When the debate adjourned last night, I was discussing the factors which determined the accuracy or otherwise of the cost-of-living index as an indicator of the relative variations in the cost of maintaining a certain standard of living. I showed that the standard of living selected as a norm might be lessened quite arbitrarily, that it might have no relation whatsoever to the actual living conditions of any section of the people, that the value of the index depends upon the degree to which the price movement of each group correctly represents the price movements of the group covered by the index as a whole. I also stated that the value of the index depends on the quantity of the articles actually consumed as well as on the prices of the articles, and that any change in the quality or quantity of the articles destroys the accuracy in the cost-of-living index as a reflex of actual living conditions. In general, I pointed out that over a period a cost-of-living index is reliable only so far as the commodities and the quantities upon which the index is based remain truly representative of the actual consumption of the class of the community to which the index is to apply.
The practical difficulty of maintaining the representative character of a cost-of-living index is very great, being due to the fact that the consumption relative to the aggregate consumption of the several commodities is not the same for all sections or classes within the community, and indeed is not the same for all individuals within a class. We find that exemplified in the report of the Cost of Living Committee itself. For instance, in a household where the income is under £3, we find that 61 per cent. of the total expenditure is upon food. Where the income is £3 that percentage decreases to 60.4. Where the income is £4 it falls to 56.3. Where the income is £5 it falls to 48, and where the income is £6 and upwards it falls to 41.4, upon food. In a household where the income is £6 and upwards 41 per cent. of the expenditure is represented by the purchase of food, whereas in households where the income is under £3, 61 per cent. of the expenditure is made upon commodities within that group. I have shown that as the income increases within a household the proportion which is spent upon food tends to decline. The converse of this is true with regard to rent and certain other groups of commodities upon which our cost-of-living index is based. For instance, in a household where the income is £3 and under, 15 per cent. of the expenditure is upon clothing, but where the income is £6 a week and over, the proportion of the expenditure upon that item rises to 17.7 per cent. In the case of rent in households where the income is under £3, the proportion of the expenditure ascribed to rent is 5.3 per cent. Where the income is £6 and upwards expenditure upon rent rises to 12.4 per cent. According to our classification here sundries include soap, pipe tobacco, cigarettes and miscellaneous items. That expenditure forms 9.2 per cent. of the total household expenditure where the income is £3 and under, and rises to 19.5 per cent. where the income is £6 and upwards.
The very fact that there is this extreme variation between the expenditures, not only, as I have said, between the several classes of the community in regard to one particular group, but also even for individuals within the same class, makes it impossible to construct a cost-of-living index for the country as a whole which, at the moment, will indicate with reasonable accuracy the change in the cost of maintaining the standard of living in any other section of the community except that in which the primary data to calculate the index have been originally secured. This fact receives attention in some countries.
In Belgium five series of cost-of-living numbers are applied to corresponding classes in the community, and in Switzerland there are three. In this country we have only one cost-of-living index. Remember that index, moreover, is based upon an entirely inadequate number of returns, as I showed last night. Our cost-of-living index is based upon 308 complete returns made between, I think, the 10th June, 1922, and 4th August, 1922. I submit that those returns are entirely inaccurate and entirely unreliable for the purpose of calculating what should be after all, so far as we can make it, a precise and reliable indicator as to the variation in the cost of living. Those returns, moreover, have been procured mainly from one class. According to the Report itself, the budgets received were from a very varied class of households, which, as will be seen from the following lists, embrace the principal industrial occupations. In that list there are unskilled labourers, fishermen, messengers, servants, pilots, barmen, carpenters, boatmen, dressmakers, bakers, embroidery workers, knitters, plumbers, road engine drivers, etc., but not a single civil servant. Out of the 308 returns which were received, according to this list, so far as I can see, there is not one return, and certainly no significant number of returns, no such number as would affect the result of the calculation, were made by civil servants. Notwithstanding that the index in this country at least is applied in the most general and most rigorous way to a large body of clerical workers, the privates and corporals in the cuff and collar brigade, whose mode of life is different, even if the real standard of living be lower, from that of the skilled industrial worker. The expenditure which the civil servant, particularly in the lower class grade, has to incur upon articles such as food, rent and furniture, I think it must be admitted, is very much higher than that which has to be met by a worker in an industrial occupation. He has got to maintain a decent appearance during all the hours of his work. The fitter or boiler-maker or motor mechanic can work all day in his dungarees, and can go home and change at night. In that way he can economise on the question of clothes, but a man who has to go into a Government office has to maintain at all times a dressy appearance, to put it that way. He has to be clean and tidy, and have clothes which are not ragged, torn or tattered. To me, looking at the rates of pay which some of the smaller civil servants have, it is a constant wonder how they are able to maintain the appearance they do. I can only assume that they do it by spending upon clothes what might be much better spent on food. Again, while the industrial worker may not have to spend so much upon clothes, he has, on the other hand, because his labour is manual, to spend a greater portion of his weekly income upon more expensive kinds of foodstuffs. For that reason it is quite clear to anyone who has examined this question that a cost-of-living index which has been constructed mainly upon household budgets which have been procured from industrial households cannot be properly applied as an indicator of the change in the cost of maintaining the standard of living which might be properly enjoyed by the lower grade civil servant or by any civil servant. In saying that, I do not mean to be taken that the standard of living which the present wages and salaries afford to the lower grade civil servant is an adequate one. I am merely discussing the cost-of-living index, to indicate the change in the standard of living, which, as I said, has no relation whatever to the actual standard on which the civil servants live.
