Kinked Experience Curve

Kinked Experience Curve

Yu Sang Chang (Gachon University, Korea & KDI School of Public Policy and Management, Korea) and Jinsoo Lee (KDI School of Public Policy and Management, Korea)
Copyright: © 2014 |Pages: 9
DOI: 10.4018/978-1-4666-5202-6.ch123

Chapter Preview



Initially discovered in 1936 (Wright, 1936), the functional relationship between performance measures such as cost, price, or efficiency index and accumulated experience has been shown to exist in nearly every fields of activities. For example, it has been recently reported (Weiss, Junginger, Patel, & Blok, 2010A) that the existence of EC relationship was verified by 207 energy-related studies as well as by 124 studies from manufacturing industries with an average learning rate of about 18%. Under EC, that indicates that performance measures such as price, cost, or index will improve 18% upon each doubling of cumulative experience. In other words, the simple concept of “practice makes it perfect” or “learning by doing” has been transformed into a mathematical relation between performance and accumulated experience. More specifically, it is defined as double log linear relationship between performance and cumulative experience. Thus, a fixed 100 percentage change of cumulative experience is correlated to another constant percentage change like 18 percentages in performance.

Theories about why such relationship exists are still not fully developed. However, we can begin with a simple black box of static learning system which converts input into output, as shown in Figure 1. The conversion rate of input into output is fixed in this system. Therefore performance measure which shows the ratio of input over output will also remain constant.

Figure 1.

A simple static learning system

If learning system were to follow the concept of EC, then we will have a dynamic learning system where the same input will never produce the same output. Following cybernetic theory (Ashby, 1964; IEA, 2000; Wene, 2011), this dynamic learning system will generate continuously changing conversion ratio between input and output as a function of cumulative experience. Therefore, the resulting performance measure will also change (improve) continuously.

As shown in Figure 2, dynamic learning system has two types of information feedback loops dealing with performance measures and deployment (or volume) rate. Due to performance information feedback, dynamic learning system will continue to improve its inner working so that performance measures can improve according to its learning rate. Similarly, deployment (or volume) information feedback will continue to adjust its deployment rate to match the requirement from environment like market or public demand.

Figure 2.

Dynamic learning system under experience curve


Main Focus

One of the basic assumptions of dynamic learning system under EC is that learning rate will remain constant throughout the period. However, as early as 2000, a theory of kinked EC has been proposed (IEA, 2000) as shown in Figure 3. The initial experience slope or PR can be bent down to generate more steep kinked slope when a radical technology is introduced in learning system.

Figure 3.

Technology structural change

Key Terms in this Chapter

Accumulated Experience: A measure which adds the amount of all the past experiences through the present. An accumulated experience is often represented by cumulative production volume, installed capacity, physical activities or social activities.

Kinked Year: The year when a kink in learning rate is observed in a kinked experience curve.

Kinked Experience Curve: Instead of a single learning rate, there may be multiple learning rates over the life cycle of experience curve. The introduction of radical technology or governmental intervention is likely to cause kinks in learning rate. In most cases, there will be only one kinked learning rate.

Performance Measure: A ratio of output to input which indicates the level of performance of a dynamic learning system. Performance measure can be defined in term of price, cost, physical, or social index.

Learning Rate: As the accumulated experience increase by 100 percent, learning rate indicates an extent of fixed percentage rate of performance improvement.

Large Dynamic Learning Systems: The scope of systems is very large covering nation, region, or the world in which input to output conversion rate undergoes continuous improvement.

Complete Chapter List

Search this Book: