Published November 18, 2026 · Financial and regional evidence through October 25, 2026. Scenarios are illustrations. Sources and assumptions

Housing · Work · Washington

What Holds Up Kirkland Housing Prices

A desirable place and a defensible purchase are different propositions.

An imagined lakeshore home is supported by workplaces, transport and local services.
Editorial illustration · AI-generatedThe place is desirable. The price still depends on the foundations beneath it.
Explore this essay · 10 sections

A million-dollar neighborhood can remain desirable long after an ordinary salary stops making its purchase price intelligible. The buildings have not suddenly become more useful. The water, schools, employment access, and familiar streets may look much as they did before. What changes is the financial system surrounding them: who arrives with equity, who receives stock compensation, who can borrow, and how few homes need to trade to establish the next comparable sale.

Kirkland makes that distinction difficult to ignore. The question is larger than whether a particular home seems overpriced. It is whether the income, wealth, employment, and public services supporting its price are durable—and whether a new buyer can withstand the conditions under which that support weakens.

That is the problem I want my dashboard work to address: make consequential decisions inspectable. The connection between my NorthlineIQ and separate REAL-IQ work is methodological: turn persuasive stories into observable conditions, then make room for evidence that contradicts them.

Washington offers several persuasive stories. Wealthy residents are leaving because of taxes. Technology jobs make Eastside housing permanently secure. Remote work is draining value out of the region. Restricted supply makes any price defensible. Each contains a plausible mechanism. None can substitute for tracing what is actually changing.

Start with three different departures

A wealthy person changing tax residence, a worker moving away, and an employer relocating jobs are separate events. They can happen together, but one does not establish the others. A founder can leave while a company keeps its employees in Washington. A remote employee can leave while retaining the same employer. A company can grow elsewhere without removing a single existing local job.

The distinction matters because the effects travel through different channels. A departing high-income household may change state tax exposure, local consumption, philanthropy, and housing demand. A departing job changes payroll geography. A departing headquarters may change neither in the proportion implied by its announcement. The relevant unit is the taxable income, employment, spending, or capital investment that actually moves.

Washington’s population was still growing in the latest state estimate reviewed here. The Department of Commerce’s July 2026 summary, citing OFM, put the April population at 8,176,300, up 61,200 over the year; net migration accounted for 72% of that increase. That is inconsistent with a simple story of statewide depopulation. It does not tell us which income groups moved, or whether domestic departures were offset by other arrivals. Washington Commerce population summary

Older IRS data expose why those questions must remain separate. In the 2022–2023 state-to-state files, Washington had 122,184 incoming domestic tax returns and 112,345 outgoing returns: a net gain of 9,839. Yet the reported adjusted gross income associated with those flows was about $10.98 billion inbound and $11.25 billion outbound. A state can gain filing households while losing associated reported income on net. These figures describe matched tax returns and year-two AGI, not every resident or a stock of accumulated wealth. IRS migration data and downloads

The IRS income-class file also shows 10,461 incoming returns and 12,518 outgoing returns in Washington’s $200,000-or-more AGI group. The net difference was 2,057 returns, with roughly $1.19 billion more associated year-two AGI outbound than inbound. That group is much broader than millionaires. It is a different table population from the state-to-state totals, and the IRS excludes negative-AGI returns from it. The figures should not be combined into a single supposedly exhaustive migration count. IRS gross-migration CSV, IRS definitions

This establishes a historical distributional concern. It does not establish that a tax enacted in 2026 caused movements recorded in 2022–2023. Nor does AGI attached to a mover equal income that would necessarily have been earned, realized, or taxed in Washington had that person stayed.

Two different lenses on migration · IRS 2022–2023

Domestic tax returns

+9,839 net incoming returns

122,184 in · 112,345 out

AGI of $200k or more

−2,057 net incoming returns

10,461 in · 12,518 out

The first is the domestic state-to-state table; the second is the gross-migration income-class table, which excludes negative-AGI returns. Different populations, not additive totals. $200k+ does not mean millionaire. Neither identifies why a household moved.