Again, as I think I pointed out last night, the difficulties in compiling an accurate cost-of-living index are created by reason of the fact that variations in the prices of the groups are not uniform throughout the whole territory, if that territory be widespread. That fact also receives recognition abroad in certain countries. Speaking off-hand, at the moment, I think in Austria and in Germany there are regional cost-of-living indices and it would seem to me that if the wages and salaries of the civil servants are to be made dependent in any way upon the cost of maintaining a certain standard of living so far as this country is concerned we would almost want to have recourse to regional indices because the cost of living in Dublin is very different from the cost of living in Cork, Waterford, Limerick, and greatly different from the cost of living in some of the smaller towns throughout the country.
Last night I referred to the statement with which the Ministry of Economic Affairs in August, 1922, prefaced its report in which they say "The inquiry applied to the whole of Ireland, and its basis is therefore broad enough to be reasonably dependable for any practical purpose to which a calculation as to changes in cost of living can usefully be applied." I deny that in toto. I think it is a ridiculous statement. I think it is generally recognised that you cannot get that sort of basis in this country which the Minister for Economic Affairs in the Provisional Government claims to have found in August, 1922. Certainly a cost-of-living index compiled on returns secured from the whole of Ireland cannot be made to apply in any fair or just way to the variations in the standard of living of the civil servant. The great mass of civil servants are concentrated in our large cities. The overwhelming mass is concentrated in Dublin. There are many factors which would make it reasonable to say that a cost-of-living index for the city of Dublin alone might be compiled, but to say that a cost-of-living index which has been compiled as a result of returns secured in the first place from an inadequate number of households and in the second place mainly from industrial workers, and in the third place from 308 families scattered throughout the length and breadth of the country, would be fair and just in its application to a section of the community which is largely concentrated in Dublin, I think, as I said before, is ridiculous and could not be defended in any way.
I would like to hear the Minister addressing himself to that aspect. There are circumstances here the mere recital of which ought to convince the Minister of that. The cost of housing in Dublin is very much higher, and it tends to go higher. I do not believe that there has been any significant fall in the cost of houses in Dublin during the past few years. Houses which for many reasons may be available in the country— fairly large houses which have become vacant owing to the fact that sections of the people have left this country and which have helped to solve the problem of housing the middle classes—are not available here in Dublin except at rents which hear absolutely no relation to the figure for rent which was selected and which was chosen by the Committee as the basis for its loading of that item in the cost-of-living index. I will give the House the exact figure and let them judge as to whether what I am saying is true or not. In page 10 of the report, paragraph (c), the Committee prefaced this part of the report by an admission which is intended to disarm criticism: "A point likely to give rise to criticism is the low percentage increase shown for rent. This figure is, however, the result of comparing the actual returns sent us by the town clerks and rate collectors." According to this the average weekly rent in July, 1914, was 4/-; March, 1922, 5/2d.; June, 1922, 5/2d. It is on that percentage, the variation of just a little over 25 per cent. for the cost-of-living index, which applies to the majority of civil servants living here in Dublin, where you would not even get a dog kennel let alone a house for 5/2d. a week, has been based.
There is one other point I should like to deal with, and that is the method of calculating the index figure. It is based on the old British cost-of-living index, which was calculated according to the system of arithmetical averages. That system has fallen greatly into disfavour. It is admitted on all hands that it gives rise to serious inaccuracies. The method of calculating the index figure by the geometric average is coming into favour, and is practically accepted as being the only correct one.
One other argument in favour of the motion is the fact that it is now universally recognised that the calculation of this cost-of-living index should be examined from time to time; that the commodities in the groups should be examined individually to see that they are actually representative of the purchases of the community as a whole; that the loadings according to the groups have to be checked periodically, and that in short there is an unanswerable case for a general periodic revision of the cost-of-living index. This cost-of-living index of ours was originally compiled in March and June, 1922, when conditions in this country were very unsettled, compiled at a time when the variations between prices in each locality were wide, when the variations from month to month were wide, and when there was really no representative figure that could be taken to form a basis of any cost-of-living index. Even on this ground alone, revision is necessary, but in addition to that ten years have passed.
I have shown that the index in its present form could only apply, if it applies at all, to certain sections of the industrial workers. It certainly could not apply to the general body of civil servants, to whom the Government apply it in its full rigour. There is undoubtedly a case, which I do not believe the Minister can answer, for investigation and examination as to the manner in which the cost-of-living index is computed. If we are to continue to try to regulate wages by means of a cost-of-living index—and we do, either consciously or subconsciously; the general tendency is in wages disputes and in matters like that to appeal to the cost-of-living index as an arbiter, as setting a norm by which the issue in dispute may be decided—it is essential that it should be as accurate as human ingenuity and mathematical skill can make it. We should have that cost-of-living index for more than one section of the community. We should have it for the industrial worker, because the commodities which he consumes and the rent which he pays, and his expenses upon certain items, such as clothes, differ considerably from the expenditure, say, of clerical workers upon the same things. There should be an index for the industrial worker, there should be an index for the clerical worker, and, if we want to regulate remuneration higher than that by means of the standard of living, there ought to be an index for the middle-class worker.
Attention called to the fact that a quorum was not present, House counted and a quorum being found present.