Keynesian System with Aggregate Demand and Aggregate Supply

The aggregate demand curve represents the set of combinations of the general price level and real output for which the goods market and the money market are simultaneously in equilibrium. In the Keynesian tradition, this derivation begins with the fundamental behavioral relations of the private sector—the consumption function and investment demand—and then incorporates the money market. By allowing the price level to vary, we expose the linkages through which price changes influence real money balances, the interest rate, and ultimately real expenditure.

The foundation of aggregate demand in a closed economy is the equality of output and the sum of consumption, investment, and government purchases. Real consumption expenditure is assumed to be a stable function of real disposable income. A widely used linear specification is C = a + b(Y - T), where a is autonomous consumption, b is the marginal propensity to consume, Y is real income, and T is real net taxes. The term b(Y - T) is induced consumption, rising and falling with current disposable income. Real investment spending is treated primarily as a function of the real interest rate, with a possible autonomous component reflecting business confidence, expected future demand, and technological progress: I = I0 - d r, where I0 is autonomous investment and d is the interest sensitivity of investment.

The goods market equilibrium condition is Y = C + I + G. Substituting the behavioral equations and rearranging yields the IS equation: Y = (1/(1-b))(a - bT + I0 + G) - (d/(1-b)) r. Define the simple Keynesian multiplier as 1/(1-b) and collect all autonomous spending components into A. Then the IS equation is Y = multiplier times A minus multiplier times d times r. This describes the IS curve: for any given real interest rate, it gives the level of output that equates planned spending to actual output. The IS curve slopes downward because a higher interest rate reduces investment, which lowers aggregate demand directly. Because of the multiplier, this initial drop is amplified: the fall in investment reduces income, which reduces induced consumption, which reduces income further.

The IS curve alone does not determine either output or the interest rate; we need the money market. The demand for real money balances follows the standard Keynesian liquidity preference: L = kY - h i, where i is the nominal interest rate. Assuming zero expected inflation for simplicity, the nominal and real rates are equal. The parameter k captures the transactions demand for money, while h captures the speculative or interest-sensitive demand. The real money supply is the exogenous nominal money stock deflated by the price level. Money market equilibrium requires M/P = kY - h r. Solving for the interest rate yields the LM equation: r = (k/h)Y - (1/h)(M/P). The LM curve is upward sloping: a higher income raises money demand for transactions, so for a given real money supply, the interest rate must rise to choke off speculative demand and restore equilibrium.

To derive the aggregate demand curve, we combine the goods and money market equilibrium conditions by substituting the LM equation into the IS equation. This yields a single relationship between output and the price level. The AD curve is downward sloping in (Y, P) space—an inverse relationship between the price level and the quantity of real output demanded. The mechanism is the Keynes effect: a rise in the price level reduces real money balances, shifting the LM curve upward, raising the equilibrium real interest rate, which in turn curtails investment spending. The drop in investment reduces aggregate demand both directly and through the induced consumption multiplier, resulting in a lower equilibrium output.

The AD curve can be written compactly as Y = gamma1 times A + gamma2 times (M/P), where gamma1 is the fiscal multiplier in the presence of money market feedback and gamma2 is the monetary multiplier. The fiscal multiplier gamma1 is smaller than the simple multiplier because of the crowding-out term in the denominator. This reduction reflects the crowding-out of investment that occurs when an expansionary fiscal policy raises the interest rate. The monetary multiplier gamma2 captures the direct effect of real money balances on aggregate demand. The AD curve shifts rightward when autonomous spending increases (expansionary fiscal policy) or when the nominal money supply increases (expansionary monetary policy). The steepness of the AD curve depends on the behavioral parameters: it is flatter when investment is highly sensitive to the interest rate, money demand is insensitive to the interest rate, or the marginal propensity to consume is high.

In the Keynesian system, the short-run aggregate supply curve captures the relationship between the overall price level and the quantity of output firms are willing to produce when some nominal frictions—typically in the labor market or in the price-setting process—prevent instantaneous adjustment to a full-employment equilibrium.

The derivation builds the SRAS curve from microfoundations by modeling the behavior of individual firms that operate under imperfect competition and have the power to set prices. Consider an economy populated by a continuum of monopolistically competitive firms, each producing a differentiated product. A representative firm faces a downward-sloping demand curve of the constant-elasticity form. The firm produces its variety using a short-run production function that depends only on labor input (capital is taken as fixed). Diminishing marginal returns to labor are the natural consequence of a fixed capital stock in the short run; they will play a central role in determining the slope of the SRAS curve.

