At the same time - are the certification costs that bad? I recently certified an unapproved oil cooler with no paperwork on a CofA vintage aeroplane. It took a couple of days of my time, and a £62 CAA fee. The big issue isn't the cost - it's that virtually nobody understands the mod approval system. I'm lucky in having a professional education that means I can do this, which is the real issue: very few people do.
That is a very valid comment.
FAA approvals are
free - subsidised by the US taxpayer. This is equally true for both a Major Alteration (337 + Approved Data) and for an STC application.
EASA major mod approvals are not expensive either. I have never sumbitted one but recall reading about one instance where somebody did one as a DIY job and it cost him of the order of £200.
The real costs come from one or both of two areas
1 - the cost of generating the data to support the application
2 - the cost of in-house staff generating the paperwork and liasing with the certification authority
Normally, with most aircraft mods, 1. is trivial, but it would not be trivial with say an engine which needs to be demonstrated as capable of restarting at 18,000ft which means screwing it to some plane and actually going up there and doing the tests. With a new airframe this is an even bigger cost because of the flight testing required.
Item 2. is the main cost with most "small" mods e.g. avionics. The agency approval fees are usually insignificant or nonexistent.
But in all cases IF you have the personnel in house that is familiar with the process (and all the big names will have that in copious quantities) none of the stuff is hard. It's a process you go through, ticking all the right boxes.
Small outfits will farm out stuff to outside people (e.g. a DER in the FAA system, to generate Approved Data) and these can be very expensive, e.g. charging $10k for a paperwork pack for the installation of a single instrument. But Lycoming won't be doing that.
but it's lack of professional skills in mod application writing that is the real issue:
It's not a professional education you need; it is being able to read and write
A rare skill these days, among the smartphone generation which cannot write more than a line or two of pidgin english...
I think the wider issue in the USA is that if you have a company like Lyco which has not introduced a new product in many years, there won't be anybody left in there with any intelligence. Smart people move on to more interesting jobs. I see this with Honeywell (GA avionics) where nothing has happened in over a decade, and there is nobody there who knows anything. Currently they are fumbling and bumbling trying to fix the major design defects in the KFC225 autopilot of which they sold a few thousand 1999-2002 and which are packing up all over the place.
You would think Lyco could screw an RSA5 fuel servo onto the bottom of an O-200 engine and get that certified, and do away with carb ice. But it would increase the cost of the engine by maybe $1000, which will be a major issue with Cessna. And pilot forums would fill up with questions on how to do a warm start on an IO-200
There is another big factor here: the market is declining and blaming certification costs or product liability for a rationalisation of the business is an easy way to do it without losing face. For example Cessna's recent dropping of piston production, citing product liability, was a dead handy excuse for dropping stuff which nobody was buying anyway.