According to Kant, morality is not based upon "hypothetical imperatives" which proposes that "If you want X (happiness for instance) then do Y." This approach to morality is Consequentialist and, as previously noted, Kant rejects Consequentialist ethics.
Kant chooses to base morality upon what he terms a "categorical imperative" which proposes simply "You must do X regardless of the outcome or consequences." The risk inherent to Consequentialist ethics and hypothetical moral imperatives lies in the reality that individuals vary in their understanding of the consequences or outcomes which should be desired, and therefore their means for attaining those goals would differ as well.
According to Kant, morality should not vary from individual to individual. Morality should not be rooted in what we "want" to do (human volition), but should be based upon the use of reason and rationality.
The designation "categorical imperative" is significant in that the word "categorical" entails Kant's conviction that morality and the basis for morality should be inclusive of all individuals regardless of their own particular context for moral decision-making. The term "imperative" reveals Kant's conviction that there exists an actual moral criterion upon which each individual "must" based moral decisions. In other words, Kant was searching for a foundation for morality, which is required of each individual regardless of his or her own life situation.
Kant's Categorical Imperative emerged in two forms. The first essentially stated that when confronted by a moral dilemma the individual should act upon that maxim or rule which they would prefer that each individual act upon. The second essentially stated that in a moral dilemma with other human beings we should always treat others as an end in themselves rather than using them as means to an end.
Kant proposed that a standard for morality should possess two criteria: universal consistency and universal acceptability. Regarding the criteria of consistency, Kant argued for instance that a maxim such as "lie when you can hey away with it" cannot be universalized consistently because if people lied when they thought they could get away with it you would never know when anyone was telling the truth. However, there are maxims, which can be universalized without contradiction, but these maxims are not universally acceptable. For instance, imagine a selfish world where no one helps out others except for personal gain; however, such a world would not be acceptable to everyone and the fact that it does not have universal acceptability makes it a maxim on which no moral action can be based.
