Incentives can be classified according to the different ways in which they motivate agents to take a particular course of action. One common and useful taxonomy divides incentives into three broad classes:
(There is another common usage in which incentive is contrasted with coercion, as when economic moralists contrast incentive-driven work—such as entrepreneurship, employment, or volunteering motivated by remunerative, moral, or personal incentives—with coerced work—such as slavery or serfdom, where work is motivated by the threat or use of violence. In this usage, the category of "coercive incentives" is excluded. For the purposes of this article, however, "incentive" is used in the broader sense defined above.)
These categories do not, by any means, exhaust every possible form of incentive that an individual person may have. In particular, they do not encompass the many other forms of incentive—which may be roughly grouped together under the heading of personal incentives—which motivate an individual person through their tastes, desires, sense of duty, pride, personal drives to artistic creation or to achieve remarkable feats, and so on. The reason for setting these sorts of incentives to one side is not that they are less important to understanding human action—after all, social incentive structures can only exist in virtue of the effect that social arrangements have on the motives and actions of individual people. Rather, personal incentives are set apart from these other forms of incentive because the distinction above was made for the purpose of understanding and contrasting the social incentive structures established by different forms of social interaction. Personal incentives are essential to understanding why a specific person acts the way they do, but social analysis has to take into account the situation faced by any individual in a given position within a given society—which means mainly examining the practices, rules, and norms established at a social, rather than a personal, level.
It's also worth noting that these categories are not necessarily exclusive; one and the same situation may, in its different aspects, carry incentives that come under any or all of these categories. In modern American society, for example, economic prosperity and social esteem are often closely intertwined; and when the people in a culture tend to admire those who are economically successful, or to view those who are not with a certain amount of contempt (see also: classism, Protestant work ethic), the prospect of (for example) getting or losing a job carries not only the obvious remunerative incentives (in terms of the effect on the pocketbook) but also substantial moral incentives (such as honor and respect from others for those who hold down steady work, and disapproval or even humiliation for those who don't or can't).
A strong incentive is one that accomplishes the stated goal. If the goal is to maximize production, then a strong incentive will be one that encourages workers to produce goods at full capacity. A weak incentive is any incentive below this level.
Incentive structures, however, are notoriously more tricky than they might appear to people who set them up. Human beings are both finite and creative; that means that the people offering incentives are often unable to predict all of the ways that people will respond to them. Thus, imperfect knowledge and unintended consequences can often make incentives much more complex than the people offering them originally expected, and can lead either to unexpected windfalls or to disasters produced by unintentionally perverse incentives.
For example, decision-makers in for-profit firms often have to decide what incentives they will offer to employees and managers, in order to encourage them to act in ways that will lead to greater success for the firm. But many corporate policies — especially of the "extreme incentive" variant popular during the 1990s — that aimed to encourage productivity have, in some cases, led to spectacular failures as a result of unintended consequences. For example, stock options were intended to boost CEO productivity by offering a remunerative incentive (profits from soaring stock prices) for CEOs to improve company performance. But CEOs could get profits from soaring stock prices either (1) by making sound decisions and reaping the rewards of a long-term price increase, or (2) by fudging or fabricating accounting information to give the illusion of economic success, and reaping profits from the short-term price increase by selling before the truth came out and prices tanked. The perverse incentives created by the availability of option (2) have been blamed for many of the falsified earnings reports and public statements in the late 1990s and early 2000s.
Similarly, throughout the 1990s and 2000s, many corporations have sought to increase individual incentives by increasing the sizes of bonuses (to the point where they exceed salaries, sometimes by a factor as high as 10) for star performers while also laying off large proportions of their workforce, hoping to cultivate fear factor-related gains. The most extreme version of this is "forced ranking", a scheme by which workers are annually ranked and a set proportion (between 10 and 15%, usually) automatically fired. The results of these programs are mixed, but in extreme cases, usually negative.
While competition among firms has often beneficial results, lowering prices and encouraging innovation, competition within firms has almost uniformly negative results. Designed to encourage production, extreme incentive schemes actually create a cut-throat working environment where office politics dominate and actually overshadow the productive goals of the company. An example of this is the now-deceased Enron corporation. According to David Callahan's The Cheating Culture, the environment at that company was so cut-throat (as a result of extreme incentive management) that employees feared leaving their computer terminals, worried that co-workers might steal information for their own purposes.