Philosophical Transactions · 1693 · Breslau, now Wrocław
Halley’s Breslau table, 1693: what a life annuity was worth, age by age
From five years of Breslau’s births and burials, Edmond Halley estimated how many of the city’s people were alive in each year of life, starting from 1,000 in their first year. From that table he priced a pound a year for life, at 6 % interest, for age 1 and every fifth age to 70, after what he called a not ordinary number of arithmetical operations. Pick an age and work it again.
- In this year of life, per 1,000
- Price, years’ purchase
- Halley printed
- The Crown’s price
- 7.14
A city that counted its dead
Breslau, in Silesia, published monthly bills listing deaths by age and sex, alongside counts of births. Henri Justel passed the figures for 1687 to 1691 to the Royal Society, and Edmond Halley, then 36, used them. In those five years 6,193 were born and 5,869 buried; Halley reported 1,238 births and 1,174 burials a year. He chose Breslau because he believed few people moved in or out and births only slightly exceeded burials, so the deaths at each age could stand for how a generation dies.
His table gives the number of people in each year of life, from 1,000 in the first year to 20 in the eighty-fourth; the 107 people from 85 to 100 he lumps together. Beside it he printed sums for every seven years, and the whole adds to 34,000, his estimate of the city. Those sums check the transcription on this page: all twelve, and the total, match. John Graunt had published an estimated survival table from London’s bills of mortality in 1662; Halley built his year by year from recorded ages at death.
The price of a pound for life
A life annuity of this kind pays a fixed sum each year for as long as a named person lives. Its fair price is the sum, over every future year, of the payment times the chance that the person is alive to collect it, discounted for the years of waiting. Halley worked it at 6 %, the legal rate of the day, for ages 1, 5, 10 and so on to 70, and printed the prices in years’ purchase: how many years of payments the buyer should hand over at the start.
Recomputed from his table, at the two decimals Halley printed, ages 1 and 10 come back exactly and the printed ages from 20 to 55 within 0.05, across all four spreads of the oldest 107. Ages 60 to 70 come back within 0.15; there the spread of the oldest people matters. In all, 14 of the 15 prices are within 0.15. Priced at 5 % instead, or paid at the start of each year, or with the oldest 107 left out, the prices miss.
| Age | Halley printed | From his table | Difference (straight spread) |
|---|---|---|---|
| 1 | 10.28 | 10.28 | +0.00 |
| 5 | 13.40 | 13.01 | −0.39 |
| 10 | 13.44 | 13.44 | +0.00 |
| 15 | 13.33 | 13.18 | −0.15 |
| 20 | 12.78 | 12.77 | −0.01 |
| 25 | 12.27 | 12.26 | −0.01 |
| 30 | 11.72 | 11.71 | −0.01 |
| 35 | 11.12 | 11.14 | +0.02 |
| 40 | 10.57 | 10.54 | −0.03 |
| 45 | 9.91 | 9.87–9.88 | −0.04 |
| 50 | 9.21 | 9.16–9.17 | −0.04 |
| 55 | 8.51 | 8.46–8.47 | −0.04 |
| 60 | 7.60 | 7.53–7.55 | −0.06 |
| 65 | 6.54 | 6.44–6.48 | −0.08 |
| 70 | 5.32 | 5.18–5.24 | −0.11 |
Years’ purchase at 6 %, paid at the end of each year, compared at the two decimals Halley printed (unrounded, the worst case for ages 20 to 55 is −0.051, at 55). The range covers the four spreads of the 107 people aged 85 to 100; it widens from a few thousandths at 45 to a tenth at 70.
Two prices that do not come back
At age 5, Halley printed 13.40; his table gives 13.01. At age 15, he printed 13.33; the table gives 13.18. D. R. Bellhouse, working back from Halley’s 10.28 at age 1, found that only these two failed to fit, and wrote that it may be safely assumed Halley calculated them incorrectly. One possible slip, offered here only as an illustration: dividing by the number alive one year later, 710 instead of 732 at age 5 and 622 instead of 628 at age 15, gives 13.41 and 13.31, close to the printed figures but not equal to them.
Halley used the table to make a point about the Million Act of 1692, which raised money for the war with France and offered, among its terms, an annuity of £14 a year for every £100 on a nominated life, young or old: about 7 years’ purchase, as Halley put it. At 6 %, an annuity on a ten-year-old nominee was worth 13.44, about 1.9 times its price on his reckoning.
What the table cannot say
Halley’s text says that of 1,238 births a year, 890 reach their first birthday, yet the table opens with 1,000 in the first year of life. He did not explain exactly how he arrived at 1,000, and historians have proposed different reconstructions. Bellhouse also found that no spread of the 107 old people fully matches Halley’s prices at the oldest ages. These are historical valuations at 6 %, based on Breslau’s records for 1687 to 1691: not a life expectancy for anyone today, and not advice about annuities.
Sources
Every figure on this page is traced in the project’s source notes.
- Edmond Halley, “An Estimate of the Degrees of the Mortality of Mankind, drawn from curious Tables of the Births and Funerals at the City of Breslaw; with an Attempt to ascertain the Price of Annuities upon Lives”, Philosophical Transactions 17 (1693), 596–610, DOI, facsimile.
- D. R. Bellhouse, “A new look at Halley’s life table”, Journal of the Royal Statistical Society A 174 (3), 2011, 823–832, DOI.
- John Graunt, Natural and Political Observations … upon the Bills of Mortality, London, 1662, copy.