Orhun Sevinç is an Economist at the CBRT.
For views, suggestions
The question of the role of exports in price formation and price volatility in unprocessed food products in Turkey has been on the agenda for a long time. However, it is not easy to find the right answer. There are two main camps in this ongoing debate. One camp claims that the rise in exports pushes prices up by reducing the amount of products supplied to the domestic market. The other camp argues that exports do not have a potential to continuously affect domestic prices. This group asserts that for many products, the exported amount is only a small portion of the domestic consumption and the products exported are not a substitute for the domestically consumed ones. It is not possible to find the right answer by looking at the data either, as product-based production and consumption data with a monthly frequency are not available. However, available data highlight two issues. First, there is a visible positive correlation between exports and domestic prices. Secondly, this correlation is still observed (albeit partially) even if the domestic product supply does not change significantly. Therefore, it would be helpful to understand which mechanisms come into play in price formation in addition to the quantitative change in product supply.
In this study, we take a close look at a relatively lesser-debated issue of changes in product quality and examine whether quality management can be an important tool in reducing price volatility. The analysis focuses exclusively on this question: how and through which mechanism are the prices affected when only the quality composition is changed provided that the quantity and the unit value of exports remain the same? Let us suppose that we can carry out an experiment. Let us assume that instead of the low-quality product that we have been exporting for some time, we started exporting the same amount of a high-quality product that finds buyers at the same price abroad. What happens to domestic prices in this case? Considering the usual arguments and assuming that the changes in quality are reflected in calculations in official statistics data, one should not expect a significant change in prices as the quantity is the same. However, our findings point to something different.
How influential is the quality composition of exports on monthly price changes?
The analysis focuses on tomatoes and cucumbers, which are both available all year around, are widely consumed both domestically and abroad, and have similar consumption patterns. These products are also items that create considerable volatility in consumer prices. Our aim is to estimate how domestic price changes are affected by export quality composition. We can determine the quality composition with a measure we call quality export ratio. The quality export ratio can be simply defined as the ratio of the quantity of products sold to countries with a higher welfare level than that of Turkey to the total exports of that product. The goal of focusing on wealthier countries is to include those with higher purchasing power which demand high-quality products and strictly apply quality standards in consumption. Thus, we can identify the markets where the best quality products of our domestic market go.
The regression model tries to explain the monthly changes in regional prices at NUTS2 level via the quality export ratio, the monthly change in the total export quantity of the product, and the monthly change in the total export unit value of the product. The model controls for fixed effects such as region, calendar month, product type and time as well as the interactions of these fixed effects. In less technical terms, the relationship between variables is adjusted for regional, seasonal, product-specific effects (and their various combinations) and regional time effects common to both types of products. Meanwhile, since the changes in total exports and the unit value of exports also affect prices, these variables are also controlled. In short, we are trying to get as close as possible to the experimental environment described in the second paragraph.
To summarize the composition effect estimated based on this model, an increase in the quality export ratio from zero to 100 percent within a month pushes up the product prices by approximately 30 percent on average. This estimate is both statistically and economically significant. Charts 1 and 2 can be regarded as rough visualizations of the analysis result. Average price changes and the quality export ratios both adjusted for the effects discussed above follow an evidently common course. This estimate, which is based on the assumption that our total exports and the unit value of exports will remain the same, reveals that the quality composition of exports can have a serious effect on price volatility. Considering the fresh fruit-vegetable group and the magnitude of this effect, we can say that the changes in the quality composition are also a factor of price volatility. In other words, in terms of price formation, what and to whom we sell are as important as how much we sell. Moreover, since the changes in the quality composition of exports also affect in a sense the quality structure of domestic demand, this result brings to the forefront the issue of quality management of market products in policy design as well.
Through which mechanisms does the quality composition of exports affect prices?
This result may be “confusing” at first glance according to the existing supply and demand-oriented point of view. However, the plausible mechanism better manifests itself when, for instance, we consider exports to EU countries, which constitute the majority of the countries with a higher level of welfare than Turkey. The main difference between Turkey’s and the EU’s approaches to these products is that the EU strictly applies a high quality threshold in general and specific marketing standards. In Turkey, on the other hand, the domestic market allows for a very wide range of quality with low predictability. When we keep other variables unchanged and only increase our exports to countries wealthier than Turkey, we will be selling a higher amount of the high-quality portion of our production. It may be thought at first that the average price of remaining products will also fall as the average quality in the domestic market will decline as a result of these additional exports. However, this is not very likely since the price indices typically reflect the prices adjusted for quality.
The changes in quality may affect prices through two mechanisms, independent of supply and demand. The first one is the acceleration in consumers’ motivation to search for the product when they know that a high-quality product is on the market. The behavioral phenomenon triggering this mechanism is that you do not quickly agree to buy an inferior product when you know there are high-quality products on the market, and so you go and see more sellers. In this way, even if the supply of product is constant, accessibility of the high-quality product due to the strong motivation to search creates a downward pressure on prices. The opposite is also true. The propensity to immediately buy any product when the best products are not on the market increases the prices. For this reason, the weakened consumer motivation to search for the good products in times of increased exports of high-quality products will lead to an upward impact on prices (contrary to the expectations), when the product quality is given.
In the second mechanism, the key factor is that lack of implementation of the quality standard, and the high and unpredictable volatility in quality make it hard for consumers to assess quality. In this case, the utility of a product for the consumer is identified via a comparison of that product’s utility with the utility of the best product on the market. Accordingly, prices in a product’s quality scale are expected to be ordered, ceteris paribus, according to quality. When the highest quality product on the market is “mediocre”, the value of even the worst product will not be low in the eyes of consumers. On the other hand, when there are good products on the market even in small amounts, the prices of all other products are revised downwards. Furthermore, this situation can be observed even in a perfectly competitive market. Again, this situation may cause the upper quality limit to become a factor of price volatility, independent of the production level.
To conclude, our observations bring to the forefront two new and important issues in the management of food price volatility. The first one is to focus particularly on countries in the high-income group and on the quality composition of products sold to these countries, in view of the export-oriented risks to price stability. According to the analysis in this study, an increase in the domestic supply of high-quality products can support stability in price formation. The second issue addresses the regulation of quality standards in the domestic market: improving the quality standards and strictly supervising the implementation of these standards can significantly contribute to the predictability of food prices by regulating the quality structure in the market.
 The experimental environment can ideally be achieved by including data on monthly regional production and consumption for each product. The absence of such data is compensated by the use of region and time-specific fixed effects. The results in this study are interpreted under the assumption that product-specific domestic supply and demand shocks are not too different for tomato and cucumber at a monthly frequency.
 In practice, quality adjustments may not be effectively reflected in the calculations even in the medium term. In this case, the average consumer prices are also expected to decline when the average quality in the domestic market decreases. Therefore, the estimated effect here can be regarded as a lower bound of the true effect. The failure to effectively reflect the changes in product quality to price indices is a much debated and well-known statistical problem around the world. For further details and review of literature on the subject, please see Boskin, Dulberger and Gordon (1998).
Boskin, M. J., E. R. Dulberger and R. J. Gordon (1998). “Toward a More Accurate Measure of the Cost of Living: Final Report to the Senate Finance Committee from the Advisory Committee to Study the Consumer Price Index.” Editor: D. Baker. Getting Prices Right: The Debate over the Consumer Price Index (Armonk, NY: M. E. Sharpe).