According to this alert, shared by user Sachin Shah, LinkedIn is no longer counting 'hibernated or restricted accounts' in your follower or connection numbers, which could see the removal of a lot of profiles from your LinkedIn stats.
According to LinkedIn

"To give members and contributors a more reliable view of engagement and reach, we'll no longer include restricted and hibernated accounts in follower and connection counts on a member's profile. We'll refresh both connection and follower counts on all members' profiles going forward to remove the restricted and hibernated accounts. This is part of our efforts to build a safe, authentic and transparent experience that helps reflect a more accurate view of your audience."

Hibernated accounts, for clarification, are LinkedIn profiles that members have shut off temporarily rather than actually deleting them.

It makes sense, really; this is probably how it should've been measured in the first place. However, Twitter and Facebook count restricted profiles in their statistics, so it's nothing unique by how other apps track this component.

"If restricted or hibernated accounts become active again, they will be reinstated to the follower and connection counts they previously were in. For members that get the 30,000 connection cap, if accounts become unrestricted again, they will not be reinserted to audience lists unless they disconnect active connections."

It makes complete sense that LinkedIn would update it at this time, as the site also says goodbye to its 59 million Chinese users as the company closes shop in the region.

Now that its membership is slatted to drop by millions anyway, I suppose it's also a good time to update all of the measures in which it reports its audience - though it's not clear if LinkedIn will stop counting the hibernated or otherwise restricted accounts in its overall member numbers.
The member stats are already somewhat misleading in this regard. That is, LinkedIn has over 930 million members. This is impressive, but we must note that 'members' and 'active users', the more common social platform usage metric, are not the same thing.

Therefore, comparing it to Instagram, which has over a billion active users, the number of 930 million seems, on paper, to be in almost the same ball-park size, but the actual count for LinkedIn's active users is just around 474 million, which is much less than that which was communicated.

But considering that idle accounts were included in this number makes it murkier - so hopefully, this might just be the kick-off point of a new trend by the app to turn over this leaf and be more transparent on this front.
But probably not. It would be helpful to know just how many real, active users LinkedIn has, but it doesn't seem that it would be in the platform's interest to report this now and thereby invite unfavorable comparison to other social apps it competes for ad spend with.

But your follower and connection count will now be more accurate, which is something, I suppose.

UPDATE LinkedIn clarifies that it is not closing its operations in China at all and will retain a part of its business for Chinese customers, no local LinkedIn app, though, simply put as LinkedIn or InCareer.