Say I offer you a 50/50 chance to either win $200 or lose $150. Would you take that chance?
Most people wouldn't.
But, from a purely mathematical standpoint, it is a good bet. That chance has an overall value of $25 (.5*200 - .5*150). Many experiments have been done on such gambles and the result is that people only start to take the chance when the gain is about double the loss (win $300 or lose $150).
Now, a purely rational agent would take any chance that has a value greater than 0. So these experiments have shown that in the case of these gambles, people are not acting as rational agents. Indeed, the conclusion is that the emotional impact of a loss is about twice that of a gain of the same amount.
If you think back to your own experiences with gains and losses, you may recall feeling the same way. But the implications are very interesting.
This is what makes negotiations hard. A negotiation is, basically, one side giving up something in exchange for the other side giving up something. But if we value losses twice as much as gains (and the other side is doing the same thing), we can understand why it's so hard to come to an agreement:
We think our giving up X has a value of $1000, but the other side only sees it as a gain of $500. Making it worse, because it's a loss worth $1000 to us, we ask them to give up something that we think has a value of $1000, but they see it as a loss of $2000.
The psychology of this can be seen everywhere. In battles, the defending army fights harder than the invading army. You can even see this in gas prices. Back when there were discussions as to whether to allow cash and credit gas prices to be different, the credit card companies said that if there was a difference, it should be called a cash discount rather than a credit surcharge. The reason is that people would rather forego a discount (gain) than pay a surcharge (loss). The numbers are the same, and to a rational agent, it would be equivalent. But to people, it makes a difference.
This is a clear case in which our emotions cloud our better judgement... as usual. So what can we do about it?
Actually, it's not so hard to fix our thinking. Let's go back to the gamble above. What if, rather than giving you one chance at winning $200 or losing $150, I gave you a hundred chances? You would likely quickly reason that the overall odds would be in your favor and its very likely you would end up with a winning (greater than 0) amount. But, really, it's the same gamble - just repeated a hundred times in the latter case.
In other words, taking a broader view helped you realize that is a good bet. In the grand scheme of things, the few losses here and there would be outweighed by the wins. When faced with any decision like this, we can apply the same logic.
The negotiation situation is not terribly different. If we can take a broad view and look holistically at everything we will be gaining and losing during the negotiation, we realize that a single loss doesn't hurt so bad.
This is just one of the themes discussed by Daniel Kahneman in his excellent, if dry, tome: Thinking, Fast and Slow: