The IS-LM model has served as a fundamental building block to short-run macroeconomic theory since its conception, through its simple interpretation of the concepts of Keynes’ General Theory, notably the joint description of goods and financial markets. The framework demonstrates the relationship between output and interest rates as a function of equilibrium in the two markets. As such it acts as a basic model of economic fluctuations for both pedagogic and descriptive/prescriptive purposes, and forms the groundwork for medium-run AS-AD analysis.

Despite its contribution to the emergence of a broad consensus on macroeconomic policy (Vercelli, 1999), the IS-LM model has received substantial criticism, particularly regarding its relevance to modern economies. In his article Keynesian Macroeconomics without the LM curve, Romer suggests that it is limited by its basic assumptions and subsequently suggests a new model. Most importantly, his IS-MP-IA model proposes an overhaul of the IS-LM’s outdated monetary targeting presumption in favour of inflation targeting.

For the purposes of consistency and clarity it revises the original model’s use of real and nominal interest rates, and develops a basis for relating output to inflation instead of the price-level for AS-AD modeling in the medium-run. This review will focus on evaluating the validity of these major alterations as well as Romer’s overriding judgment that the IS-LM model is no longer valid in its original form.

The LM curve plots the value of the interest rate associated with any value of income for a given money stock and price level (Blanchard & Sheen, 2009), its basis being the relation between money demand and central bank determined money supply. However, this assumption of monetary targeting has become increasingly irrelevant as central banks have largely shifted to an inflation targeting objective in many industrialised countries including the US, UK, Australia, Canada, Sweden.

Accordingly, Romer’s model acquires the more realistic assumption of central banks following interest rate rules. This is achieved through the substitution of the LM curve with a horizontal monetary policy (MP) curve that reflects the central bank’s choice of the real interest rate, given an inflation rate. It is implicit that increased inflation will necessitate increased real interest rates, causing an upward shift of the MP and thus a decrease in output.

This makes clear, intuitive sense and is perhaps more straightforward than the LM approach since it provides a direct view of the central bank’s behaviour instead of one implied by interactions in the money market. Romer’s MP curve is specified as a real interest rate rule despite most central bank’s use of a nominal rate as their short-run instrument. Whilst a nominal rate rule is applicable for the very short-run, in deciding whether to change their nominal rate central banks necessarily consider expected inflation.

Since the nominal interest rate is the sum of the real rate and expected inflation (Fischer Equation), they effectively set the real interest rate. As such, Romer’s model addresses one of the major failings of IS-LM: that it often ignores inflation expectations and thus fails to distinguish between real and nominal interest rates (Sumner, 2004). Since the real rate is relevant to the IS curve and the nominal rate to the LM curve in traditional IS-LM models, the IS-MP also improves the model’s consistency and coherence, which are particularly important in its pedagogic role.

When extending to medium-run AS-AD analysis, the IS-LM is viewed as a “subsidiary model,” (Colander, 2004) and representation of the aggregate demand side of the economy (Vercelli, 1999) that is expressed in terms of output and the price level. In contrast, Romer’s approach facilitates a relation between output and inflation, the latter involving a sustained increase in the price level. The AD curve is retained but becomes a derivation of the IS and MP curves, whilst the AS curve is replaced with a horizontal inflation adjustment (IA) curve.

This relies on the assumption that inflation is given and in the absence of shocks, rises when output exceeds its natural rate – this causes a downward shift of IA as inflation falls until a stable equilibrium is reached. The opposite is true for below-natural output. By relating output directly to inflation rather than the price level, Romer’s IS-MP-IA offers a more straightforward analysis of two of the most examined economic variables. Unlike IS-LM-AS-AD it allows the inference of the magnitude of price changes as well as the sign. It is also more easily interpretable in that it portrays higher inflation and ower inflation in boom and recession periods respectively, as opposed to the inflation and deflation depicted by the IS-LM model. Although it is widely appreciated that the IS-LM is a core element of macroeconomic theory, as asserted by Blinder (1997), Vercelli (1999), and Revier (2000) it is evident that it requires various modifications. Romer provides an extremely convincing argument for his model as a response to the increasingly apparent irrelevancies of the IS-LM. His IS-MP-IA constitutes a much-needed update that addresses some of the IS-LM’s fundamental inconsistencies whilst retaining a clear, intuitive approach.

The result is a readily understandable model for teaching that has the potential to be extended for professional descriptive and prescriptive purposes. However, it has not been readily adopted into university textbooks (Weerapana, 2003) and thus some further extensions may be beneficial. A minor adjustment to improve clarity would be the renaming of Romer’s AD curve (e. g. the output adjustment, OA curve) to facilitate clearer distinction between the models. Further, Romer’s use of a single interest rate may be a misrepresentative (Bordo & Schwartz, 2004) over-simplification of policy decisions and consumer/investor behaviour.

Its lack of microeconomic foundations regarding the constraints inherent in consumption/investment decisions may also detract from its applicability. Indeed Colander (2004) asserts that the field of macroeconomics is evolving into one centered around complex, non-linear relationships and interrelationships between variables. As such the model’s simplicity may become a vice rather than a virtue. So, while it is currently clearly valuable for teaching concepts and an understanding of basic economic relationships, the future of any IS-LM-type model remains uncertain.