Romer explicitly and directly ties the changes that have been made to the model to the historical shifts in the world's (and the United State's) economy; by noting that inflation was simply not a major issue for the first decades of the models' use and existence, the model's efficiency and accuracy during that period are easily explained. Just so are the inefficiencies of the model in a world where inflation is a major factor in the global economy (as well as in the economies of many independent nations), which the inclusion of aggregate supply (AS) into most contemporary uses of the model addresses. This also creates complications in the model that are not necessarily warranted by the increased accuracies that they provide, and it is for this reason that Romer advocates another major change to the model before it is put to further use. It is at his juncture that the tone of the article changes somewhat, and where Romer begins to get to the central point of his argument -- namely, that the IS-LM model is in drastic need of change and that he has developed the correct method for changing it. The author's language here becomes somewhat more dense than it is in his discussion of the model's historical progression and current applicability (or lack thereof), but he still manages to achieve a great deal of clarity and certitude in his explanation. Essentially, Romer argues, the model in its past iterations failed to account fully (or at al) for the influence of the central bank on real interest rates, which are themselves in essential part of the model.

Romer's model incorporates the effects of monetary policy into the calculations of interest rates and output that the model was originally designed to graph and establish points of equilibrium for, creating what he asserts is a far more accurate model for use in contemporary times. The second half of the paper is devoted to the evidence that author has compiled/created that backs up this assertion in remarkable detail and with a large degree of fairly irrefutable conclusions. Much of this evidence is developed through highly technical hypothetical situations, yet the model's performance as demonstrated by the author is also in keeping with relatively recent historical shifts in interest and output. Again, the depth of detail and clarity with which the author presents his argument and supporting facts allow for a huge amount of confidence on the part of the reader that the conclusions reached in this article are correct.
Conclusion

Despite the successful presentation of Romer's argument, the IS-LM model is still fairly ubiquitous in macroeconomics courses. Most sciences are inherently resistant to rapid change, and economics is often worse than other fields in this domain. With continued perseverance, however, the IS-MP-IA model could catch on with educators and theorists yet.