California's Anchor Firms Are Hiring. Just Not in California.

California has missed out on up to a quarter million jobs created elsewhere by established firms, and likely more from startups that chose other states. A wealth tax would make it even worse.

As California’s proposed wealth tax moves closer to the ballot, the debate has settled into a familiar exchange.

Supporters ask why billionaires shouldn’t pay more. Opponents reply that billionaires can leave. Defenders of the tax then point out that the companies are largely still here. Tesla still builds cars in Fremont. Oracle still operates a sprawling Bay Area campus. Charles Schwab still has a meaningful presence in San Francisco. If a founder buys a house in Austin or Miami, doesn’t the rest of California just carry on?

No. In fact, the argument that firms still have a presence in California is speaking to the wrong question. Global demand for the products these California-anchored firms sell has skyrocketed over the past decade, yet California's share of the jobs that growth produced has not. The question is how much hiring still lands here to the benefit of Californians, and how much will land here over the coming decades.

California’s Anchor Public Firms

We tracked US and California employment from 2015 through early 2026 at 23 of California's largest publicly traded employers, using workforce data from Revelio Labs. We selected the panel on three criteria: large publicly traded firms with material California operations at some point in the period; either currently or formerly headquartered in California, or anchored by a significant California presence; and sufficient firm-level Revelio coverage to produce reliable annual headcount estimates.

The panel spans tech, semiconductors, hardware, telecom, finance, and energy: Alphabet / Google, Amazon (incl. AWS), Apple, Applied Materials, AT&T, Broadcom, Charles Schwab, Charter Communications, Chevron, Cisco Systems, Hewlett Packard Enterprise, IBM, Intel, Intuit, Meta Platforms, Microsoft, NVIDIA, Oracle, Qualcomm, Salesforce, ServiceNow, Tesla, and Western Digital.

Of these 23 anchor firms, thirteen remain California-headquartered in 2026, five had significant employment presence in California but were never headquartered in the state, and five departed during the decade.

Most of California's high-growth private firms — Stripe, Databricks, OpenAI, Anthropic, SpaceX — fall below the depth of Revelio coverage required, hence our general exclusion of non-public firms. Several recently public firms with sparse pre-IPO histories are also excluded. The most consequential exclusion is Palantir, which moved its headquarters from Palo Alto to Denver in 2020 and has announced a further move to Miami in 2026. Including firms like Palantir, or the broader cohort that was never California-headquartered and now hires at scale in Texas, Washington, and Florida, would widen the gap we document.

Declining Job Shares, Missed Opportunities

Together, the 23 anchor public firms employed 1.62 million people in the United States in 2026, up from 1.09 million in 2015 — a national expansion of about 49 percent. California headcount at these firms also grew, from roughly 301,000 to roughly 390,000.

But California’s share of the US workforce fell significantly, representing lost opportunities for job creation that other states enjoyed.

On a headcount-weighted basis, the 23 firms employed 27.6 percent of their US workforce in California in 2015 and 24.1 percent in 2026. On an equal-weighted basis — which prevents national giants like Amazon from dominating the average — the decline is steeper, from 39.5 percent to 34.6 percent. Figure 1 shows both series falling steadily across the period, with the sharpest drop between 2019 and 2023.

Hiring kept happening, but most of it happened outside California.

Holding each firm's 2015 California share of US employment constant through 2026 and applying it to actual 2026 US headcount yields a counterfactual California workforce for these anchor public firms of roughly 453,000 — about 62,900 jobs larger than the realized figure, or 16 percent above today’s actual. Figure 2 visualizes the gap between US and California job growth at these firms.

Importantly, this assumes that the US total is identical in the actual and counterfactual columns: we hold each firm’s realized 2026 US headcount fixed and ask only where those jobs were sited. This implicitly assumes that California’s policies did not change the firms’ overall US job growth, or the extent to which they offshored jobs abroad. If California’s cost and regulatory environment pushed some of these firms to expand overseas rather than in another US state, the true domestic loss would be larger than the counterfactual shows, and our estimate is again conservative.

The California jobs gap varies among firms. It is strongest in Alphabet, Meta, Tesla, Amazon, Salesforce, Cisco, Intel, and Oracle. Three firms run the other way: Apple, whose California share has risen since 2015, accounts for roughly 2,600 more California jobs than the share-constant counterfactual implies, with NVIDIA and IBM contributing small positive offsets of the same sign.

Imputing Beyond the Sample

The 23 anchor public firms employ only 2.2 percent of California’s 18.0 million nonfarm workers. Their 62,900 California job gap is not small, but whether it is meaningful at the state level depends on whether the within-firm share decay we observe extends to firms and sectors outside the panel.

Within-sector extrapolation suggests a much larger total. Our 23 anchor firm panel concentrates on two BLS supersectors. Eleven of the 23 firms are classified primarily in Information Technology — software publishing, internet platforms, telecom — the panel alone accounts for about 33 percent of California’s entire Information workforce. Nine classify in Manufacturing — computer and electronic products, motor vehicles, where the panel covers about 11 percent of the California sector.

If the employment dynamics in the remaining firms in these two sectors mirror the dynamics in the anchor firms, then the implied sector-wide shortfall is roughly 234,000 jobs. This is a very plausible and indeed conservative estimate, as the firms that did not make it to anchor-firm size include exactly the relocating and never-Californian cases that would tend to deepen the decay. Furthermore, we deliberately do not gross up the three single-firm sectors in the panel (Amazon in trade and transportation, Chevron in mining, Schwab in financial activities), as one firm is too thin a basis on which to extrapolate a sector — but these sectors may well have displayed similar dynamics.

Two further channels remain entirely outside the data. The first is the sectors our panel does not touch at all. The second, and likely larger on a prospective basis, is the firms that in an earlier era might have been founded in California but instead chose Austin, Miami, Seattle, or Denver from inception — a loss that never appears in any California headcount series because those jobs were never here to begin with. While our data do not allow us to quantify these channels, any reasonable accounting adds significantly to the total.

In sum, the estimated within-firm effect alone is a quarter-million jobs in the two sectors we can measure; the full state-level number is larger by an unknown amount.

Share Decay and Founder Migration

The fact that California’s share has declined does not, by itself, identify a cause. Many of the steepest decliners neither lost their founder nor moved their headquarters. Meta’s California share fell from 59 percent to 44 percent with Mark Zuckerberg residing in Palo Alto the entire time, although since February 2026 he is now in Miami. Alphabet, Salesforce, and ServiceNow show similar trajectories without an HQ event, although the founders of Google did leave California in late 2025.

However, for a subset of the panel, founder relocation and headquarters migration did occur during the sample period, and those cases reveal the mechanism by which California loses share when founders leave. Specifically:

Tesla. Elon Musk moved to Texas in late 2020. Tesla redesignated its headquarters from Palo Alto to Austin the following year. Fremont remains a critical manufacturing asset, and Tesla’s California headcount rose in absolute terms over the period. But California’s share of Tesla’s US workforce fell from 57 percent in 2015 to 44 percent by 2026. The Gigafactory, the AI compute infrastructure, the executive presence — Tesla’s marginal investment increasingly anchors in Texas.

Oracle. Larry Ellison left for Hawaii in 2020. Oracle redesignated its headquarters to Austin the same year and later announced a further move to Nashville, paired with 8,500 promised Nashville jobs and a $28 billion acquisition of Cerner. California’s share of Oracle’s US headcount fell from 25 percent to 21 percent. The California footprint persists, but the strategic center of gravity does not.

Palantir. Palantir is one of the firms for which there was insufficient data to enter our panel, but news reports are instructive. Peter Thiel relocated to Florida in March 2024. Palantir, which had already moved its headquarters from Palo Alto to Denver in 2020, has announced a further move to Miami in 2026. The most commonly cited driver is regulation rather than tax rates, but the sequence is the same one visible at Tesla and Oracle: founder relocates, headquarters follows within 12 to 36 months, major campus and acquisition commitments land in the new state, and marginal hiring tilts toward the new location while the existing footprint is preserved.

Most of the decline in California’s employment share of the anchor firms reflects broader forces — most importantly housing costs, regulatory burden, and offshoring to India. The migration of founders and headquarters is one channel, and the one most directly responsive to tax policy.

What this means for the wealth tax

Research earlier this year by Rauh, Jaros, Kearney, and Doran found the proposed billionaire tax would not generate an increase in the net present value of resources for the state government. The study found a point estimate effect of negative $24.7 billion on state coffers, once lost income-tax revenue from departing billionaires is netted against projected wealth tax collections.

The employment lens adds a longer-horizon channel that the NPV calculation on individual tax payments does not incorporate. The proponents’ implicit assumption that firms stay put even when founders leave is incorrect. “The firm” is not a single decision. Companies rarely leave all at once. What moves first is the next round of hiring, expansion, and executive functions. In the cases where we observe a founder departure, those activities increasingly follow the founder out. A wealth tax landing on top of an already-elevated set of costs faced by firms and their founders would accelerate that sequence.

California’s high-net-worth residents and their firms have already been actively recruited, well before the wealth tax. Tennessee gave Oracle $65 million. Texas has built much of its modern economic-development apparatus around exactly these types of relocations. Palantir’s announced Miami move cited Colorado’s AI regulation. The wealth tax would be a compounding shock.

The tax base at stake includes not only the departing executive’s personal income, but also the incomes of thousands of high-earning employees who never get hired in California, along with the sales taxes generated by their spending.

But the issue is larger than tax revenue alone. The data suggest that when founders leave, future economic growth leaves with them. The logos stay. But the next hires, the next campuses, and the next acquisitions increasingly happen elsewhere. Californians lose the upside of the companies they once built.

---

Source note: Workforce data underlying the panel analysis is from Revelio Labs and covers 23 California-headquartered or California-anchored publicly traded firms across technology, telecom, semiconductors, energy, and financial services. Headquarters and executive-relocation dates are drawn from company 10-K filings, annual reports, and contemporaneous news coverage. California nonfarm and sectoral employment totals come from the BLS Current Employment Statistics survey, December 2025 (most recent available).