Skip to main content
diversification.com
← Back to L Definitions

Learning curve

What Is a Learning Curve?

A learning curve is a graphical representation illustrating the relationship between the cumulative production of a good or service and the average time or cost required to produce each unit. It postulates that as individuals or organizations gain experience through repetitive tasks, the time and resources needed to complete each subsequent unit decrease. This concept is a fundamental principle in production economics and is widely applied in fields ranging from manufacturing and aerospace to healthcare and project management to predict improvements in efficiency and cost reduction. The core idea is that practice leads to perfection, or at least, significant improvement in productivity.

History and Origin

The concept of the learning curve, while intuitively understood for centuries, was first formally described and quantified in the field of psychology by Hermann Ebbinghaus in 1885, as he studied the process of memory and forgetting. However, its application in an industrial and economic context gained prominence with the work of Theodore Paul Wright. In 1936, Wright, an engineer at Curtiss-Wright, observed that in aircraft manufacturing, the direct labor costs for producing an airplane decreased by a consistent percentage each time the cumulative quantity of aircraft produced doubled. This observation became known as "Wright's Law" and laid the foundation for the industrial application of the learning curve.17 His findings demonstrated that production costs could be systematically reduced through accumulated experience and process improvements, significantly impacting defense contracts and later, commercial manufacturing.

Key Takeaways

  • A learning curve illustrates that the average time or cost per unit decreases as cumulative production volume increases.
  • The effect is attributed to increased experience, leading to improved efficiency and streamlined processes.
  • It is a crucial tool for forecasting costs, setting prices, and making strategic planning decisions.
  • The rate of improvement, known as the learning rate, varies across industries and tasks.
  • While significant in many areas, the benefits of a learning curve can eventually plateau as processes become fully optimized.

Formula and Calculation

The learning curve is typically represented by a power law function. The most common formulation, derived from Wright's observations, relates the average time or cost per unit to the cumulative number of units produced.

The formula for the cumulative average time or cost per unit is:

YX=Y1⋅Xlog⁡2(LR)Y_X = Y_1 \cdot X^{\log_2(LR)}

Where:

  • ( Y_X ) = The cumulative average time or cost per unit after ( X ) units have been produced.
  • ( Y_1 ) = The time or cost required to produce the first unit.
  • ( X ) = The cumulative number of units produced.
  • ( LR ) = The learning rate, expressed as a decimal (e.g., for an 80% learning curve, ( LR = 0.80 )). The learning rate indicates the percentage of the cost (or time) remaining when output doubles. For example, an 80% learning curve means that the average cost per unit falls to 80% of its previous level each time cumulative production doubles.14, 15, 16
  • The exponent ( \log_2(LR) ) will be a negative value, reflecting the inverse relationship between experience and unit cost/time.

This formula allows for the forecasting of future production costs based on historical performance and a determined learning rate.

Interpreting the Learning Curve

Interpreting the learning curve involves understanding its shape and the implications of its learning rate. A steep initial slope indicates rapid initial improvements in efficiency as new processes are mastered and early errors are reduced. As cumulative production increases, the curve typically flattens out, indicating diminishing returns to experience. This flattening occurs because initial, easily identifiable inefficiencies are addressed quickly, while further improvements require more subtle adjustments or technological breakthroughs.

A low learning rate (e.g., 70% or 75%) signifies faster cost reduction per doubling of output, implying a steeper learning curve, which can be a significant source of competitive advantage. Conversely, a high learning rate (e.g., 90% or 95%) suggests slower cost reduction. Understanding the specific learning curve for a product or process is vital for strategic pricing, capital investment decisions, and resource allocation.

Hypothetical Example

Consider a hypothetical manufacturing company, "InnovateTech," producing a new type of circuit board. The first unit takes 100 hours of direct labor. InnovateTech has determined, based on similar processes, that they can achieve an 80% learning curve.

  • First unit (X=1): 100 hours.
  • Second unit (X=2): Average time per unit for 2 units = ( 100 \cdot 2{\log_2(0.80)} \approx 100 \cdot 2{-0.3219} \approx 100 \cdot 0.80 = 80 ) hours. This implies the second unit itself took 60 hours (2 * 80 - 100).
  • Fourth unit (X=4): Average time per unit for 4 units = ( 100 \cdot 4^{\log_2(0.80)} = 100 \cdot (2^2)^{\log_2(0.80)} = 100 \cdot 2^{2 \cdot \log_2(0.80)} = 100 \cdot (0.80)^2 = 100 \cdot 0.64 = 64 ) hours.
  • Eighth unit (X=8): Average time per unit for 8 units = ( 100 \cdot 8{\log_2(0.80)} = 100 \cdot (0.80)3 = 100 \cdot 0.512 = 51.2 ) hours.

As the cumulative production doubles (from 1 to 2, 2 to 4, 4 to 8 units), the average time per unit consistently decreases by 20%, demonstrating the power of the learning curve in driving down labor costs and improving efficiency.

Practical Applications

The learning curve has numerous practical applications in business and finance: