The Turkish demonstrations - and the usefulness of the AS/AD framework
Peter Dorman has a blog post that have gotten quite a bit of attention in the blogosphere on the AS-AD model and why he thinks it is not a useful framework. This is Peter:
"Introductory textbooks are supposed to give you simplified versions of the models that professionals use in their own work. The blogosphere is a realm where people from a range of backgrounds discuss current issues often using simplified concepts so everyone can be on the same page. But while the dominant framework used in introductory macro textbooks is aggregate supply—aggregate demand (AS-AD), it is almost never mentioned in the econ blogs. My guess is that anyone who tried to make an argument about current macropolicy using an AS-AD diagram would just invite snickers. This is not true on the micro side, where it’s perfectly normal to make an argument with a standard issue, partial equilibrium supply and demand diagram. What’s going on here?"I am somewhat surprised by Peter's statement that the AS-AD framework is never mentioned on the econ blogs. That could indicate that Peter has never read my blog (no offense taken - I never read Peter's blog before either). My regular readers would of course know that I am quite fund of using the AS-AD framework to illustrate my arguments. Other Market Monetarists - particularly Nick Rowe and Scott Sumner are doing the same thing quite regularly. See Nick's discussion of Peter's post here. The purpose of this post, however, is not really to discuss Peter's critique of the AS-AD framework, but rather to show the usefulness of the framework with a example from today's financial news flow. Furthermore, I will do a Market Monetarist 'spin' on the AS-AD framework. Hence, I will stress the importance of monetary policy rules and what financial markets tell us about AD and AS shocks. The Istanbul demonstrations as an AS shock Over the weekend we have seen large street protests in Istanbul in Turkey. The demonstrations are the largest demonstrations ever against the ruling AKP party and Prime Minister Erdogan. Mr. Erdogan has been in power for a decade. The demonstrations today triggered a 10% drop in the Istanbul stock exchange so there is no doubt that investors think that these demonstrations and the political ramifications of the demonstrations will have a profound negative economic impact. I believe a core insight of Market Monetarist thinking is that financial markets are very useful indicators about monetary policy shocks. Hence, we for example argue that if the US dollar is depreciating, market inflation expectations are rising and the US stock market is rallying then it is a very clear indication that US monetary conditions are getting easier. While the focus of Market Monetarists have not been as much on supply shocks as on monetary policy shocks (AD shocks) it is equal possible to 'deduct' AS shocks from financial markets. I believe that today's market action in the Turkish markets are a pretty good indication of exactly that - the combination of lower stock prices, higher bond yields (higher risk premium and/or higher inflation expectations) and a weaker lira tells the story that investors see the demonstrations as a negative supply shock. It is less clear whether the shock is a long-term or a short-term shock. A short-run AS shock - mostly about disruption of production My preferred textbook version of the AS-AD model is a model similar to the one Tyler Cowen and Alex Tabarrok use in their great textbook Modern Principles of Economics where the model is expressed in growth rates (real GDP growth and inflation) rather than in levels. It is obvious that the demonstrations and unrest in Istanbul are likely to lead to disruptions in production - roads are closed down, damages to infrastructure, some workers are not coming to work and even some lines of communication might be negatively impacted by the unrest. Compared to the entire Turkish economy the impact these effects is likely quite small, but it is nonetheless a negative supply shock. These shocks are, however, also likely to be temporary - short-term - rather than permanent. Hence, this is a short-run AS shock. I have illustrated that in the graph below. This is the well-known illustration of a negative short-run supply shock - inflation increases (from P to P') and real GDP growth declines (from Y to Y'). This might very well be what we will see in Turkey in the very short run - even though I believe these effects are likely to be quite small in size. However, note that we here assume a "constant" AD curve. Peter Dorman is rightly critical about this assumption in his blog post: "...try the AD assumption that, even as the price level and real output in the economy go up or down, the money supply remains fixed." Peter is of course right that the implicit assumption is that the money supply (or money base) is constant. The standard IS/LM model suffers from the same problem. A fact that have led me to suggest an alternative ISLM model - the so-called IS/LM+ model in which the central bank's monetary policy rule is taken into account. Obviously no analysis of macroeconomic shocks should ignore the monetary policy reaction to different shocks. This is obviously something Market Monetarists have stressed again and again when it for example comes to fiscal shocks like the 'fiscal cliff' in the US. In the case of Turkey we should therefore take into account that the Turkish central bank (TCMB) officially is targeting 5% inflation. Therefore if the TCMB was targeting headline inflation in a very rigid way (ECB style) it would have to react to the increase in inflation by tightening monetary policy (reducing the money base/supply) until inflation was back at the target. In the graph above that would mean that the AD curve would shift to the left until inflation would have been brought down back to 5%. The result obviously would be a further drop in real GDP growth. In reality I believe that the TCMB would be very unlikely to react to such a short run supply. In fact the TCMB has recently cut interest rates despite the fact that inflation continue to run slightly above the inflation target of 5%. Numbers released today show Turkish headline inflation was at 6.6% in May. In fact I believe that one with some justification can think of Turkish monetary policy as a "flexible NGDP growth targeting" (with horrible communication) where the TCMB effectively is targeting around 10% yearly NGDP growth. Interestingly enough in the Cowen-Tabarrok version of the AS-AD model that would mean that the TCMB would effectively keep the AD curve "unchanged" as the AD curve in reality is based in the equation of exchange (MV=PY). The demonstrations could reduce long-term growth - its all about 'regime uncertainty' While the Istanbul demonstrations clearly can be seen as a short-run supply shock that is probably not what the markets are really reacting to. Instead it is much more likely the the markets are reacting to fears that the ultimate outcome of the demonstrations could lead to lower long-run real GDP growth. Cowen and Tabarrok basically think of long-run growth within a Solow growth model. Hence, there are overall three drivers of growth in the long run - labour forces growth, an increase in the capital stock and higher total factor productivity (TFP - think of that as "knowledge"/technology). I believe that the most relevant channel for affecting long-run growth in the case of the demonstrations is the impact on investments in Turkey which likely will influence both the size of the capital stock and TFP negatively. Broadly speaking I think Robert Higgs concept of "Regime Uncertainty" comes in handy here. This is Higgs:
“The hypothesis is a variant of an old idea: the willingness of businesspeople to invest requires a sufficiently healthy state of “business confidence,” … To narrow the concept of business confidence, I adopt the interpretation that businesspeople may be more or less “uncertain about the regime,” by which I mean, distressed that investors’ private property rights in their capital and the income it yields will be attenuated further by government action. Such attenuations can arise from many sources, ranging from simple tax-rate increases to the imposition of new kinds of taxes to outright confiscation of private property. Many intermediate threats can arise from various sorts of regulation, for instance, of securities markets, labor markets, and product markets. In any event, the security of private property rights rests not so much on the letter of the law as on the character of the government that enforces, or threatens, presumptive rights.”I think this is pretty telling about the fears that investors might have about the situation in Turkey. It might be that the Turkish government is not loved by investors, but investors are clearly uncertain about what would follow if the demonstrations led to "regime change" in Turkey and a new government. Furthermore, the main opposition party - the CHP - is hardly seen as reformist. And even if the AKP does remain in power the increased public disconnect could lead to 'disruptions of production' (broadly speaking) again and again in the future. Furthermore, the government hard-handed reaction to the demonstrations might also "complicate" Turkey's relationship to both the EU and the US - something which likely also will weigh on foreign direct investments into Turkey. Hence, regime uncertainty therefore is likely to reduce the long-run growth in the Turkish economy. Obviously it is hard to estimate the scale such effects, but at least judging from the sharp drop in the Turkish stock market today the negative long-run supply shock is sizable. I have illustrated such a negative long-run supply shock in the graph below. The result is the same as in the short-run model - a negative supply shock reduces real GDP growth and increases inflation. However, I would stress that the TCMB would likely not in the long run accept a permanent higher rate of inflation and as a result the TCMB therefore sooner or later would have to tighten monetary policy to push down inflation (by shifting the AD curve to the left). This also illustrates that the demonstrations is likely to become a headache for the TCMB management. Is the 'tourism multiplier' zero? Above I have primarily described the Istanbul demonstrations as a negative supply shock. However, some might argue that this is also going to lead to a negative demand shock. Hence, Turkey very year has millions of tourists coming to the country and some them will likely stay away this year as a consequence of the unrest. In the Cowen- Tabarrok formulation of the AD curve a negative shock to tourism would effectively be a negative shock to money velocity. This obviously would shift the AD curve to the left - as illustrated in the graph below. Hence, we initially get a drop in both inflation and real GDP growth as the AD curve shifts left. However, we should never forget to think about the central bank's reaction to a negative AD shock. Hence, whether the TCMB is targeting inflation or some kind of NGDP growth target it would "automatically" react to the drop in aggregate demand by easing monetary policy. In the case the TCMB is targeting inflation it would ease monetary policy until the AD curve has shifted back and the inflation rate is back at the inflation target. This effectively means that a negative shock to Turkish tourism should not be expected to have an negative impact on aggregate demand in Turkey for long. This effectively is a variation of the Sumner Critique - this time, however, it is not the budget multiplier, which is zero, but rather the 'tourism multiplier'. Hence, from a macroeconomic perspective the demonstrations are unlikely to have any major negative impact on aggregate demand as we would expect the TCMB to offset any such negative shock by easing monetary policy. That, however, does not mean that a negative shock to tourism will not impact the Turkish financial markets. Hence, there will be a change in the composition of aggregate demand - less tourism "exports" and more domestic demand. This likely is bad news for the Turkish lira. Mission accomplished - we can use the AS-AD framework to analysis the 'real world' I hope that my discussion above have demonstrated that the AS-AD framework can be a very useful tool when analyzing real world problems - such as the present public unrest in Turkey. Obviously Peter Dorman is right - we should know the limitations of the AS-AD model and we should particularly be aware what kind of monetary policy reaction there will be to different shocks. But if we take that into account I believe the textbook (the Cowen-Tabarrok textbook) version of the AS-AD model is a quite useful tool. In fact it is a tool that I use every single day both when I produce research in my day-job or talk to clients about such 'events' as the Turkish demonstrations. Furthermore, I would add that I could have done the same kind of analysis in a DSGE framework, but I doubt that my readers would have enjoyed looking at a lot of equations and a DSGE model would likely have reveal little more about the real world than the version of the AS-AD model I have presented above. PS Please also take a look at this paper in which I discuss the politics and economics of the present Turkish crisis. PPS Paul Krugman, Nick Rowe and Mark Thoma also comment on the usefulness of the AS-AD framework.