.

​Sources and further reading.
​
WSJ reporting on Meta’s court victory: https://www.wsj.com/us-news/law/meta-defeats-ftcs-antitrust-case-alleging-social-media-monopoly-504b2323

My X thread: https://x.com/vanceginn/status/1990878777795359214?s=46&t=Zv07DS2UC3mLPAOmcJxg_Q
​
Brian Albrecht’s X thread: https://x.com/briancalbrecht/status/1990866831553310762?s=46&t=Zv07DS2UC3mLPAOmcJxg_Q

NetChoice statement: https://x.com/netchoice/status/1990867855668068534?s=46&t=Zv07DS2UC3mLPAOmcJxg_Q