7
do
8
4
of Births and Deaths should be the Director of Medical Services or the govern- ment's statistical officer. If the Medical Department wishes to get rid of the work of registration as it did in Singapore the Statistical Department can take it over. If it wishes to keep the registration it can do so without seriously affecting the Colony's statistics. The same is true of almost any department-in some cases it may be more convenient, it may even be necessary, that the statistics should be collected by the appropriate department rather than by the statistical office. In other cases it may be easier, or more economical or more efficient for the statistical department to do the work. In any case a certain amount of collaboration be- tween the two is inevitable, but in the final result the question is not, I think, one of very great importance.
This is the more so in that as time goes on the practical work of collecting statistics must inevitably become a matter of the merest routine which will require little or no attention from the administrative officer in charge (whether at the head of the statistical office or of some other department). This is in effect what has happened in Singapore-the Registrar General of Statistics is an administrative officer who is not concerned to any appreciable extent with the details of the running of his department, while executive control is in the hands of the Deputy Registrar General. I think it is almost inevitable that sooner or later the statistical office will tend to take over more and more of the statistics of other departments but it would be perhaps a mistake to attempt too rapid progress in this direction.
For these reasons again I am inclined to suggest that the development of the new office should be a gradual one. It might for example in the first instance con- fine itself to a minimum of activities connected with imports and exports, general trade, financial and shipping returns. The next step might be to take over birth and death registrations, health and hospital, sanitary and water statistics. The final step might be to take over labour, unemployment, social, police, prison and judicial returns. The three broad classes outlined above, economic, biometric and social overlap to some extent-for example the census of population concerns all three and there would be. I think, no point in making any rigid distinction between the three groups.
I simply offer them as a proposal, on the broadest lines, for the gradual development of the statistical department.
I hope that it will be possible, even at an early date, for the statistical depart- ment to offer its resources for the benefit of other departments. For example if, say, the educational department desired to obtain a survey of the school children shewing the social strata from which they come, their physical condition, their ability, physical and mental, etc., it would be possible for the whole of the work of enumerating, sifting, counting and analysis to be done by the statistical office staff. In other words the existence of a statistical office would, I think, make it possible for the various departments of government (or even for the university or the chambers of commerce) to undertake colony wide surveys on a fairly ambitious scale on any subject of administrative importance at small cost to the government. It has of course always been open to any department in the past to make such a survey but any really comprehensive survey, in the absence of a trained and equipped count- ing-sorting staff. has I think always been ruled out of the sphere of practical pos- sibilities on the score of time and expense,
are certain
The solution to this difficulty lies in methods of rapid sifting and enumeration. One method is to use electrical sorting and counting machines. Apart from the question of expense (a minimum of this equipment would cost perhaps $100.000 together with the salary of a permanent European technician) there objections to mechanization of statistical work. I should add that this is my personal view which would not I think be generally subscribed to. In the first place it is inelastic. If two machines just suffice for the work in hand and one goes out of order the result is chaos. If two machines each do 50% of the work, working at full capacity, and the work increases by 10% a third machine has to be purchased which will be lying idle for fourth-fifths of its time, and $40,000 of idle capital is a heavy charge even on the biggest installations. Another objection is that a machine is a bottle neck, all the returns have to pass through it, and if there is any delay at any stage before the returns come out of the machine the whole process is held up.
There are however methods of sorting which are even more rapid than machine sorting, and I think that methods of counting can be devised which will prove almost as rapid as machine counting. If this proves to be the case in practice it is hoped that our statistical office will be able to make use of sorting-counting equipment at a negligible cost as efficient as the more expensive electric installations. If the CX- periment proves successful it should simplify the 1941 census considerably.
Another way in which the statistical office might from the beginning he of assistance to other government departments would be for it to undertake the publication of a weekly or monthly or quarterly statistical gazette containing every variety of government statistics. Such a gazette might, inter alia, contain the following statistics :—
Revenue and expenditure.
Meteorological.
Health.
Indices of wholesale and retail prices, stock exchange prices, cost of
living.
Birth, death, marriage and migration returns.
Shipping, railway and air returns.
Police, prison and justice returns.
Imports and exports.
are
If such a gazette came into being, it is hoped that it might be possible to present at any rate some of the statistics in a more digestible form than is usual with government statistics. It is perhaps true that a very great proportion of the
statistics wasted because the labour expended on government statistics is presented in detail and not at large. Presumably somebody is interested in the export on a Greek steamer of two cases of rubber shoes to Ecuador, but many more people will be interested in the wider implications that our trade with the Americas It is in rubber manufactures generally or in foreign ships generally is increasing. hoped that a statistical gazette would, while preserving details for anyone who is interested, enable emphasis to be placed on matters of real importance.
Statistics generally are liable to get out of hand. The enthusiasm of statis- tical departments in almost every government tends I think to run away with their sense of proportion and the result is that economic statistics are produced in such detail that their publication ceases to be an economic proposition and social statistics are produced with an energy that might be better employed in improving social conditions. In Hong Kong the production of unnecessary statistics is to some extent forced upon us by the home government, the League of Nations and other bodies, but the danger still remains that our purely domestic statistics may get out of hand. For this reason I am inclined to suggest that a statistical advisory committee, comprising perhaps representatives from government and public of the economic and social sides, might be of value. Such a committee would I think see the general value of statistics in a wider perspective than would be possible for the officer in charge of the statistical office. But this again is a matter that could be put aside for future consideration,
In Singapore the Statistics Department which has evolved as the result of many years growth and might profitably be taken as a model for our own deals with:-
Trade Statistics, Shipping Statistics,
Foodstock Statistics,
Birth and Death Registration,
and certain statistics of local interest such as Rubber.
Pa
9
END
6
The arrangement of the Department is somewhat as follows:-
STATISTICS DEPARTMENT.
Registrar General of Statistics
Deputy Registrar General
Pa
Assistant Registrar
Assistant
Imports and
Registrar Rubber
Portuguese Officer in Charge
Assistant
Registrar
Exports
Births and Deaths Registration
Shipping
European Lady Officer in charge of counting and sorting
Staff
Staff
Staff
Staff
Staff
Food Stocks
Staff.
The department operates under the sanction of the Statistics Law which lays down in effect that upon requisition by the Registrar General of Statistics any person, firm or company is bound to afford any information in their possession regarding the circumstances of their business, household, etc., while the Governor in Council may make rules regarding the collection of information.
Finally I should like to record my gratitude to the Colonial Office and other government departments in London for the very willing and courteous assistance they afforded me, and to the Government of Hong Kong for detailing me to under- take this very interesting work.
To summarize what has gone before :-
(1) It is suggested that the new statistical department should be of gradual growth.
(2) It is hoped that on the theoretical side the mathematical theory of statistics may occasionally be of value in matters of government policy.
(3) It is suggested that a monthly statistical gazette be published embrac- ing all government statistics and containing in addition such simplification of the data as may be possible.
(4) It is hoped that by the introduction of simplified counting and sorting methods surveys on matters of administrative importance may be made at small cost for any department or other public body that is interested.
(5) It is suggested that an advisory committee on statistics be formed. (6) It is suggested that the Straits Settlements Statistics Department might be taken as a model of what the local department might ultimately become.
D. KELVIN-STARK.
15th May, 1939.
7
7
53921
No comments yet.
Private notes are available after approval.