These simple illustrations may serve to remind us that there are difficulties other than those of statistical measurement. Not only can the index of consumers' benefit rise over time without any actual experience of benefit - simply because ' real' income is increasing, and therefore people are prepared to pay more for the ith unit of any good, or service, whose actual utility to them in fact remains unchanged - but such an index can rise concurrently with an actual reduction of benefit. The more significant parts of the other factors which may affect any consumer's surplus (and rent) analysis relate to the constancy of the prices, and/or availabili ties, of the close substitutes, and complements, for the good or service in question. The more effective is the good y as, say, a substitute for x, the smaller will be the consumer's surplus on purchase of x at the given prices. Raise the price of y, and ultimately withdraw it from the market, and this simultaneously reduces the welfare of the consumer and increases his measure of con sumer's surplus on purchases of x.