What Washington actually changed

The current debate has a specific legal timeline. Washington’s Department of Revenue says the new 9.9% income tax begins January 1, 2028, with the first returns due in 2029, for income above the statutory $1 million threshold. As of this essay’s cutoff, it is enacted future policy, not a wage tax already being collected from ordinary earners. Anticipation can influence decisions before a tax starts, but that possibility requires evidence of its own. Department of Revenue income-tax guidance

Separately, the capital-gains tax has a 7% base rate and an additional 2.9% on Washington capital gains exceeding $1 million, beginning with tax year 2025. These are distinct taxes with distinct bases, deductions, credits, and timing rules. Adding headline rates together does not produce a household’s effective tax burden. Department of Revenue tiered-rate notice

A house worth more than $1 million is not itself $1 million of annual income. The governor’s signing notice explicitly distinguishes the new income tax from a tax on assets such as homes. A resident’s exposure depends on income and applicable law, not on whether a neighborhood’s sale price has crossed a psychologically important number. Governor’s signing notice

The political disagreement concerns a genuine tradeoff. Higher taxes can reduce the after-tax advantage of living in the state, particularly for households with portable income or unusually large realizations. Public revenue can also support the schools, infrastructure, childcare, and services that help an employment cluster function. The governor’s office emphasizes that second channel; opponents emphasize migration, competitiveness, and uncertainty. The useful economic question is the net result, including implementation, service quality, behavioral response, and revenue volatility.

Research does not justify either absolute. Studies summarized by Cristobal Young find small tax-induced migration effects among the populations they examine. Research by Akcigit, Baslandze, and Stantcheva finds tax sensitivity among superstar inventors in international location decisions. Different people, tax systems, and relocation margins can produce different effects. Neither study supplies a ready-made forecast for Washington’s new law. Young’s research and papers, Inventor-mobility research

Testing the Washington claim requires more than collecting departure announcements. It requires inflows and outflows, income groups, timing, comparison regions, and controls for housing costs, retirement, stock-market changes, and the technology cycle. Tax-return data arrive with a substantial lag. We do not yet have a mature post-implementation record for a tax that begins in 2028.

Count jobs without inventing their causes

Washington’s September 2026 labor figures show a difficult but mixed picture. Seasonally adjusted nonfarm payroll employment was about 3.64 million, down 15,400 from August but up 8,000 from September 2025. The unemployment rate was 4.8%, compared with 4.7% a year earlier. The figures describe statewide employment conditions; they are not Kirkland job counts. Employment Security Department facts and figures

King County’s separate September report puts its payroll job base at 1,476,100. That remains a substantial concentration of employment even as conditions become more difficult at the margin. It is a county observation, not a Kirkland count, and its unadjusted unemployment measures should not be spliced into the seasonally adjusted statewide series. King County labor profile

That distinction changes the interpretation of layoffs. Gross separations can be severe for the people affected while employment elsewhere offsets some of the aggregate loss. A decline in one month is different from a year-over-year contraction. Neither measure, by itself, identifies jobs eliminated by AI, taxes, offshoring, or remote work.

WARN notices add useful evidence about announced reductions, but they are not a census of all job losses or proof that every announced separation occurred as originally scheduled. Their dates, locations, amendments, and potential overlaps must be reconciled. A dashboard should show verified notice totals separately from payroll changes and job openings. Assigning all of them to “AI disruption” would replace analysis with a label. Washington WARN requirements and database

The honest answer to “how many jobs are actually being disrupted?” therefore has several parts. We can quantify net payroll change. We can track announced layoffs where records permit. We cannot derive a reliable Washington-wide count of jobs causally displaced by AI or tax policy from those series. Vacancies never opened, slower hiring, contractor reductions, and changes in job content are also different forms of adjustment that headline layoff totals miss.

For housing, the composition matters as much as the count. Losing a concentration of highly compensated buyers can affect a particular price band even while the state adds jobs elsewhere. Conversely, a slowdown in hiring need not immediately force existing owners to sell. Their mortgages, accumulated equity, and other income influence the timing.

Remote work changes where value lands

Remote work separates a company’s address, a worker’s residence, the location of production, and the place where the paycheck is spent. A Washington company employing someone in another state can retain intellectual property and business income while some household spending moves away. A Washington resident working for an out-of-state company can import purchasing power. A suburban worker staying home can shift spending from a downtown lunch counter to a neighborhood business without leaving Washington at all.

These are distinct spatial changes. Downtown commercial weakness can coexist with resilient suburban demand. Employment can become more geographically distributed while valuable professional networks remain concentrated. In-person coordination, specialized labor, suppliers, and reputation can continue to support a cluster; remote hiring expands the set of alternatives.

