Introduction and Katie Johnson's Story

• Steven Dubner introduces archive episode about bad bosses
• Katie Johnson, 33-year-old data scientist from England
• Background: grew up near Bristol, University of Birmingham, series of IC (individual contributor) jobs
• Loved being IC - analyzing data, very good at job
• Decided to pursue management for three reasons:
  • Wanted more autonomy over work projects
  • Desired more impact/"seat at the table"
  • Believed success meant moving upward in company hierarchy
• Father was network engineer with no management aspirations

Katie's Management Experience

• Became head of data and analytics at software firm (customer acquisition)
• Management training included extensive personality test
• Strengths: critical thinking, attention to detail, courage, internal thinking characteristics
• Weaknesses: compassion, empathy, relationship building
• Managed roughly 10 people, promotion came with more money/prestige/leverage
• Massive increase in meetings: Tuesdays had 7 hours of meetings vs. 30 minutes as IC
• Virtual meetings during pandemic made it more draining
• Received good performance reviews as manager
• Personal misery: would finish workday, go to living room, put blanket over head and cry from boredom
• Satisfaction dropped from 8.5/10 as IC to 4-5/10 as manager (6 on great days)

The Horrible Boss Phenomenon

• Cultural stereotypes: screamer, sadist, idea stealer, passive aggressive
• Film examples: Blake (Alec Baldwin) in Glengarry Glen Ross, Peter's boss in Office Space, Miranda Priestly in Devil Wears Prada
• Scott Rudin Hollywood producer example: alleged abuses including smashing assistant's hand with computer monitor
• Gallup poll: 50% of American employees left job due to bad boss at some point
• European survey: only 13.3% rated current boss as bad
• Suggests Hollywood caricature may be overblown

Research on Boss Impact - Steve Tadelis

• Steve Tadelis, UC Berkeley economics professor
• Worked as manager at eBay and Amazon during sabbaticals
• Formal feedback surveys with 1-5 scale questions about manager effectiveness
• Research question: Do manager survey ratings correlate with actual firm value?
• Academic literature on boss impact not very advanced due to multiple variables
• Most research focuses on single quantifiable metric: productivity

Lazear, Shaw, and Stanton Study

• 2015 paper analyzed single firm (appears to be call center)
• Moving from average boss to high-quality boss increased productivity by 50%
• Example: 100 calls per shift under average boss → 150 calls under excellent boss
• Data couldn't identify what made good bosses effective

Tadelis-Hoffman Research

• Partnership with Mitchell Hoffman (University of Toronto)
• Studied high-tech knowledge-based firm (unnamed, possibly like eBay/Amazon)
• Data included:
  • Manager "people management skills" ratings from employee surveys
  • Employee performance scores, promotions, raises, patents filed, retention rates
• Key finding: manager ratings had no meaningful impact on employee performance, earnings, promotions, or patents
• One major exception: employee retention
• Moving from poor-rated to high-rated manager associated with 60% drop in attrition
• Effect strongest for retaining better employees (where it matters most)

Limitations of Boss Research

• Survey data reliability issues - economists prefer "revealed preference" approach
• Steve Tadelis privacy example: people say they value privacy but give away personal data for Snickers bar
• Gap between what people say vs. how they behave
• Compassion identified as important but unmeasurable factor
• Nicholas Bloom (Stanford): 20 years of leadership research, still no "secret recipe"
• Describes management literature as "airport bookstore pulp fiction"

The Peter Principle Research - Kelly Shu

• Kelly Shu, Yale School of Management finance professor
• Paper with Alan Benson and Danielle Lee in Quarterly Journal of Economics
• Lawrence J. Peter's 1973 theory: employees rise to level of incompetence
• Peter's observation: competent mechanic promoted to incompetent foreman
• Book "The Peter Principle: Why Things Always Go Wrong" sold millions of copies
• Peter declined corporate consulting offers to avoid rising to own level of incompetence

Peter Principle Study Methodology

• Data from sales performance management software company (like Salesforce)
• 40,000 business-to-business sales workers across 130+ US firms
• 5,000 managers, 1,500 promotion events
• Sales chosen because: 10% of US workforce, clear performance metrics
• Findings likely apply to other fields where skills differ between levels

Key Findings on Promotions

• Doubling sales performance = 30% increase in promotion probability
• Top sales worker in team of 5-6 has triple promotion probability vs. average
• Manager quality measured by value added to subordinate sales
• Critical finding: managers with double pre-promotion sales led to 6% decline in subordinate sales
• Low-sales managers before promotion were better at managing subordinates
• Firms could increase sales by 30% by promoting based on collaboration/experience vs. sales numbers

Why Firms Continue Peter Principle

• Firms aware of Peter Principle problem but choose to live with it
• Evidence: firms put less weight on sales numbers when selecting managers for large teams
• Not a mistake - firms face tradeoff between manager quality and worker motivation
• Two ways to motivate workers: higher pay or promotion opportunity
• Promotion cheaper than pay increases and provides resume value/social status
• Cultural taboo against discussing compensation makes promotion more attractive

Alternative Career Tracks

• Tech companies (Silicon Valley) offer dual career tracks
• IC (Individual Contributor) roles can reach VP level without managing anyone
• Engineering progression: engineer → distinguished engineer → lifetime distinguished engineer
• Consulting firms also use technical specialist tracks for PhDs
• These remain exceptions rather than widespread practice

Katie Johnson's Resolution

• Quit management job despite positive LinkedIn review from manager calling her "unicorn"
• Returned to data scientist role at different firm
• Satisfaction increased from 5-6/10 to 9.5/10
• Actually earning more money as IC than as manager
• Believes technical specialist route becoming more prominent and rewarded
• Notes dual career tracks still exclude technical specialists from high-level conversations
• Management retains connotation of seniority that technical roles lack
• Recently became freelance data coach/consultant

Broader Implications

• Management science not yet very scientific
• Most firms stick with traditional promotion patterns
• Peter Principle alive and well despite seeming absurd
• Katie Johnson's perspective: people rise to "ceiling of competence" rather than "level of incompetence"
• Status loss concern prevents many from stepping back from management
• Katie Johnson's advantage: doesn't care about others' opinions, prioritizes job satisfaction