H.—2

8

the facts for the "members sick," it will be perceived that only 24 per cent, of them withdrew during the same period. The experience in respect of the entrants into the society discloses precisely the same features. The proportion of deaths among members who never received any sickness-allowance appears high, and points certainly to a large number of sudden deaths. That the proportion of members who were either sick, not sick, or died whilst under observation varies with the age of the assured the following statement shows: —

As life progresses, the proportionate number of "not sick" members diminishes, and the percentage to the credit of the deaths column materially augments. Taking the members as a whole—in other words, ignoring for a moment their age-distribution— it will be found that, in respect of the 127,269 years of life at risk, — The average mortality per cent, per annum was ... ... ... 0-817 „ weeks of sickness per member per annum was ... ... 1-416 „ „ to every member sick was ... ... 5-365 To every 100 members the number sick was ... ... ... 26-394 It must be understood that these ratios are not applicable for comparisons with the results of the workings of other friendly society organizations, unless it is strictly true that not only the agedistribution of their members, but likewise the locality, occupation, &c, influences are similar in both organizations. On no other basis can scientific deductions be drawn. In ignorance of the correct methods of dealing with rates of sickness and mortality most extraordinary comparisons are often made between the workings of society and society ; and the deductions drawn from the results are most fallacious. To take merely one point, namely—the age-distribution of the members—the extent to which the rates of sickness and mortality vary with age is well shown in a later portion of this report ; and, thereby, it would be extremely easy to draw up several hypothetical organizations, all of which should experience a normal rate of sickness and mortality, and yet disclose the most divergent general rates of sickness and mortality when compared with one another irrespective of the variable ages of the assured in the different groups. If this is the case when only one disturbing element —namely, age-distribution —is brought into account, how much more so will it be when is borne in mind the influence of locality, occupation, density of population, &c? It would be well to remember these circumstances when unskilled statisticians are found drawing most erroneous comparisons between the working of one society compared with another. Only a skilled expert can venture upon such a task, and only then when complete data is at his disposal, so that due care can be taken to compaxe as far as practicable only like with like. It is not uncommon to find persons drawing comparisons between the death-rate prevailing among some recently - formed friendly society organization and the rate of mortality occurring among the whole population of the country. Needless to say this is a most fallacious process. Like is not being compared with like. In the friendly society, for example, there are no children, and few, if any, very old persons. Now, in the death-rate for the whole population, the high mortality for infant life, and the deaths among those advanced in years, account for quite one-half of the gross rate of mortality. For a comparison which could take credit for being-of at all an accurate character, therefore, the rate of mortality in the friendly society should be gauged by the deaths which occur in the whole population for persons of a similar age-group. ... In the following statement there is displayed the rates of entry, withdrawal, and sicknessclaimants in your society, with the proportional age-distribution in the respective groups : —

To every 100 Members in the Number who, whilst undi Ten Yeai Tents, 1st Jam ir Observation, ■s 1878-87, jary, 1878, the during the To every 100 Members who became Free during the Ten Years 1878-87, the Number who, whilst under Observation, Ages. Were Sicknessclaimants. Were not Sick. . claimants. B10K - Died. Total. Died. Total. 15 — .. 25 — .. 35 — .. 45 — .. 55 — .. 65 and upwards 57-0 53-4 58-8 57-4 49-9 26-4 1 39-3 42-2 35-4 30-0 21-0 3-9 3-7 4-4 5-8 12-0 28-5 69-7 100-0 100-0 100-0 100-0 100-0 100-0 40-4 38-2 40-1 39-5 50-0 I 58-1 60-7 57-9 56-9 44-4 1-5 11 2-0 3-6 5-6 100-0 100-0 100-0 100-0 100-0

Entri uits. Withdi vawals. Sickness-claimants. Ages. Rate of Admission per Cent, per j Annum. Proportional Agedistribution. Rate of Withdrawal per Cent, per Annum. Proportional Agedistribution. Number of n ,. , Sick-claims Pro P° rtl0nal p , Age- " . ' \ distribution, per Annum. Weeks of Sickness per Member per Annum to every Member sick. Under 25 25 — 35 — 45 — 55 — 65 — 75 and upwards 43-08 19-91 17-18 9-52 0-33 0-12 52-52 27-98 14-87 4-56 0-06 0-01 12-94 10-79 8-20 5-03 1-18 0-37 39-21 37-30 1-7-22 5-73 0-49 0-05 27-94 24-56 25'60 26-94 26-35 40-58 62-54 i 30-64 31-05 19-94 11-60 4-22 1*99 0-56 3-78 4-15 5-27 6-92 11-07 20-85 32-46