Great, great question.
I just designed a kind of “quick start” experience that would sign users up for a series of options in one click. It looked good, tested well… But it would also take advantage of something I call “usability for evil.” It helped the company more than it did the customer (some of the options the user agrees to reduced company costs and/or signed the customer up for additional fees). I’m all for showing the ROI of my designs… But just as I don’t want to be manipulated, I don’t want my work to manipulate others.
I killed the design.
(Another example of “usability for evil” I often cite is when Facebook changed thier privacy settings over a year ago, they set all the public settings to be the default selections – which helped them a lot more than their users. Facebook’s designers know that most users accept the defaults, and they were exploiting that fact. I’m no longer on Facebook.)
So, going back to the question, yes, design can absolutely be manipulative. You see examples of that all over the place if you look for it hard enough. But design can also be supportive, assisting users when they want to do something. The best design, in my opinion, is just there, waiting and available and neutral. It provides value that is obvious and helps users. And by helping users, it makes those users happy and content – sometimes so happy that they become advocates for your design to friends and family.
See question on Quora