Every California tax sale shrinks between the county approving it and holding it. The board adopts a resolution, the list is published, and then owners redeem right up to the last business day. Kern’s September list lost 203 of 1,238 parcels — one in six — in three months.
The obvious question for anyone reading a list early is whether that attrition is selective. If the good parcels get rescued and the junk is left behind, an early list is worth less than it looks.
We can now answer it, because we read Kern’s sale in May from the board’s approving resolution and the county has since published its own list. The same 1,238 parcels, three months apart.
On price, redemption looks random
Compare the 203 parcels that left against the 1,035 that stayed:
| Redeemed | Still on offer | |
|---|---|---|
| Median opening bid | $4,300 | $4,300 |
| Median assessed value | $2,899 | $2,899 |
| Median bid ÷ assessed | 1.25× | 1.32× |
That is about as close to no signal as real data gets. Whatever decides who pays off a tax debt, it is not the size of the debt or the county’s valuation of the land.
This is worth stating plainly because the opposite is widely assumed. An early list is not systematically stripped of its best-priced parcels. On the ratio we screen by, what survives looks like what left.
On one thing it is not random at all
Whether there is a building on it.
44 of the 203 redeemed parcels carried improvement value — 21.7%. Among the 1,035 still on offer, 78 do — 7.5%. A parcel with something built on it was about three times as likely to be rescued (odds ratio 3.4; χ² = 38.2, p < 0.001).
The obvious objection is that this is really about value — buildings are worth more, and owners of valuable things fight harder. So we split both groups by assessed value and looked again:
| County assessed value | Redeemed, with a structure | Still on offer, with a structure |
|---|---|---|
| Under $5,000 | 0 of 135 | 0 of 676 |
| $5,000 – $25,000 | 16 of 38 (42%) | 13 of 191 (7%) |
| Over $25,000 | 28 of 30 (93%) | 65 of 168 (39%) |
The gap does not close inside a value band. It widens. Among Kern parcels the county assesses above $25,000, 93% of the ones that redeemed had a building, against 39% of the ones that did not. It is not a proxy for value. It is the building.
It holds in a second county
One county is an anecdote. Los Angeles gives an independent test, because its Assessor publishes a live tax status against a list adopted in May: 62 of the 1,252 parcels in the October sale had already cleared the default that put them there when we read the records in August.
50 of those 62 — 81% — carry improvement value. Among the 1,190 still in default, 198 do — 17%. Odds ratio 20.9. Different county, different data source, different definition of “gone”, same direction and a much larger effect.
Two counties is still two counties, and the LA figure is a status flag rather than removal from a published list, so the two are not measuring exactly the same event. But the direction is not in doubt.
What it means if you are reading a list early
The interesting items are the fragile ones. On a desert list, a parcel with a structure is the row people stop at — it is the one that looks like a house rather than a square of scrub. It is also, on these numbers, three to twenty times more likely to be gone before the gavel.
So the practical rule is about where you spend your effort. Title work, a site visit, a call to county planning — that is real time, and spending it on the single most redeemable category is how it gets wasted. If a structure parcel is your plan, have a second plan, and re-check it against the county the week of the sale rather than the month before.
The reverse is the quiet good news: bare land is what actually reaches the auction. An early list of vacant parcels is a much better forecast of the real sale than an early list of improved ones. That is most of what these lists are — 92.5% of what is left in Kern is bare ground — and it is the part you can plan around.
What this does not say
It does not say redeemed parcels were better buys. On price they were the same, and we have no idea what any of them were worth beyond what the county assesses them at.
It does not explain why. A building usually means a lender, an insurer, an occupant or a family, and any of those can produce the money a bare lot never does — but we are measuring an association across two counties, not a motive.
And it is measured on two sales in one state, at one point in the redemption window. If you have a county list from before and after its own sale, we would genuinely like to see it — reply to the newsletter and tell us.
Sources
- Kern County Board of Supervisors Resolution No. 2026-186, adopted May 12, 2026, Exhibit A — 1,238 parcels — against the county’s own list on GovEase (auction 1348), read August 15, 2026: 1,035 parcels.
- Kern County Assessor parcel service, for land value, improvement value and acreage on all 1,238.
- Los Angeles County Board of Supervisors letter 26-2339, adopted May 5, 2026 — 1,252 parcels — against the Los Angeles County Assessor’s public parcel record read in August 2026, for tax status and improvement value.