Thus it appears there are three modes in which modern capitalist methods of production cause temporary unemployment. (1) Continual increments of labour-saving machinery displace a number of workers, compelling them to remain wholly or partially unemployed, until they have "adjusted" themselves to the new economic conditions. (2) Miscalculation and temporary over-production, to which machine industries with a wide unstable market are particularly prone, bring about periodic deep depressions of "trade," temporarily throwing out of work large bodies of skilled and unskilled labour. (3) Economies of machine-production in the staple industries drive an increasing proportion of labour with trades which are engaged in supplying commodities, the demand for which is more irregular, and in which therefore the fluctuations in demand for labour must be greater. Most economists, still deeply imbued with a belief in the admirable order and economy of "the play of economic forces," appear to regard all unemployment not assignable to individual vice or incapacity as the natural and necessary effect of the process of adjustment by which industrial progress is achieved, ignoring altogether the two latter classes of consideration. There is, however, reason to believe that in an average year a far larger number of the "unemployed" at any given time owe their unemployment to a temporary depression of the trade in which they are engaged, than to the fluctuations brought about by organic changes in the economic structure of the trade. The size and importance of the "unemployment" due primarily to trade depressions is very imperfectly appreciated. The following statistics of the condition of the skilled labour market in the period 1886-92, based upon the reports of twenty-two trades unions, have an important bearing on this point:-- Year. Percentage out of work. 1886 10.1 per cent.[195] 1887 8.6 " 1888 4.4 " 1889 1.8 " 1890 2.6 " 1891 4.45 " 1892 7.33 " 1893 7.9[196] " When it is remembered that these figures apply only to the well-organised trades unions, which, as a rule, comprise the best and most highly-skilled workers in the several trades, who are less likely than others to be thrown out in a "slack time," that the building and season trades are not included in the estimate, and that women's industries, notoriously more irregular than men's, are altogether ignored, it will be evident that these statistics very inadequately represent the proportion of unemployment for the aggregate of the working classes at the several periods. The _Report on Principal and Minor Textile Trades_ deducts 10 per cent. from the normal wages to represent unemployment, though the year 1885, to which the figures refer, is spoken of as "fairly representative of a normal year."[197]