Centre for Strategy and Performance

The Learning Curve for Industry

The learning curve concept for industry states that the input cost, or time, per unit produced decreases by a set percentage every time the cumulative production output doubles. The roots of the learning curve concept go back more than a century to studies of how an individual’s performance at a task improves with experience (e.g., Thorndike, 1898; Thurstone, 1919). Wright (1936) introduced the concept to an industrial environment by showing that the decrease in direct labour cost fell by 20% every time the cumulative production doubled for airframe manufacture. Since Wright’s study, a similar effect has been shown to exist in the case of a small group, an organisation and an industry (see Argote et al., 1990, for references). In the literature the phenomenon is variously referred to as "learning curve", "experience curve", "learning by doing" or "learning by use".

The most common form of the relationship between input per product is a log-linear model in the form of the function:

(Equation 1)

where

y = input cost for the x^th unit

x = cumulative number of units produced

a = input cost for the first unit

b = progress rate

An Example

Say you have a production process with an input cost of £1,000 for the first unit and a 20% reduction in per unit cost for every doubling in the cumulative production output. The values for the parameters are then:

a = 1,000

b = -log(1 - 0.2) / log(2) = 0.322

The resulting curve (a straight line) for the unit cost plotted against cumulative production using Equation 1 is shown on a log-linear graph in Figure 1.

Although a 20% reduction in costs per doubling in cumulative production has been used as a rule of thumb in many industries, it should be cautioned that this can differ even for similar industries, within companies and for subsequent runs of the same product in the same plant.

Literature Reviews

1. Dutton, J.M. and Thomas, A. (1984) "Treating Progress Functions as a Managerial Opportunity", Academy of Management Review Vol. 9 No. 2, pp. 235-247.
2. Yelle, L.E. (1979) "The Learning Curve: Historical Review and Comprehensive Survey", Decision Sciences Vol. 10, pp. 302 – 328.

References

1. Thorndike, E.L. (1898) "Animal Intelligence: An Experimental Study of the Associative Processes in Animals", The Psychological Review: Monograph Supplements Vol. 2, pp. 1-109.
2. Thurstone, L.L. (1919) "The Learning Curve Equation", Psychological Monographs Vol. 26, pp. 1-51.
3. Wright, T. (1936) "Factors Affecting the Cost of Airplanes", Journal of Aeronautical Science Vol. 4 No. 4, pp. 122-128.
4. Argote, L., Beckman, S.L. and Epple, D. (1990) "The Persistence and Transfer of Learning in Industrial Settings", Management Science Vol. 36 No. 2, pp. 140-154.

Interesting website

• Learning Curve Calculator:
  http://www.jsc.nasa.gov/bu2/learn.html

 by Erik van der Merwe, < ev211eng.cam.ac.uk >