The firm chooses its price to maximize profit, taking the nominal wage, the aggregate price level, and aggregate demand as given. In a flexible-price equilibrium, the first-order condition is the familiar equality of marginal revenue and marginal cost. This yields the optimal price as a markup over nominal marginal cost. The desired gross markup is determined solely by the elasticity of demand. Because all firms are symmetric, they will choose identical prices in equilibrium, so the aggregate price level satisfies P = markup times MC.

Nominal marginal cost derives from the firm's labor hiring: to produce one more unit of output, the firm must increase labor, at a cost of the nominal wage divided by the marginal product of labor. Under symmetry, aggregate employment equals each firm's labor input. Because the marginal product of labor is strictly decreasing, we can define the marginal product of labor as a decreasing function of output. Intuitively, producing more output requires more employment, which drives down the marginal product of labor. The equilibrium price level is an increasing function of output.

The short-run character of the Keynesian AS curve is now obtained by introducing a crucial friction: a temporarily fixed, or sticky, nominal wage. Suppose that at the beginning of the period the nominal wage is predetermined, either because of labor contracts, staggered wage setting, or a simple behavioral rigidity. With a sticky nominal wage, a firm that wishes to sell more output must hire more labor. As employment rises, the marginal product of labor falls (diminishing returns) and the real marginal cost rises. To restore the desired markup, each firm raises its price. In the aggregate, the price level therefore increases as output expands. Conversely, when demand is weak and output is low, employment and marginal cost are low, pushing the price level down. Hence, the SRAS curve slopes upward.

Two polar cases illuminate the range of slopes the SRAS curve can take. If the production function is linear with constant returns to labor, the marginal product is constant, and the SRAS curve is horizontal. Firms can supply any amount of output at the same price as long as the nominal wage is fixed. This corresponds to the "extreme Keynesian" case: the price level is fully demand-determined in the short run, and output is completely demand-driven. If, instead, the nominal wage were free to adjust continuously so as to maintain labor market equilibrium at the full-employment level, then output would be fixed at the full-employment level regardless of the price level. The SRAS curve would be vertical—the classical case. In the Keynesian short run, however, wage stickiness prevents such adjustment, giving rise to an upward-sloping aggregate supply curve.

The traditional Keynesian analysis often conceptualizes the aggregate supply curve as having three distinct ranges. The Keynesian range is horizontal: in deep recession, the economy operates with massive excess capacity, and firms can increase output without experiencing diminishing marginal returns. Output is entirely demand-determined. The intermediate range is upward-sloping: as the economy recovers, bottlenecks emerge and diminishing marginal returns set in, so the price level must rise to induce firms to expand production. The classical range is vertical: at full employment, all available resources are fully utilized, and further increases in aggregate demand only generate inflation.

The defining feature of the short-run Keynesian framework is the assumption of fixed money wages (or nominal wage rigidity). This assumption captures institutional realities such as long-term labor contracts, union bargaining, or money illusion. Under this regime, the economy can operate below full employment, and aggregate policy interventions transmit to the real economy primarily through price-level adjustments that alter the real wage.

Consider an expansionary open-market operation that raises the nominal money supply. At the initial price level, the LM curve shifts right. For a given price level, the IS-LM system yields higher output and a lower interest rate. Hence the AD curve shifts rightward: at each price level, output demanded is higher. With the upward-sloping AS curve, the increase in AD raises the equilibrium price level as well as output. The rise in the price level dampens the expansionary impact on real output by reducing real money balances—the higher price level partially offsets the LM shift. Output and employment unambiguously rise. The price level rises, but by less than the increase in nominal money, so real balances increase. The interest rate may rise or fall, depending on the relative strengths of the transactions demand for money and the interest sensitivity of investment. The transmission mechanism is: an increase in the money supply shifts LM right, lowering the interest rate, stimulating investment, and raising output (and via the multiplier, consumption). The rise in output increases employment; the resulting increase in the price level partially curtails the rise in real balances.

Now consider an increase in government expenditure, financed by bond sales. At the initial price level, the IS curve shifts right. For a given price level, the IS-LM system yields higher output and a higher interest rate. Thus the AD curve shifts right. With upward-sloping AS, both output and the price level increase. The rise in the price level reduces real money balances, shifting LM left. This crowds out some of the initial expansion. Output and employment increase, but by less than the IS-only multiplier would suggest, because of the interest-rate-induced crowding out and the price-induced crowding out (the latter unique to the flexible-price, fixed-wage model). The price level rises. The interest rate unambiguously rises. The increase in government spending directly boosts aggregate demand and income, raising transactions demand for money. With a fixed nominal money supply, the interest rate must rise to restore money market equilibrium. This rise in the interest rate dampens investment (partial crowding out). The additional rise in the price level (which reduces real balances) further raises the interest rate, reinforcing the crowding-out effect.

