RDP 9407: Explaining Import Price Inflation: A Recent History of Second Stage Pass-through 7. A Measure of the Mark-up

Central to the empirical analysis in the preceding section is an assumption that the mark-up follows a stationary process. This assumption of stationarity permits us to take advantage of all known variables and estimated long-run relationships to construct a measure of short-run variations in the mark-up as deviations from their long-run equilibrium. Specifically, from equation (9) it follows that:

where is 0.66 and is 0.34, and the mark-up is stationary by construction.^[33]

The resultant series exhibits both significant variation and persistence as illustrated in Figure 7. The persistence is in keeping with the notion that price adjustment by firms is not costless; it can involve menu costs that reduce the frequency with which firms are prepared to adjust prices (Ball and Mankiw 1992). The oscillations are in keeping with shocks to the cost of imported or domestic inputs.

Oscillations in the mark-up occurred around a trend decline until the September quarter 1986, after which there was a sharp increase in the mark-up. In fact, by late 1988 the mark-up reached a record high in the period for which it has been estimated. Certainly, perceptions of a disproportionate rise in importers' mark-ups precipitated an inquiry by the Prices Surveillance Authority (PSA 1989).^[34]Subsequently, the mark-up has drifted down, but it remains at a level higher than has been evident for most of the period shown.

Given that the mark-up moves to offset the impact of changes in the cost of imported and domestic inputs, of interest is the behaviour of the mark-up during periods when major changes in such costs have occurred. For example, changes in the cost of the imported input – the largest component of total costs – have been primarily associated with episodes of exchange rate movement. Consequently, we shall focus on the behaviour of the mark-up during these episodes.

We expect that retailers increase their margins during times of appreciation, and decrease them during times of depreciation, so as to offset the changes in the free-on-board price that result from such currency movements. As shown in Figure 8, changes in the mark-up are, in general, positively correlated with changes in the exchange rate. However, at times, the strength of this correlation has been weakened. For instance, during the historic depreciation from the December quarter 1984 to the September quarter 1986, the mark-up fell by significantly less than did the exchange rate; in fact, the fall in the mark-up was not exceptional when compared with that in other episodes of depreciation. In the period of appreciation that immediately followed, the mark-up rose by more than the exchange rate. Indeed, in the late 1980s importers appear to have taken an opportunity to substantially increase their mark-up.

The mechanics of these changes in the mark-up are depicted in Figure 9 where episodes of major exchange rate movement are identified. During the sharp depreciations (from December 1984 to September 1986, late 1987 and recently) it is clear that over-the-docks import prices rose dramatically, driving up the total costs faced by local distributors. A significant share of this increase in costs was passed on to retail import prices. Consequently, despite our general finding of very slow pass-through, the mark-up was, on these occasions, squeezed by no more than in most other periods in which it has fallen.^[35]

Conversely, the appreciations (from September 1986 to September 1987 and from March 1988 to March 1989) led to a decrease in the total costs faced by retailers that was driven by a fall in the free-on-board price of imports. Retailers chose not to lower the retail prices, but instead increased their mark-up to a record high level.

A number of factors might have contributed to this especially large increase. Corresponding to a lengthy period of depreciation in the first half of the 1980s, there was a protracted fall in the mark-up. Given that we expect the mark-up to return to a long-run equilibrium, some reversal of the fall was inevitable as retailers sought to “catch up” on accumulated losses. However, the mark-up appears to have increased beyond such “catching up”. This outcome raises several important questions about its recent behaviour.

A popular view is that, in the short run, given downward stickiness in final retail prices, the mark-up responds asymmetrically to changes in the exchange rate; that is, the increase in the mark-up during appreciations is greater than the fall in the mark-up during depreciations (PSA 1989). The second stage results reported in the previous section lent support to this view by demonstrating asymmetry in the response of changes in the retail import price to changes in their landed price during periods of appreciation and depreciation.

However, an important aspect of the asymmetry result is that exchange rate depreciations have often been precipitous, whilst appreciations have been less steep. Given that price adjustment for firms is not costless, they tend to be more responsive to large shocks than to small ones (Ball and Mankiw 1992). For example, the historic depreciation of the mid 1980s may have forced firms out of their “band of inaction” so that pass-through of total costs was higher than would otherwise have been the case, containing the fall in the mark-up. Alternatively, during the subsequent and less dramatic appreciation, firms may have remained within their band of inaction and failed to pass on the reduction in total costs that stemmed from such appreciation, permitting a disproportionate rise in the mark-up.

Nonetheless, it is by no means clear that the extent of the rise in the mark-up during the late 1980s is attributable solely to an asymmetric response to exchange rate movement. One might expect a role for domestic demand conditions. During the late 1980s, actual output exceeded potential creating an opportunity for firms to increase their mark-up with little fear of loss of customers (Benabou 1992). Certainly, the PSA (1989) concluded that the appreciation, combined with buoyant economic conditions, allowed firms to recover lost margins. In our estimation of second stage pass-through, though, the output gap proved to be insignificant. This is not to say that domestic demand conditions are irrelevant to the mark-up. Rather, in our model, there is no statistical indication that the output gap is important. (The information afforded by it may well be swamped by that contained in more volatile exchange rate movements.)

At issue, however, is the future direction of the mark-up, and the attendant pressure on inflation, as Australia's economic conditions change. Given prospective currency appreciation and consolidation of economic recovery, it is of interest whether a situation will re-emerge that is reminiscent of the late 1980s. In this regard, there are several important differences between the present environment and that which prevailed in the late 1980s.

The first of these differences is the prevailing level of the mark-up. As importers emerged from the mid-1980s depreciation, there was a clear motivation to recover lost margins. In contrast, as importers emerge from the recent episode of depreciation, the motivation for “catch up” is less evident. This is because the mark-up was established at such a high base in the late 1980s.

The second difference in the present environment compared with that in the late 1980s stems from the process of microeconomic reform which has now reached greater maturity. Consequently, there is an increased focus on competition. We might expect that importers will be more inclined to preserve viable margins by reducing cost rather than increasing final prices.^[36]

The third difference, which is relevant in the case of depreciation, is the prevailing rate of inflation. Ultimately, the extent to which retailers can alter their mark-ups is governed by the scope for changing the relative prices of final goods. When the prices of domestically produced goods are increasing rapidly, as was the case in the second half of the 1980s, the scope for raising retail import prices is considerable. If, however, domestic inflation is low, relative price changes are conspicuous. Certainly, if the price of a domestic substitute has not changed, the scope for altering the corresponding retail import prices is limited. Thus, a continuance of Australia's low inflation can impose an exigency in this regard.

Footnotes

In fact, λ = r −0.66 pl −0.34c is simply the error-correction term from the cointegration analysis in the previous section. [33]

Public perceptions that the sharp appreciation of the late 1980s, and the associated fall in free-on-board import prices, was not being passed on to consumers, resulted in the PSA inquiry. [34]

An outcome facilitated by the subdued movement in the cost of domestic inputs during these episodes. [35]

Such an effect would, however, be difficult to separate from the discipline imposed by low domestic inflation on importers seeking to alter the price of imports relative to substitute goods. [36]