Market Power and the PC firm
A fundamental assumption of traditional (neoclassical) theory is firms in a perfectly competitive market are price takers and face a perfectly elastic demand curve. However, the assumption of a perfectly elastic demand curve and a negatively sloped demand curve are contradictory. As Keen notes the market demand curve is the summation of the firm demand curves and that one cannot add flat line segments to derive a negatively sloped demand curve. If you add line segments all you get is one long line. One response to this criticism is that the firm "perceive" the demand curve as perfectly elastic even though it is not. However, this response violates the assumption of perfect knowledge. It is impossible to fool someone with perfect knowledge. A second response is that the output of each firm is infinitely small and that each firm's demand curve is represented by a point on the market demand curve. This explanation shares the problems of flat firm demand curves - if you add together a bunch of points you end up with a point (points have no dimensions). The answer is provided by one of the leading neoclassical theorists. Varian states, the firm demand curve has a "little bit" of negative slope and that the market demand curve should be viewed simply as an approximation of aggregate demand. Of course, first approximations lack the mathematical precision infused into the theory of the firm by the mathematization of neoclassical microeconomics and undercuts some of the assumptions of welfare economics particularly. The ultimate problem is that you simply cannot aggregate consumer indifference curves to generate a social utility function. The positions and values of the curves will constantly shift as you change relative prices. If PC firms face a negatively sloped demand curve then the market demand curve is negatively sloped. The firm would also have a marginal revenue curve and would produce where MR = MC rather than where P = MC.