There is regional evidence of a changed commute. PSRC’s November 2025 transportation report, drawing on its 2023 household travel survey, found that more than half of employees with workplaces in Bellevue and more than two-thirds of those with workplaces in Downtown Seattle were no longer commuting to the office every day. That is an earlier survey of work patterns, not a current count of jobs moved out of Washington. It helps explain why the geography of daily spending can change while employment relationships remain. PSRC transportation report

The latest reviewed output data also resist the collapse narrative. Washington’s Economic and Revenue Forecast Council reported second-quarter 2026 real GDP growth of 0.6% at a seasonally adjusted annual rate, below the national 2.2%; state personal income grew at a 3.4% nominal annualized rate. Slow growth and continued production are both visible. These figures do not isolate remote work’s causal effect, but they do not describe economic value simply disappearing from the state. Washington output and income report

The necessary measurements are residence-based income, workplace employment, commuting patterns, office utilization, and local business activity. Remote-work survey shares are not counts of jobs exported. A statewide headquarters count cannot tell us where all the household purchasing power lands. The Congressional Research Service’s review of remote work likewise treats productivity, workers, geography, and business effects as separate channels. CRS review of remote work

For Kirkland, the relevant risk is not simply that an employer permits working from home. It is whether the workers most able to bid for local housing continue to choose, or need, to live nearby—and whether their compensation supports the prices at which new transactions occur.

Expensive housing can lose seller power

The public Redfin snapshot retained in REAL-IQ covers all residential sales in Kirkland for the three months ending September 2026. Its median sale price was about $1.22 million, down a reported 9.67% year over year. Median selling time was 28 days; end-period inventory was 426 homes, with a reported 23.04% annual increase. About 36.49% of active listings had price reductions, and the average sale-to-final-list ratio was 97.72%. These describe a rolling three-month city window, not a single month’s detached-house market. Redfin Kirkland market page, Data Center definitions

The combination points toward weaker seller bargaining conditions, even though the price level remains high. It does not establish that every comparable house lost 9.67% of its value. Medians change when the mix of homes changes, and concessions are not fully represented by a sale-to-list ratio. Repeat-sales evidence and property-level comparables are needed before assigning that change to a particular home.

Everett provides a useful comparison that checks the temptation to narrate the entire region as one market. In the same retained window, its median price was also lower year over year, but selling time was shorter and sale-to-original-list performance firmer than the year-earlier observations. The dashboard’s reason for keeping cities separate is visible here: a shared macro environment can produce different local bargaining conditions. Redfin Everett market page

Layer of evidenceWhat it helps establishWhat it cannot establish alone
Income and migrationWho may carry purchasing power into or out of the regionWhy each household moved
Payroll and compensationEmployment demand and potential buyer incomeA property’s defensible price
Inventory and contract activitySupply relative to current demandFuture forced sales
Selling time and reductionsLiquidity and bargaining pressureA repeat-sales appreciation rate
Property cash flows and carrying costsThe economics of a particular purchaseThe buyer’s whole financial resilience
A high level can coexist with a weakening market
$0.76M2019
$1.01M2021
$1.44M2024
$1.22M2026

Kirkland all-residential median sale prices in selected July–September windows. Zero-based bars. Nominal, mix-sensitive medians—not a repeat-sales index. Public Redfin snapshot retrieved October 20, 2026. Definitions.

Conceptual model · different populations

Existing owners

Past purchase price
Accumulated equity
Existing financing

The next transaction

One willing buyer
One available seller
A new comparable sale

Transaction prices emerge at the margin. This diagram does not measure the share of buyers or owners in any category.

How the purchase price becomes possible

A market does not require every resident to afford its median sale. It requires enough marginal buyers to transact with the available sellers. Some arrive with two high incomes, stock proceeds, prior-home equity, family assistance, or substantial accumulated assets. Others rent. Existing owners may have bought at a lower price or financed at a lower rate. These mechanisms can sustain transactions without making a fresh purchase affordable to a typical wage earner.

Consider an illustrative $1.2 million purchase. A 20% down payment requires $240,000 before closing costs and reserves. Financing $960,000 for thirty years at 7.32% produces approximately $6,595 in monthly principal and interest. The rate is Freddie Mac’s October 22 national benchmark, not an individual loan offer. Freddie Mac historical mortgage survey

Assume property taxes of 0.9% of value annually and insurance of $200 a month. Those are modeling assumptions, not a parcel’s actual tax bill or an insurance quote. Principal, interest, taxes, and insurance total about $7,695 monthly. At a chosen 30% housing-to-gross-income planning ratio, that implies roughly $308,000 of annual household income. A 1%-of-value annual maintenance reserve adds $1,000 a month, lifting the corresponding planning income to roughly $348,000. HOA charges, utilities, closing costs, other debts, and childcare are outside the calculation. The 30% ratio is not an underwriting approval rule.

At a hypothetical 5% rate, the same principal would require about $5,153 monthly before taxes and insurance. That difference illustrates why prices can feel disconnected from the budgets of new buyers even when existing owners remain comfortable. The payment depends on the financing date as well as the house.