Both policies shift the AD curve right, raising output, employment, and the price level. The critical difference lies in the behavior of the interest rate. Monetary expansion tends to lower the interest rate (unless supply-side constraints are extreme), while fiscal expansion raises the interest rate. This distinction is central to the policy assignment debate: monetary policy works through the interest-rate channel, stimulating investment; fiscal policy directly adds to demand but crowds out private spending through higher interest rates. The efficacy of these policies in stabilizing output and employment hinges critically on the interest elasticities of money demand and investment, as well as the degree of nominal rigidity prevailing in the economy.

The general multipliers elegantly nest several classic macroeconomic extremes. In the liquidity trap, where money demand is perfectly interest-elastic, fiscal policy is fully effective with zero crowding out, while monetary policy is entirely impotent. In the classical case, where money demand is entirely insensitive to the interest rate, fiscal expansion results in complete crowding out, as the entire increase in demand is absorbed by a higher price level and interest rate, leaving output and employment unchanged. When investment is interest-inelastic, monetary policy fails to stimulate output, while the fiscal multiplier reverts to the simple Keynesian value because the interest rate rises without depressing investment.

A shift of the aggregate supply curve occurs when, for any given price level, the quantity of output that firms are willing to produce changes. Such shifts can originate from three broad sources. First, nominal wage changes: an exogenous increase in the nominal wage—due to higher minimum wages, stronger union bargaining power, or supply-side wage-push pressures—raises the real wage at every price level, reduces labor demand, and therefore lowers the quantity supplied at each price level. Graphically, the SRAS curve shifts upward (leftward): a higher price is now required to elicit any given level of output.

Second, productivity shocks: a rise in the technology parameter raises the marginal product of labor at any employment level and therefore shifts labor demand to the right. For a fixed nominal wage, this implies that any given real wage is compatible with a higher level of output, and the SRAS curve shifts downward (rightward): more output is forthcoming at any given price. Similarly, an increase in the capital stock also raises labor productivity and shifts the SRAS to the right.

Third, supply-side changes in the labor market: growth of the labor force or a reduction in structural unemployment raises the maximal feasible employment and hence full-employment output. This shifts the vertical segment of the AS curve to the right. Additionally, shocks to the prices of intermediate inputs (e.g., oil, imported raw materials) can be treated as factor-price disturbances that operate much like a change in the nominal wage. An oil-price hike raises the cost of production at given price and nominal wage, shifting the SRAS upward.

When an adverse supply shock shifts the AS curve upward (e.g., an exogenous spike in oil prices, a decline in productivity), the new intersection with the downward-sloping AD curve reveals a simultaneous increase in the equilibrium price level and a decrease in equilibrium output. This toxic combination of rising inflation and falling output—which implies rising unemployment via Okun's Law—is known as stagflation. The impact on employment is unambiguously negative: the rise in the price level partially offsets the adverse shock by lowering the real wage, but not enough to prevent a net decline in employment.

The analysis of AS shifts highlights a fundamental asymmetry in Keynesian macroeconomics. While demand management policies can effectively navigate AD shocks by stabilizing both output and prices, they face a severe policy dilemma when confronted with AS shocks. Policymakers must choose between accommodating the shock to preserve employment (sacrificing price stability) or contracting demand to fight inflation (sacrificing output and employment). If the central bank accommodates the shock by expanding the money supply (shifting AD to the right), it can stabilize output and employment, but at the cost of permanently higher prices and an upward shift in inflation expectations. This shifts the short-run AS curve up further, potentially triggering a wage-price spiral. If the central bank instead maintains a strict nominal anchor (holding AD fixed), the economy endures a painful recession with a negative output gap. Over time, the prolonged unemployment puts downward pressure on nominal wages and expected prices. As expected prices fall, the short-run AS curve shifts back down, eventually restoring output to its natural rate, albeit after a period of significant economic slack.

Modern New Keynesian models incorporate the expected price level explicitly. The short-run AS curve is formulated as output equals natural output plus alpha times the difference between actual and expected price level, plus a supply shock. An adverse supply shock shifts the short-run AS curve up. Therefore, mitigating the adverse effects of supply shocks often requires structural, supply-side policies—such as labor market reforms, investments in human capital, and technological subsidies—rather than mere aggregate demand manipulation.