The video generally challenges the idea that people are motivated to work harder by money. That particular idea is pervasive and almost universal in modern corporate thought. It's no mystery why this happened. Most people are aware that economics assumes that people are self-interested. But what is less often explained is that they are interested in. In classical economics, people are self-interested utility maximisers. Utility is the economic measure for the satisfaction derived from a particular behaviour. In economics, it's a relatively short step from utility to value and from value to money. So if money and utility are the same, it's easy to see that people will behave in ways to maximise the money they make. If that's all true, then I need to pay someone for the behaviour I want and I will get it.
There are a number of problems with this view. The RSA Animate clip talks about one of these: that assumption just doesn't stack up against the empirical evidence. Money is not a generic motivator: it motivates people under specific circumstances. For example, when you don't have enough money (however you define that), then the ability to earn more will motivate you to try and earn that extra. Another case where money motivates is when choosing between jobs. If given the choice of two jobs, I'd look first to the job that paid more. If the pay gap was large, then the lower paying job would have to be significantly better in other ways for me to consider it.
Another point is hinted at by RSA Animate: they show evidence that pay for performance works with simple, mechanical tasks but doesn't work when tasks require cognitive skill. This isn't very surprising, and there is a lot of evidence for this finding. In a sense, pay for performance is a form of behaviour modification, where pay (response) is linked to behaviour (stimulus). Behaviour modification works best when the link between the stimulus and response is immediate and unambiguous. You can achieve this with simple tasks. But for more complex tasks, that link breaks down.
The final factor, which isn't in the video, is that money is what James March and his colleagues called a garbage can. This is a decision making model, where decisions tend to attract problems, which all get dumped into the garbage can. There is often no real relationship between the solutions and the problems, apart from the fact they all sit in the same organisation. For example, if you have a disenfranchised and unmotivated workforce, this could be a result of a number of different factors: company culture, poor management, a few disaffected staff, or any one of hundreds of other problems. As a result of all these problems, people complain. One of their complaints may be about the rate of pay, but this isn't always the real problem; and it is rarely the only problem. But pay is relatively simple to change - at least compared to the effort involved in improving management or changing company culture. Conversely, if you run a good, happy workplace, then you don't need to pay the highest wages to satisfy your staff.
So, money is a motivator, but not the simple motivator assumed by economic theory. The Welfare reforms proposed by Iain Duncan Smith take the relatively simple assumptions of money as a motivator and use it to reform the welfare system. The problem is, this assumes that the presence of benefits is a disincentive for people to enter work. The system is being re-organised to reduce this disincentive (and, as the FT points out, this will only be partially successful). The problem is this approach is based on an ideology: that money motivates people to work. This ideology just doesn't stack up against the empirical evidence.