Economists were the first social scientists to note the existence of- people who sharply discount the future.*  The early writers generally described the problem as Jevons (1871/1911) did:

To secure a maximum of benefit in life, all future events, all future pleasures or pains, should act upon us with the same force as if they were present, allowance being made for their uncertainty. The factor expressing the effect of remoteness should, in short, always be unity, so that time should have no influence. But no human mind is constituted in this perfect way: a future feeling is always less influential than a present one. (pp. 72-73)

There has not been much speculation about the nature of this imperfection. It has often been dismissed as "improvidence" due to "intellectual as well as moral causes" (Mill, 1848/1909, pp. 165-7) or a "defect of will" (Bohm-Bawerk, 1891, pp. 253-9). One author guessed that it arises from an inability to clearly imagine distant goals (Bohm­Bawerk, 1891, pp. 253-9; see also Rae, 1834/1905, pp. 52-65), which could imply either incomplete learning of the contingencies of reward or an innate limitation in the amount of this kind of learning that can occur. Samuelson (1937) spoke of it as a "perspective phenomenon," implying that it is an innate property of the way we perceive time. Strotz (1956), the only modern author to analyze the problem in detail, noted that people often change their preferences as time passes, even though they have found out nothing new about their situation. To account for this he hypothesized that a goal's "utility," or rewarding effect, decreases with delay according to an inborn function. The curve he drew to portray this has a pronounced upward concavity, suggesting that utility falls off rapidly for relatively short delays and declines more gradually as delays get longer.

* That is, economists were the first to see impulsiveness as a discounting phenomenon. [added June 2018]

Freud (1911/1956) said that impulses were a product of the operation of the "pleasure principle" and were controlled insofar as a person learned the "reality principle." Of the competition between the two principles, he said:

The superiority of the reality-ego over the pleasure-ego has been aptly expressed by Bernard Shaw in these words: "To be able to choose the line of greatest advantage instead of yielding in the direction of least resistance." (Man and Superman).

And in the same paragraph:

Actually the substitution of the reality principle for the pleasure principle implies no deposing of the pleasure principle, but only a safe-guarding of it. A momentary pleasure, uncertain in its results, is given up, but only in order to gain along the new path an assured pleasure at a later time (1911/1956, p. 223)

He was clearly talking about the same issue as the economists and sociologists just dis­cussed, although he broadened it to include purely intrapsychic behaviors such as autistic thinking. He seems to have had a mechanism in mind, but he formulated it vaguely. When he said the "momentary pleasure" is "uncer­tain in its results" he must have meant that there is a risk of later—larger punishment or loss of reward, since the pleasure itself is at hand and presumably the least uncertain of any of the factors. In a later paper he im­plied a simultaneous competition between "id-pleasure" and an "ego-pleasure" which has been somehow differentiated from it through learning (1920/1956). He seems to have wondered why behaviors based on id-pleasure persisted in competition with much greater ego-pleasure. In this regard, he postu­lated a "repetition compulsion" by which be­haviors could become immune to extinction (1914/1956) and later a "death instinct," which might make punishment itself reward­ing (1920/1956). He never defined how these principles interacted with the pleasure principle.

Reviewing analytic theory in 1950, Rapaport remarked that little was known about the origin of impulse controls, but he did say:

The core of this change from the primary to the secondary process appears to be change in the character of delay. The delay to begin with was due to external circumstances (i.e., the unavailability of immediate reward) and is turned into an ability to delay, into an internal control. (p. 164)

Hartmann (1956) listed "postponement of gratification and a temporary toleration of unpleasure" as a function of the reality principle, possibly developing because "the pleasure principle is a less reliable guide to self-preservation" (pp. 35-36). He implied that processes which obtain long-term benefit gradually supplant those which obtain only short-term benefit, but he suggested no specific mechanism. Singer (1955) suggested that people learn to control impulses by converting motor impulses to thoughts. Perhaps this is the process of conceptualizing distant goals, which has been studied by the social psychologists (see above). Klein (1954) spoke of "delay mechanisms" exerting a "force" to regulate a person's response to needs. He said that these may act by developing "cognitive attitudes" which can "detour the approaches to need-satisfying objects . . . alter the consummatory process . . . modify the intensity of the need itself, or . . . draft the energy of the need to some other adaptive intention prominent in the situation" (p. 227). None of these authors specified rules relating the force of impulses to the influence of higher processes.