David Nebenzahl wrote:

I'm here to say that the conventional wisdom that one gets here--that
"backstabbed" wiring is bad, evil, and always leads to failure--may not
be correct.

[To the perplexed, "backstabbed" means that instead of using screw
terminals to connect wires to devices such as outlets and switches, the
stripped (solid) wire is pushed into a connector that grabs the wire
inside the device. Very commonly used "in the field".]

The opinion one reads here most often is that this is an inferior wiring
method that must always be suspected when there are electrical problems,
that it should be avoided and that it should be corrected if found.

I'm not sure that's correct.

First of all, it is an approved, UL/CSA tested, and, most importantly,
code-approved (US building code) wiring method. If it was as all-fired
bad as folks here claim, why would it still be allowed? After all, the
building codes tend to err on the side of caution.

My own experience, as limited as it might be, has not shown backstabbed
connections to be the source of any trouble. I recently worked on a
house built in the 1960s in which all devices were backstabbed. I was
called to add a circuit, not to correct any problems. There was no
current problem with any device that I could see, nor was there any
history of any such problems.

I'd like to see some more evidence for the badness of backstabbed
connections. Everything I read here is either based on anecdotal
evidence, or just speculation and personal preference.

I will say that I personally don't like backstabbed connections; as
tempting as they are (a lot faster than
stripping/bending/screwing/crimping using screw connections), I prefer
the "old-school" method. But I do think they've gotten an unfairly bad
rap. Furthermore, I refrain from automatically correcting them
(replacing backstabbed connections with screwed ones) when I see them,
on the theory of "if it ain't broke, don't fix it", and I suggest this
to others. Especially newbies and DIYers; I think it's bad advice to
automatically suspect backstabbed connections as the source of a fault,
and to imply that they should all be ripped out and redone.

Let the brawling commence.

--
Found--the gene that causes belief in genetic determinism

I've seen a number of problem push-wire / back-stabbed devices
personally, and they represent a sizable percentage of the number of
problem devices I've dealt with.

I have a house that happily has all 20A circuits with 12ga copper wire
(other than the larger dedicated circuits of course), and has many 12ga
push-wire connections that are no longer allowed.

I've not had specific failures of these connections as the devices seem
to be of reasonable quality, however where I have had opportunity to
replace these devices for other reasons (adding GFCIs or changing to
Decora style devices) I have seen evidence of less than optimum
push-wire connections, i.e. discolored area at the connection point from
heating.

The devices I have used for replacement have generally been the "spec
grade" variety and have had the much better clamp type back-wire
connections. These back-wire connections allow the same convenience of
inserting the stripped wire in a hole without the need to wrap around a
screw, but instead of relying on a feeble spring connection the wire is
captured between full size metal plates that clamp around it with screw
pressure, not a spring.

These back-wire devices also work nicely in my shop where I have
stranded wire pulled through conduit. Stranded wire doesn't work well
wrapped around screw terminals and not at all with the push-wire type
devices.