GE vs LG microwaves: an incident-report comparison

What this GE vs LG comparison is based on#

A comparison page is supposed to find a difference. This one does not, and the absence is the honest result rather than a gap in the work.

The two records are 418 CPSC SaferProducts incident reports naming a component on a GE microwave and 70 doing the same for an LG microwave, both filed between 2011 and 2026. Set the columns side by side and the largest gap anywhere is 2 points, on the control board. The next two rows differ by a single point each, and the table marks both as one-report noise.

What the tables actually contain#

There is a second thing worth noticing before reading any row, and it is easy to miss: in both records the named-component rows account for a small minority of the filings. The control board leads GE's column at 8%, wiring or the power cord and the motor follow at 2% each, and the remaining rows round to nothing. LG's column is shorter still — the board at 6%, then two single filings.

That means most reports in a microwave record do not resolve to any of the components this table tracks. What those reports describe instead is not represented here, so the table should be read as a view of one slice of each record rather than as a summary of it. It is a real limit and it applies equally to both columns, which is part of why they look so similar.

Reading a page with no separation#

Within that slice, the two brands are indistinguishable. The board is the most-named component for both, at shares that differ by 2 points on corpora of very different depth. Wiring or the power cord is 2% for GE and a single filing for LG. The motor is the same story in reverse order.

None of that is surprising once the corpus sizes are in view. LG's 70 filings put roughly a point and a half behind every single report, so a 1-point difference against GE's 418-report column is well inside what one filing would move. The correct reading is that these records do not distinguish the two brands, not that they establish equivalence — a null result from a thin corpus is weaker than a null result from a thick one.

What is still usable#

Two things. First, for either brand, a microwave report that does name a component names the control board more often than anything else, by a clear margin within each column. As a first suspect on an electronics complaint that is a reasonable prior for both.

Second, the shape of both records argues against treating this table as a picture of microwave reliability at all. These are filings people chose to make, weighted toward safety incidents, with no count of machines behind them — and in the microwave case, most of them do not name a component in this taxonomy. That is three separate reasons why no percentage here converts into how often anything fails.

GE vs LG microwaves: what the reports mention#

Both columns are shares of their own corpus, and the corpora differ in size. GE: CPSC SaferProducts incident reports for GE microwaves, 2011-2026, n=418. LG: CPSC SaferProducts incident reports for LG microwaves, 2011-2026, n=70. A difference no larger than one report on either side is marked "within one-report noise".

About the corpus: these are CPSC SaferProducts reports — incidents someone chose to report, not units sold — so the shares on this page are shares of the reports and never a failure rate. The corpus is weighted toward safety events: 12,007 of its 20,581 records (58%) are fire, smoke, burn or melt reports, which over-represents the components that burn and under-represents the ones that merely stop working.

GE microwaves: the reported component split#

GE microwaves — every share in this table is a share of: CPSC SaferProducts incident reports for GE microwaves, 2011-2026, n=418.

Shares of reports, not failure rates — see the corpus note above.

LG microwaves: the reported component split#

LG microwaves — every share in this table is a share of: CPSC SaferProducts incident reports for LG microwaves, 2011-2026, n=70.

Shares of reports, not failure rates — see the corpus note above.

How to use these numbers#

What this comparison can settle: which components the reported incidents of each brand actually concentrate on, for this appliance. What it cannot settle: which brand fails more often. That would need units sold, and no public source states it. Use the component split to plan for the repair each brand is most reported for, not to rank the two.