Recent empirical data indicate that the marginal propensity to consume may not decline as much with rising income as Keynes and others assumed it would. If this is true, how does it affect the validity of the income-expenditure approach as a tool of analysis?