Learning curve

Learning curve (also called experience curve in broader applications) describes the phenomenon whereby the cost or time required to produce a unit decreases as cumulative production volume increases, reflecting efficiency gains from repetition and accumulated experience (Wright T.P. 1936, p.122)^[1]. The first airplane off the line takes twice as long to build as the hundredth. The first surgery of a new type takes far longer than the thousandth. Workers, processes, and organizations learn from experience—and that learning has predictable, quantifiable patterns that dramatically affect costs over time.

Theodore Wright first quantified the effect in 1936, studying aircraft production at Curtiss-Wright. He found that every time cumulative production doubled, labor hours per aircraft fell by approximately 20%. This "80% learning curve" became famous, and variations have been observed across industries—from semiconductors to solar panels to surgical procedures.

The Wright Law formula

The basic learning curve relationship is:

Y = aX^b

Where:

• Y = cost or time per unit
• X = cumulative units produced
• a = cost or time for first unit
• b = learning rate exponent (negative)

The learning rate is expressed as the percentage retained when cumulative volume doubles. An 80% learning curve means the 2nd unit costs 80% of the 1st, the 4th costs 80% of the 2nd, and so on^[2].

Calculating the exponent

The exponent b relates to the learning percentage:

b = log(learning rate) / log(2)

For an 80% learning curve: b = log(0.80) / log(2) = -0.322

This negative exponent ensures costs decrease as cumulative volume increases.

Sources of learning

Multiple mechanisms drive learning effects:

Labor efficiency

Practice effects. Workers performing repetitive tasks become faster and more accurate. Motor skills develop. Cognitive processing becomes automatic.

Method refinement. Workers discover shortcuts and better techniques through experience. "Tricks of the trade" accumulate^[3].

Reduced errors. Experienced workers make fewer mistakes requiring rework. Scrap rates decline.

Process improvement

Tooling optimization. Tools and fixtures are redesigned based on production experience.

Layout improvements. Work flow and material handling become more efficient.

Scheduling refinement. Production planning improves with experience managing the specific product.

Product redesign

Design for manufacturability. Products are redesigned to simplify production based on manufacturing feedback.

Component standardization. Experience reveals opportunities to use standard parts.

Specification refinement. Unnecessarily tight tolerances are relaxed; necessary ones are tightened.

Organizational learning

Management systems. Quality control, inventory management, and coordination systems improve.

Supplier development. Supplier relationships mature, improving component quality and delivery.

Knowledge codification. Tacit knowledge becomes explicit and transferable^[4].

Learning curve vs. experience curve

Related but distinct concepts:

Learning curve

Narrow focus. Traditionally focuses on direct labor efficiency—the improvement in labor time per unit.

Single product. Typically analyzes a specific product or process.

Manufacturing emphasis. Most applicable to production environments.

Experience curve

Broader scope. The Boston Consulting Group extended the concept in the 1960s to encompass total costs—labor, materials, marketing, distribution, administration.

Industry level. Often applied to industry-wide cost patterns over time.

Strategic implications. BCG used experience curves to justify market share strategies—larger volume means lower costs means competitive advantage^[5].

Industry variations

Learning rates vary across industries:

Aerospace. 75-85% learning curves typical. Complex assembly benefits substantially from experience.

Electronics. 70-80% curves observed in semiconductor manufacturing.

Automotive. 80-90% curves. Process automation limits pure labor learning.

Energy. Solar panel costs have followed approximately 20% learning rate (80% curve) over decades.

Services. Learning effects exist but are harder to measure. Medical procedures show significant learning curves.

Simple manufacturing. 90-95% curves for highly standardized, automated production with limited learning opportunities^[6].

Strategic applications

Learning curves inform strategic decisions:

Pricing strategy

Penetration pricing. Price below current cost, anticipating that volume-driven learning will eventually make the price profitable.

Experience-based pricing. Set prices based on expected future costs rather than current costs.

Investment decisions

Volume projections. Forecast costs based on cumulative volume to evaluate project economics.

Capacity planning. Account for productivity improvements when planning facility requirements.

Make-vs-buy decisions

Learning ownership. Internal production captures learning benefits. Outsourcing transfers learning to suppliers.

Competitive strategy

First-mover advantage. Early entrants accumulate experience, achieving cost positions followers can't match without sustained losses.

Market share emphasis. The BCG growth-share matrix reflects experience curve thinking—market leaders benefit from lower costs^[7].

Limitations and cautions

Learning curves have constraints:

Assumption risks

Constant learning rate. Real learning may not follow smooth power curves. Rates often diminish as easy improvements are exhausted.

Volume causation. Correlation doesn't establish causation. Cost reductions may result from factors other than experience.

External factors. Technology changes, input price changes, and market conditions affect costs independently of learning.

Strategic dangers

Price war spiraling. Competitors pursuing volume through price cuts may destroy industry profitability.

Quality neglect. Focusing exclusively on cost reduction may sacrifice quality.

Flexibility loss. High-volume strategies may create inflexibility when markets change.

Implementation challenges

Measurement difficulty. Isolating learning effects from other cost drivers requires careful analysis.

Forecasting uncertainty. Extrapolating future costs from learning curves assumes continued learning—which may not occur^[8].

Organizational disruption. Personnel changes, process changes, and product changes reset learning progress.

Disruption and resets

Learning progress can be lost:

Product changes. New product versions may not carry forward all learning from predecessors.

Process changes. New manufacturing methods restart learning curves.

Personnel turnover. Knowledge leaves with departing workers.

Extended breaks. Production interruptions cause skill decay.

Organizations must actively manage knowledge retention to preserve learning gains.

Learning curve — recommended articles

Production management — Cost management — Strategic management — Operations management

References

• Wright T.P. (1936), Factors Affecting the Cost of Airplanes, Journal of the Aeronautical Sciences, Vol. 3, No. 4.
• Henderson B.D. (1984), The Logic of Business Strategy, Ballinger Publishing.
• Yelle L.E. (1979), The Learning Curve: Historical Review and Comprehensive Survey, Decision Sciences, Vol. 10, No. 2.
• Argote L. (2013), Organizational Learning: Creating, Retaining and Transferring Knowledge, 2nd Edition, Springer.

Footnotes

1. ↑ Wright T.P. (1936), Factors Affecting the Cost of Airplanes, p.122
2. ↑ Yelle L.E. (1979), The Learning Curve, pp.302-328
3. ↑ Argote L. (2013), Organizational Learning, pp.34-56
4. ↑ Henderson B.D. (1984), Logic of Business Strategy, pp.19-38
5. ↑ Henderson B.D. (1984), Logic of Business Strategy, pp.45-67
6. ↑ Yelle L.E. (1979), The Learning Curve, pp.329-345
7. ↑ Argote L. (2013), Organizational Learning, pp.112-134
8. ↑ Yelle L.E. (1979), The Learning Curve, pp.346-358

Author: Sławomir Wawak