Multigenerational arrangements and shared housing can be rational responses to costs. The transaction data do not establish that crowding is how most buyers qualify, however. To test crowding requires occupancy and household data, not an inference from a listing price. The more firmly supported conclusion is that the market draws on wealth and income distributions that a single salary does not describe.

The entry cost · an illustrative new purchase
20% down payment$240k
Monthly PITI$7,695
Gross income at 30% PITI$308k/yr

Principal and interest: $6,595 / month. Maintenance reserve: $1,000 / month in addition.

Fixed assumptions: 30-year amortization, 20% down, property tax 0.9% of purchase price annually, insurance $200/month, maintenance reserve 1% annually. PITI excludes maintenance, HOA, utilities, closing costs and other debts. The 30% income ratio is a chosen planning assumption, not lender approval. Starting rate: Freddie Mac, October 22, 2026; individual terms differ.

What protects Kirkland and what leaves it exposed

The defenses are tangible: access to a deep regional employment cluster, established neighborhoods, limited desirable locations, and amenities that buyers may be willing to pay to keep. These qualities can preserve a premium relative to less connected places. In an MBA resource-based analysis, location and established networks can be valuable and difficult to reproduce.

But a premium is a relative advantage, not a floor under every nominal price. Technology compensation and stock-market wealth can move together. Mortgage rates can reduce bidding capacity without reducing a property’s appeal. Insurance, maintenance, taxes, and financing can erode rental economics. A highly desirable asset can still be purchased on terms that leave its owner vulnerable.

Supply policy also matters. Kirkland’s middle-housing work implements state requirements to allow additional dwelling types and greater residential capacity. More legal capacity creates possibilities; financing, infrastructure, development costs, and actual completions determine how quickly supply responds. A rezoning is not a completed home. It can also change the value of a specific lot differently from the value of the existing structure on it. Kirkland middle-housing materials

Support or vulnerabilityObservable test
Employment-cluster durabilityLocal payroll by industry, hiring, compensation, and employer expansion or contraction
Scarcity of desirable housingListings, permits, completions, and absorption by location and property type
Dependence on stock wealthBuyer-financing patterns and income composition, where credible public data exist
Ability to hold through a downturnCarrying costs, rent alternatives, debt terms, and liquid reserves
Policy competitivenessEffective tax burdens, public-service quality, business formation, and comparable migration cohorts

Leverage deserves its own test. On the illustrative purchase, a 10% price decline would cut initial equity from $240,000 to $120,000 before transaction costs and principal repayment. A 20% decline would erase that initial equity. Those are stress tests, not forecasts. Their purpose is to expose the difference between a durable location and a financially durable ownership position.

Why an unaffordable price can persist without being safe

The apparent contradiction becomes clearer if three propositions are kept separate. Kirkland can be a desirable location. A limited group can still afford its asking prices. A particular purchase can nevertheless offer poor financial terms. The first two explain how a market clears; they do not prove the third is attractive.

Mortgage lock-in is one mechanism. An owner with a low fixed rate may decline to sell rather than replace it with expensive financing. Federal Reserve research on the earlier rate-hike episode finds that lock-in reduced mobility and affected market tightness and prices. That is evidence of a mechanism in the studied markets and period, not an estimate of Kirkland’s present lock-in share or a guarantee that supply will stay constrained. Federal Reserve lock-in research

The result can be fewer sales rather than immediate price discovery. Asking prices are cheap to maintain when owners can afford to wait. Completed transactions carry more information, but a small group of wealthy buyers is not proof that the broader population can support the same price. If listings accumulate or sellers become forced to move, the balance changes. Scarcity is a condition to measure, not a permanent promise.

There is also a financial test independent of neighborhood prestige. In a hypothetical $1.2 million home renting for $4,000 a month, gross rent is $48,000 a year. Subtract the earlier illustrative $10,800 property tax, $2,400 insurance, and $12,000 maintenance reserve: $22,800 remains before vacancy, management, HOA, and financing. That is a 1.9% unlevered cash yield after the reserve and before the omitted costs. The rent is invented for sensitivity analysis, not a Kirkland rent estimate. At a chosen 4% required yield, the same rent supports approximately $773,000 if tax and maintenance remain modeled at 0.9% and 1% of purchase price: ($48,000 − $2,400) ÷ (4% + 0.9% + 1%). That is an illustration of required-return sensitivity, not a fair-value appraisal. Actual parcel taxes, comparable rents, and operating costs must replace these assumptions.

Owner-occupiers also purchase stability, control, and access to a place they value. They may rationally pay more than a rental investor. But the additional consumption value should be named rather than disguised as guaranteed appreciation. Mortgage principal repayment builds equity and is different from an economic expense; interest, upkeep, transaction costs, and the opportunity cost of the down payment still need to be compared consistently.

The transmission to Kirkland has missing links
Corporate AI gainsRevenue and productivity can rise.

Local work expands

More well-paid roles or stronger compensation can increase purchasing power.

Work becomes less labor-intensive

Profit can rise while headcount, entry roles, or seat demand weaken.

Neither route settles the housing questionWho receives the income or equity—and do they buy here?

Buyer budgets

Available homes

Actual clearing prices

Scenario mechanisms, not estimated effects or probabilities. Existing owners’ ability to wait can delay adjustment. A location premium does not establish a floor under a purchase price.

The AI boom does not guarantee the next buyer

The companion AI essay follows the economic chain from power and memory to digital work. Its implication for Kirkland is conditional: corporate AI gains support home values only insofar as income and wealth reach people who bid for homes here. Global revenue, headquarters employment, local payroll, and household purchasing power are different measures.

Two scenarios expose the vulnerability. If the trade disappoints, weaker share prices and slower hiring can hit salary expectations and down payments together. If the technology succeeds while reducing staffing needs, corporate margins can improve without reproducing the same local career ladder. In either case, high-value work can remain concentrated in the region, and new businesses can emerge. The mechanism identifies exposures; it does not establish which scenario will dominate.

As a pure illustration, a prospective buyer whose entire $240,000 down-payment fund is held in exposed shares would have $144,000 after a 40% decline, before taxes. Maintaining a 20% down-payment convention would then limit that fund to a $720,000 purchase, even if salary and mortgage rates stayed unchanged. This is a hypothetical buyer, not a claim about the financing of typical Kirkland transactions. Prior-home equity, cash reserves, and other sources can produce very different outcomes.

This is why treating a technology salary, employer shares, a technology-heavy portfolio, and a local home as four independent protections is dangerous. They can respond to the same shock. Equally, a national AI rally can enrich shareholders who never become Kirkland buyers. Geography and distribution interrupt the inference in both directions.

The burden of proof should therefore be on the price. Which comparable rent or household income supports it? How many buyers are closing at that price, and with what concessions? Does inventory grow faster than contracts? Are local hiring and compensation improving in the relevant occupations? How much resilience comes from existing owners, and how much from fresh purchasing power? An explanation for yesterday’s transaction is not a defense of tomorrow’s asking price.

The dashboard is an instrument for changing your mind

REAL-IQ becomes useful when it preserves contradictions instead of averaging them away. Population can rise while higher-income filers leave. A company can grow revenue while cutting local employment. A city can remain expensive while buyers gain negotiating room. A statewide tax story can matter politically without yet being demonstrated as the cause of a neighborhood’s price change.

That requires a visible distinction among observations, interpretations, and triggers for reconsideration. A new layoff notice belongs beside payroll and hiring data. A median price belongs beside selling time, inventory, and transaction mix. A tax claim belongs beside the law’s effective date and the migration data’s observation period. Missing data should remain missing rather than becoming a reassuring zero.

My working assessment is that Kirkland retains meaningful regional advantages while the reviewed housing indicators show weaker seller conditions and formidable entry costs. The evidence supports examining purchase prices and financing more critically. It does not support declaring Washington’s economy doomed, predicting a specific housing crash, or assuming that technology wealth makes expensive property immune.

The question behind the dashboard is therefore practical: what would have to remain true for this decision to work, and what evidence would tell me those conditions were changing?

A useful system gives that question somewhere to live before a conviction becomes an obligation.

Sources and further reading · 21
  1. Washington Commerce population summary
  2. IRS migration data and downloads
  3. IRS gross-migration CSV
  4. IRS definitions
  5. Department of Revenue income-tax guidance
  6. Department of Revenue tiered-rate notice
  7. Governor’s signing notice
  8. Young’s research and papers
  9. Inventor-mobility research
  10. Employment Security Department facts and figures
  11. King County labor profile
  12. Washington WARN requirements and database
  13. PSRC transportation report
  14. Washington output and income report
  15. CRS review of remote work
  16. Redfin Kirkland market page
  17. Data Center definitions
  18. Redfin Everett market page
  19. Freddie Mac historical mortgage survey
  20. Kirkland middle-housing materials
  21. Federal Reserve lock-in research
Back to the beginning ↑

Cite and share

Vale, Thaddeus. “What Holds Up Kirkland Housing Prices.” ThaddeusVale.com, November 18, 2026. https://thaddeusvale.com/ideas/what-holds-up-kirkland-housing-prices

Structured article (JSON)