“But rates haven’t gone up in ten or fifteen years!” is a common complaint from freelance translators, and it’s true, rates at the bulk end of the market have certainly not risen over the last ten to fifteen years, and have dropped quite considerably in real terms. Explanations for this phenomenon vary, but usually include some combination of rapacious agencies, miserly end-clients, competition from India and/or China and/or machine translation, young translators who don’t know their worth, translators without business sense or who don’t how to market themselves and a host of other excuses. Because all of these are just excuses, and it’s quite easy to see through them. If it was really so easy to make money running a translation agency, we’d all be doing it! It isn’t, you have to be able to sell your translation services at about twice the price you’re paying your translators, who are also your main competitors. Of course an agency wants to sell its services at as high a price as possible, but there are economic limits to the price it can charge… including miserly clients! Well, no, of course the end-client wants to get their translation services at as cheap a price as possible, but surely the clients we have now are the same as the clients we had ten or fifteen years ago, so how can that be the cause of a drop in price in real terms? The translators who don’t know their worth or don’t know how to market themselves, they were around ten or fifteen years ago as well, so they can’t be the reason either.
So what about the competition from India and China, where translators are charging less than one U.S. cent per word, or from machine translation, which is virtually free? But have you seen what you get for that? Google Translate might be useful for getting the gist of a website in a foreign language, but it is simply not fit for most commercial translation purposes. And the one-cent translations from India and China are basically just the output from Google Translate or similar programs. The machine translation programs used by the large agencies all around the world are not significantly better than Google Translate in this respect, whatever their slick marketing might pretend.
In my opinion, the real reason rates have been dropping in real terms for the last ten to fifteen years is much more interesting than these moans over conference drinks, and much more optimistic for the individual freelancer. As well as all the above phenomena, the last ten to fifteen years have shown an explosion in the availability and use of translation memory (TM) software, of which SDL Trados is the best known and putative market leader. I should quickly point out for the non-specialists that TM software is not machine translation: instead, it is a way to be able to “remember” all the translations a translator has ever done, so that when a sentence comes up a second time, there is no need to “think” about the translation. Even better, it makes it possible to combine the “memories” of several translators, which is invaluable if you are trying to maintain consistency over time and between different translators, as large users of translation such as the automotive industry would very much like to do. Like machine translation, TM software has been around for a long time, but it only really came into its own when the necessary computing power – quite a lot, as many a translator’s laptop battery will attest –became affordable enough for the software to be used by freelancers and not just by the in-house translation departments of large companies. Microsoft started using Trados for its enormous translation needs in 1997, and that can be seen as a convenient start date for the current “TM era” in translation.
TM software doesn’t just make it far easier to maintain consistency in translation. It can also, in certain situations, vastly speed up the translation process. Suddenly translators weren’t having to type so much when translating repetitive documents. It’s difficult to estimate the gain in efficiency across translation as a whole, because there are so many vastly differing situations, but some lucky translators were able to translate several times as many words per day: and translators are paid by the word!
Most translators are individual freelancers: we don’t have a fixed salary. We depend on the profit of our one-person business for our income. We are also professionals, part of a professional environment, and human beings with certain aspirations as to our standard of living. We are overwhelmingly graduates, usually with postgraduate qualifications, often with significant professional experience in other fields, and that colours our aspirations. Some of us want to be rich, and most of us want at least a comfortable middle-class lifestyle. I earn about the same as an experienced classroom teacher, or a specialist nurse, or a well-established lawyer, or a junior doctor; these are jobs with similar levels of qualification and experience to my own, I don’t feel “cheated” by my earnings, even if I wouldn’t say no to a bit more! So what happened when TM software came along and multiplied the incomes of some translators several times over? Well most of translation is not a regulated profession, and the economic barriers to entry are very low compared with just about any other job. So someone came along, saw that there were now excellent incomes to be made in some areas of translation, and realised that he or she could have that business by charging slightly less and being satisfied with a very good income instead of an excellent one. The real per-word price of translation began to drop with the new gains in efficiency, just as classical economics would predict.
But the story doesn’t end there. Translators do not work in isolation, even if most of us work on our own. We are part of the overall economic process, we add value. But most of our clients see translation as a cost that has to be integrated into the price of their goods or services. What did our clients do as translation costs started to fall? Some of them took the welcome extra profit; but a lot of them responded by ordering more translation for the same or even higher budget. Because, with the dropping cost, it became economic to translate certain documents that before it would not have been profitable to translate, because a translation of that particular document didn’t add as much value to the client’s business as the translations of the other documents. And therein lies the explanation to a paradox: the price of translation may have been falling in real terms for the last ten to fifteen years, but the “translation industry”, the sum total of the value of the translation services provided, has been growing over the same period, usually at double-digit annual rates. Because there’s now a lot more translation going on, and that’s not counting the free machine translation provided by Google Translate and Bing and their ilk, that’s only counting the translation that’s paid for.
So what are the morals of this story? Well the drop in rates in real terms at the bulk end of the translation market, the segment of the market that requires competent human translators but where there is effective competition between translators, is due to a real gain in efficiency due to increasing use of TM software. It’s not due to increased competition. Competent human translators will not willingly work for a rate that cannot provide them with the minimum degree of comfort to which they aspire, and this provides some sort of floor to rates all the time that competent human translators are rare compared to the demand for their services. The fact that the demand for competent human translators increases as the price falls is, in my mind an effective guarantee that competent human translators will remain rare in comparison to the demand for their services for the foreseeable future. The sky is not about to fall on our heads!