Dan J. Harkey

Educator & Private Money Lending Consultant

Lies, Damned Lies, and Statistics: The Origin, Evolution, and Real-World Lessons

The phrase “lies, damned lies, and statistics” is not just a cultural shorthand for skepticism toward data-driven arguments, but a powerful reminder of the potential for manipulation in numerical data. How would we feel if we realized that government statistics were one giant reservoir of lies and misstatements? So, what do you think?

by Dan J. Harkey

Share This Article

Summary

It suggests that while numbers appear objective, they can be twisted to mislead. But where did this phrase come from, and why does it remain so relevant today?

The Origin and Evolution of the Phrase

Pre-1890s: Early Roots

Before the phrase took its modern form, British humor often mocked unreliable testimony and expert opinion. Variants like “liars, damned liars, and experts” circulated in the late 19th century (e.g., a widely quoted 1885 Nature item referenced “simple liars, damned liars, and experts”). 

1891: First Printed Version

The earliest known print reference appeared in the National Observer (June 13, 1891):

“There are three kinds of falsehood: the first is a ‘fib,’ the second is a downright lie, and the third and most aggravated is statistics.” 

1895: Leonard Courtney’s Speech

Leonard H. Courtney—later President of the Royal Statistical Society—used a nearly modern form:

“There are three degrees of untruth—lies, damned lies, and statistics.” 

1907: Mark Twain Popularizes It

Mark Twain introduced the phrase to many American readers in Chapters from My Autobiography, attributing it to Benjamin Disraeli, although no evidence places it in Disraeli’s writings. 

Why It Endures

The phrase resonates because it highlights a timeless truth: numbers can be manipulated. In today’s era of big data, its relevance has only grown. It reminds us to question not just the data, but the motives and methods behind it, instilling a sense of caution and vigilance in our data interpretation.

Real-World Examples of Statistical Misuse

Tobacco Industry and Health Risks

For decades, certain industry-funded studies cherry-picked or framed data to cast doubt on the smoking–cancer link, illustrating how quantitative “evidence” can be weaponized.

Financial Markets and Risk Models

Before the 2008 crisis, structured-finance models relied on historical data that excluded systemic shocks, overstating the safety of mortgage-backed securities.

Political Polling

Campaigns tout favorable polls while ignoring margins of error, nonresponse bias, or unrepresentative sampling—manufacturing momentum from fragile numbers.

COVID-19 Data Interpretation

 Competing narratives frequently misread base rates, confound correlation with causation, or cherry-pick denominators when discussing infection and mortality rates.

Advertising and Consumer Behavior

Claims like “up to 50% off” or “average savings” often ride on outliers, not typical consumer outcomes.

Timeline

“Timeline from 1882 to 1907 showing the phrase’s evolution: 1882 ‘liars, great liars, and scientific witnesses’; 1885 ‘simple liars, damned liars, and experts’; 1891 first print with ‘statistics’; 1891 Charles Dilke uses ‘fib, lie, statistics’; 1892 Arthur Balfour quoted; 1895 Leonard H. Courtney; 1907 Mark Twain popularizes; then 20th century onward.” (Sources: Wikipedia and Quote Investigator.) 

 Key milestones:

  • 1882 – “liars, great liars, and scientific witnesses” appears in a UK context. 
  • 1885 – “simple liars, damned liars, and experts” appears in Nature (Nov 26, 1885). 
  • June 13, 1891 – First print linking statistics to the triplet in the National Observer
  • Oct 1891 – Sir Charles Dilke uses the “fib, lie, and statistics” formulation. 
  • June 28, 1892 – Arthur James Balfour quoted with near-modern form. 
  • 1895 – Leonard H. Courtney adopts modern phrasing and later becomes the President of RSS. 
  • 1907 – Mark Twain popularizes the quote in his autobiography, attributing it to Disraeli (no direct evidence in Disraeli’s works). 

How to Spot Misleading Statistics 

Be aware of Indological, extremist, religious, and financial beneficiaries who present the statistics.

·          Check the source: Who funded and collected the data?

·        - Examine the sample: Size, representativeness, response rate.

·        - Mind the metric: Mean vs. median vs. percentiles.

·        - Margins & intervals: Confidence intervals, error bars.

·        - Denominator discipline: Rates per relevant population.

·          Beware cherry-picking: Time windows and subgroups.

·        - Watch axes & scales: Truncated or dual axes.

·        - Correlation ≠ causation: Evidence beyond association.

·        - Replicability: Independent datasets/methods.

·        - Context & baselines: History, peers, benchmarks.

 Conclusion

“Lies, damned lies, and statistics” remains a powerful reminder: statistics are tools, not truth. They can illuminate reality—or obscure it. The key is critical thinking.

Sources & Further Reading