How is this different from “Alternate page with proper canonical tag”?

There, a canonical exists and Google agrees with it. Here there's no canonical, so Google decided on its own which of the similar pages is the main one.

Do duplicates lower rankings for the whole site?

Google doesn't penalise duplicates, but it splits signals between several URLs and may show a version you didn't intend. That's why it's worth choosing yourself.

Which URLs does Google treat as duplicates?

Pages with very similar content: with and without a trailing slash, with and without www, http and https, parameter variants, near-identical products, empty categories and tag pages.

Duplicate without user-selected canonical

Google decided the page is a copy of another one and picked which of them to index on its own, because the page has no canonical tag. The fix: choose the canonical yourself — this page if it should rank, or the main version if it's a duplicate.

What this status means

Google found several pages with very similar content. None has a canonical tag, so Google chose one as the main version and marked the rest as duplicates left out of the index. This page was marked as a duplicate. To see which page Google chose, use URL Inspection → "Google-selected canonical".

Common causes

the same URL available in several versions: with and without a trailing slash, with and without www, over http and https,
URL parameters: filters, sorting, session IDs, utm_*,
products that differ only by colour or size with identical descriptions,
empty or near-empty categories, tag pages, internal search results.

How to fix it

Technical versions of the same URL (www, https, slash) — set up a 301 redirect to one version.
Parameters and variants — add a canonical tag pointing to the main version.
The page should rank on its own — add a self-referencing canonical and make its content distinct from similar pages.
Low-value pages (empty tags, search results) — consider noindex.

After the changes, request indexing for the most important URLs.

How to keep an eye on it

Duplicates often appear after a template change that stops outputting the canonical tag.

IndexRise records the canonical of every monitored page and compares it nightly with the baseline. When the canonical disappears or starts pointing elsewhere, you'll see it the next day — along with the value before the change.