Supply vs Demand Shock – An Economist Writes Daily

By the time most students leave undergraduate school, they become familiar with the aggregate supply-aggregate demand model. I think this pattern is so important that my Macro Principles class spends twice as much time on it as on any other topic. The pattern is nice because it uses the familiar supply and demand tools and gives them a macro twist. Below is a chart of the short-term AS-AD pattern.

Quick start: the AD curve increases to the right and decreases to the left. Both the Federal Reserve and the Federal Government can affect AD by increasing or decreasing total spending in the economy. Economists differ on the circumstances in which one authority is more relevant than another.

The AS curve reflects inflation expectations, short-term productivity (intercept) and nominal rigidity (slope). If inflation expectations increase, the AS curve moves vertically upwards. If there is a transient drop in productivity, it moves vertically up and left horizontally.

Nominal rigidity refers to the total expenditure elasticity of the quantity produced. In simple terms, nominal rigidity describes how output changes when there is a short-term increase in total expenditure. The figure above shows 3 possible SR-AS. AS0 reflects that firms will simply produce more when they spend more and do not raise prices. AS2 reflects that producers mainly increase prices and only slightly increase production. AS1 is an intermediate case. One of the things that determine nominal rigidity is the accuracy of inflation expectations. The more accurate the inflation expectations, the more vertical the SR-AS curve appears.*

The AS-AD model has many typical S&D features. The initial equilibrium is the intersection between the original AS and AD curves. There is a price and quantity implication when one of the curves moves. An increase in AD results in a combination of higher prices and greater output – depending on nominal rigidities. An increase in the SR-AS curve results in a combination of lower prices and higher output – depending on the slope of aggregate demand.

Of course, the real world is complicated – sometimes multiple shocks occur and multiple curves move simultaneously. If so, we can simply tell which curve has “moved more”. We should also expect that the long-term productive capacity of the economy has increased over the past two years, for example due to technological improvements, so that the new equilibrium production is several percentage points to the right. We cannot directly observe the AD and AS curves, but we can observe their results.

The big questions are:

  1. What happened during and after the 2020 recession?
  2. Was there more than one shock?
  3. When did the shocks occur?

Below is a chart of real consumption and consumer prices as a percentage of the February business cycle peak before the recession (see this post I made last week exploring only the real side). What can we say from this figure?

We can say that in the 12 months before the peak, production and prices rose steadily, in line with an accommodative monetary policy. The peak occurs at the origin. After that, we see a sharp and immediate drop in prices and real consumption. This is consistent with a mostly flat SR-AS curve and a sharp decline in the MA. 12 months after the peak, consumer spending had almost fully recovered. Lower production and higher prices imply that there has been a negative supply shock (less production despite a recovery in consumer spending). Fourteen months after the pre-recession peak, real consumption has grown at about its pre-recession rate. BUT, prices started to rise MUCH FASTER. In fact, the past few months have indicated that consumer spending is growing much faster than our productivity. The implication is that people – including those in business – are getting pretty good at predicting inflation. As a result, they adjust prices in the face of increased expenditure rather than increased production.

Consumption expenditure can also be split into several parts: Durable goods, Not durable, Services& Food. The charts below repeat the above exercise to reveal heterogeneous supply shocks and the composition of AD changes. What can we see?

Demand for durable goods fell, but then rebounded in four months. You can almost see the SR-AS until about 15 months after the peak. Then we start to see prices rise and output weaken – the telltale sign of higher inflation expectations.

Non-durable goods came under significantly different pressures. Immediately after the peak, quantities increased while prices fell. It appears there was an initial expansion in supply as companies attempted to unload inventory. Again, we can see the SR-AS curve rising slightly. Indeed, spending on non-durables spiked just ahead of April’s declines (are you one of those hoarding toilet paper?).

Total food expenditure was nearly flat in March 2020 and never looked back. It immediately rose 23%, then stabilized at about 15% above the pre-recession peak for the next 10 months. Thirteen months after the pre-recession peak, almost all of the increase in AD has been accompanied by rising inflation expectations and no change in output.

The outlier among all these expenditures is spending on services, which did not recover to the previous peak until 16 months later. It’s not that spending on services went down that much – it was typical. Rather, actual purchases never recovered to the previous peak. There seems to have been a one-time increase in the cost of producing services and the increase in demand seems to reveal that there has even been a decrease in long-term productivity. After all, what else could explain the rise in prices and more than a full recovery in other categories of real consumption?

In summary, it is clear to me that the recent “Covid recession” was largely an aggregate demand shock. Yes, food experienced higher actual production, but that may be explained by decrease in inventory and consumer substitution and storage. Otherwise, we are witnessing a story characterized by a sharp decline in the MA, followed by a strong recovery. Overall, there was no negative supply shock for goods. After about a year we see AD continuing to grow, but SR-AS falling comparably (left shift, rising inflation expectations) so there has been little real consumption growth compared to substantial price growth. Services are the only component of consumption to experience a long-term decline in productivity.

*Note that a perfect inflation expectation implies a vertical SR-AS curve. It can also be interpreted as any sloping SR-AS curve that moves instantaneously and perfectly so that it intersects the AD curve at the initial output level.

About Chris McCarter

Check Also

Business-level effects of the Russian power cut

Raphaël Lafrogne-Joussier, Andrei Levchenko, Julien Martin, Isabelle Méjean April 24, 2022 Editor’s note: This column …