Motivation: Why Classic Rewards Fail and Intrinsic Drivers Matter

Topic: Human motivation, specifically how rewards & punishment affect performance.
Presenter: Researcher with several years of experience reviewing behavioral-science literature.
Key Claim: Classic incentive logic ("reward more → get more; punish → get less") fails in many real-world, cognitive contexts.

Widespread corporate practice: tiered monetary bonuses (top performers = large cash prize).
Assumption rooted in classical economics & behaviorism: higher extrinsic rewards ⇒ higher performance.
The "science is freaky": empirical results repeatedly contradict this when tasks require cognition.

Participants: MIT students.
Task suite:
- Memorize digit strings.
- Solve word puzzles.
- Spatial puzzles.
- Simple physical skill (throw ball through hoop).
Incentive structure (3 tiers):
1. Small reward ("pretty well").
2. Medium reward ("medium well").
3. Large reward ("top performers").
Findings:
- For tasks requiring only mechanical skill 👉 $\text{Higher pay} \rightarrow \text{Better performance}$ .
- For tasks requiring even rudimentary cognitive skill 👉 $\text{Higher pay} \rightarrow \text{Worse performance}$ .
Significance: Contradicts textbook economics. Funded by the conservative $\text{Federal Reserve Bank}$ (i.e.
not a “socialist plot”).

Rationale: Test if larger relative stakes change result (≈ $50\text{–}60$ USD = sizable local income).
Incentive tiers converted to salary multiples:
- Small = $2$ weeks’ salary.
- Medium = $4$ weeks (≈1 month).
- Large = $8$ weeks (≈2 months).
Findings replicated & amplified:
- Medium reward ≈ Small reward performance.
- Large reward produced worst performance of all.

Replicated by psychologists, sociologists, and economists "over and over".
Pattern:
- Algorithmic / rule-based tasks → $\text{If–then bonuses}$ excel.
- Conceptual / creative tasks → $\text{If–then bonuses}$ hinder.

Paradox: $\text{Money is necessary but not sufficient}$ .
Two-step rule:
1. Pay enough to remove $\text{money anxiety}$ .
2. Motivation then pivots to internal drivers.

Autonomy – "Our desire to be self-directed."
- Traditional management → focuses on compliance.
- Modern work → demands engagement, therefore self-direction works better.
Mastery – "Urge to get better at stuff."
- Drives weekend musicians, open-source coders.
Purpose – "Desire to contribute to something larger than ourselves."
- Organizations flourishing today tie profit to a transcendent cause.

Policy: Once per quarter, 24-hour autonomy sprint.
- Choose anything to work on.
- Choose how & with whom.
- End with playful demo session (beer, cake).
Outcomes:
- Burst of software fixes & new-product ideas not produced under normal schedules.
- No cash "innovation bonus"—just freedom.

Hypothetical 1983 pitch (volunteers donate 20–30 hrs/week, give product away) would seem crazy.
Reality:
- Linux powers $\frac14$ of Fortune 500 servers.
- Apache dominates web-server market.
- Wikipedia crowdsources global knowledge base.
Motivators identified by economists:
1. Challenge.
2. Mastery.
3. Making a contribution (purpose overlay).

When $\text{Profit motive}$ becomes unmoored from $\text{Purpose motive}$ :
- Ethical lapses.
- "Crappy products," "lame services," uninspiring workplaces.
Emerging trend: Blur lines between profit & purpose.
- Skype founder: "Disruptive for making the world a better place."
- Steve Jobs: "Put a ding in the universe."

Treating people as self-motivated humans ≠ carrot-and-stick horses →
- Higher engagement & innovation.
- Healthier organizational culture.
- Potential societal benefits.

$\text{If–then}$ monetary rewards:
- Great for algorithmic tasks.
- Counter-productive for creative/cognitive tasks.
Effective pay strategy: \text{Pay enough} \rightarrow \text{remove concerns}.
Sustainable high performance rests on Autonomy + Mastery + Purpose.
Case studies (Atlassian, Linux/Apache, Skype, Apple) exemplify intrinsic-motivation success.

Mechanical task performance: P \propto R (Performance proportional to Reward).
Cognitive task performance (beyond rudiments): P \propto \frac{1}{R} (inverse relationship once baseline needs met).
Motivation utility model: U = f(\text{Autonomy}, \text{Mastery}, \text{Purpose}) $when$ Salary \geq S_{\text{threshold}}$$.