Here's a central question, is proofreading any faster or easier than translating from scratch? Recently I've done a couple of proofing jobs and had some discussions about the subject with some clients. It seems there are a few important points to keep in mind. I'm refering only to Chinese to English for these points:
1. Economic disparity.
This lies behind a lot of the reasoning of some translation agencies. Basically for my language combination, Chinese to English, it's much cheaper to get a Chinese native than to get an English native. As a guide a Chinese native might be around 20GBP per day (although they sneakily charge US and UK based companies around double that!), an English native is going to want something like 100-200GBP per day. So the question is, is it worth the extra 80GBP to get a native English speaker. The answer is simple, why not just hire Chinese people to do all your English writing then? Why not have Chinese based writers doing the Cambridge English dictionary? If being a native speaker makes your English better (it does) then if you really want a decent document, you have to get an English native to do it.
This leads to the question of taking the "middle road" and getting a Chinese person to do the original translation, and then hiring a native speaker for "proofreading" to make it sound more English. I would say I get about 3-5 emails a week from clients who run that kind of service. I always refuse to do these jobs because I find it takes me just as long to proofread the translation (which is ususally unspeakable quality) as it would to do it from scratch. In other words, these companies are not producing a "virtually" native translation that just needs a bit of work. They are producing a terrible translation that needs "re-doing" from scratch.
Needless to say, if there was not such an economic disparity for my language pair. I think most of these clients and companies would disapear.
2. Quality of translation
I'm sure it's possible to do a great translation into a foreign language. In fact there is plenty of research showing that it is possible. Here are my questions. Can you translate as quickly into a foreign language? Can you read through the tonnes and tonnes of required information as quickly in the foreign language? Can you keep doing that, all day, everyday for years and years? I would say the answer is no. Even though my Chiense is extremely good. I still prefer reading in English, and I read much faster in English. When I get a job on a difficult subject, I'll always read up on that subject in English first to get an idea of what's happening. If I did the same in Chinese, it would add a few extra minutes or hours here and there, and over weeks and years, I'd be far less profitable.
Thus, I think the idea of "quality" needs to be viewed not in the context of "doing it with unlimited time and resources" but in doing it at a cost of time and effort which is sustainable and reasonable to the translator. So if I was a Chinese native, I would much prefer to work English into Chinese, so I can have all the advantages of using my native language. In other words, I'm claiming that every translator would prefer to work into their own native language if possible.
Combining my two points. I claim that proofreaders don't really have much to gain from proofing non-native translation. I also claim that every translator would prefer to work into their mother tongue, so for both sides, the best thing would be for translators to get into their own mother tongue and for proofreaders to work in the same direction.
3. Agency savings?
I've worked with so many agencies who have come to me with jobs that were translated terribly and are now well overdue and they need me to go back over the document from scratch. This costs them a great deal of extra money, not to mention stress. The attempt to save a bit of money usually ends up costing a lot more than it would have saved in the first place.
The key question is, how good does the target need to be? If the client doens't need a brilliant translation, then of course it's better for everyone for the cheaper direction to work backwards. If the client needs a good quality translation, then that's when it's worth the extra cost. I would aruge that, a lot of Chinese clients don't realize how bad their Chinese to English translations are. They listen to their agencies (we get native's to proofread!) and think the English must be good. The clients don't have the resources to check the translation so they assume it's native English. The results are hilarious!
So personally, I hate proofreading anything done by a non-native translator, but not because I don't think a non-native speaker can't do a good job, more because I think over years and years of work, the non-native speakers will not be keeping such a consistently high quality level. I think clients need to better understand the needs of the end companies, and need to be able to provide what they actually ask for, not what they think the client needs.