A lot of translators charge lower rates than they could otherwise get, especially when they're starting out. There are a few possible reasons for this; here are a couple.

1. Lack of knowledge about the market
2. Desire to get more work
3. Desire to avoid haggling
4. Lack of confidence

Lack of knowledge about the market

Information isn't evenly distributed in the translation market. This is one of the reasons why translation agencies thrive: they spend a lot of time figuring out how much clients are willing to pay for translation, and how much translators are willing to work for.

Finding out how much clients will pay is a lot of work, and as a freelancer I'm generally happy to leave this legwork to the agencies. But without a little knowledge of what translation buyers are paying, and what other translators are charging, it's easy to get short changed. That's why I recommend finding this out. Translation conferences like IJET are a great way to do this.

Desire to get more work

Especially when you're starting out, it's tempting to set your rates low in order to get more work. As Corinne says in the Thoughts on Translation blog , this is a typical beginner's mistake. While it's fine to set your rates a bit low until you get a decent amount of work, and gradually raise them after, setting your rates very low in an attempt to get work quickly usually backfires.

As the About Translation blog points out, low rates lead to quality problems, and worse yet, give you a reputation as a low-quality translator. We have a term for this in the industry — "bottom feeder" — and there's a reason why the term isn't flattering.

Desire to avoid haggling

Translators generally aren't the most assertive people on the planet. If our idea of a good time was pressing the flesh and making the big sale, we'd probably go nuts cooped up in an office with a compiler manual for company. A lot of translators thus want to avoid haggling over rates in order to avoid conflict, and so they set their rates a bit lower than where they perceive the "market" rate to be. Believe me, I get this.

The problem is, no matter how low you go, they'll always want you to go lower. Price in a large way sets expectations of quality. Studies have shown that when people are given blind taste tests of wine, they report enjoying the wine more when told that it costs more — and brain scans even show greater activation of their pleasure centers when drinking "more expensive" wine.

This means that when you set your rates low, you're creating an expectation of poor quality. The client then figures that since you're delivering lousy quality, they might as well get it cheaply.

Take for example this tale of woe on the translatorscafe.com forums. The original poster related that despite charging a depressingly low 1.5 euro-cents per word for book translation, her client was demanding still lower rates. When you make low rates your selling point, clients are going to pressure you to lower your rates no matter how low you go.

Lack of confidence

Translators also might set their rates lower because they're not confident of being good enough. Although you might think your translations aren't the best, when you set your rates low you're telling this to your clients. I say do your best, and let the clients decide whether the quality is good enough. And as About Translation mentions in the linked post, setting low rates will actually prevent you from getting better, since you'll be swamped with sweatshop work and won't have time to improve your skills.

Conclusion

Translators set low rates for many different reasons, most of them misguided. What a lot of them fail to realize is that under-selling themselves actually sends the message to clients that their quality is poor, and creates a vicious cycle that prevents them from getting out of the low-rate ghetto.

Finally, while the cost of living varies greatly depending on where and how you live, I strongly believe in earning a living wage. I sometimes get offers from India to subcontract my translation work for 1 cent per word. I always refuse for many reasons, but two of them are because I don't believe that would be a living wage even in India, and because I'm extremely suspicious of why a good translator would offer to work so cheaply, when they can earn much more working directly for agencies or end clients.

Over on the Honyaku mailing list, Matt Stanton has a rather unique take on this topic: he believes in providing the entire spectrum himself.

Then all you do is adjust the time you spend on a translation to reflect the rate you're getting. Suppose you think you're worth 5,500 yen per hour. At 7 yen a character, you need to get through 786 chars per hour; at 5 yen, 1,100 chars. (At 4 yen, you'd have to get through 1,375 chars, which you might decide is unfeasible, in which case you would set 5 yen as your absolute minimum rate.)

The 5 yen version is not going to be very well written, there might be a few chokuyakus ["literal" translations -- Ryan] in there, some of the terminology might be a bit off, etc. – but so what? As long is there are no serious mistranslations or omissions of key material, you've provided a reasonable product at the
price. If the client wants better quality, they can either pay you more or find someone who's prepared to work for 4,500 yen an hour.

I personally would never take this approach; I can't stand sending work out the door that I know is sub-par. Sometimes with extreme rush jobs, I might not have enough time to do the translation to my own satisfaction, but in that case my quality is deadline-constrained, not rate-constrained.

It seems to me that intentionally doing shoddy work could backfire, by giving you a reputation as a poor translator. This approach seems to be working for Matt, but I think I'll keep turning down the low-paying work, rather than accepting it and doing a lousy job on it.

Matt goes into quite a bit more detail about his unorthodox work style in this thread.