Rising prices in a deflationary environment
SUMMARY Regarding the ongoing extremely loose monetary policy in the United States, Japan and Europe, some observers worry about inherent inflationary dangers. Those worries are ridiculed by others who point to the lasting tendency of declining prices everywhere as reflected in major price indices. Which side is right is difficult to decide:
– Even in a deflationary environment there are constantly some prices rising, while others fall.
– The current state of the economy and our economic knowledge does not allow deciding at an early stage, which would enable monetary policy to act adequately, whether the general level of prices has begun to rise persistently.
– Official price indices are poor early-warning signs. In simply adding up weighted price changes of goods and services for special groups of economic actors they fail to capture the underlying price dynamics in an economy. Hidden price changes and people’s reactions to rising prices further complicate the diagnosis.
– The calculation of the German consumer price index may serve to illustrate the way in which official statistics account for a changing environment as well as its shortcomings.
No sign of inflation anywhere, as Izabella Kaminska and other authors are claiming these days? Although being aware that in current debates observers focus on the looming danger of deflation (see, for example, the contributions quoted by Noah Smith), I cannot help noticing that in the region where I live many prices are rising – and rising markedly. Friends and neighbours are complaining about ever higher cost of living, my weekly grocery list has perceptibly become more expensive and people planning a bigger replacement are often irritated remembering a much lower price – not to mention the development of rents and real estate prices.
Of course, I and my immediate surroundings are no representative sample. In November, the German Federal Statistical Office (Destatis) wrote that
“consumer prices in Germany rose by 1.2% in October 2013 compared with October 2012. The inflation rate as measured by the consumer price index thus decreased again [my emphasis] (September 2013: +1.4%). The last time a lower rate of inflation was observed was in August 2010 (+1.0%). Compared with September 2013, the consumer price index was down 0.2% in October 2013.”
So, do we err? Not necessarily. In every economy, prices are moving constantly for many reasons. Shifts in demand and supply play a role, but also changing cost of production, exchange-rate changes, taxes and other factors.
In the following chart Destatis has listed some items which illustrate the point. While in October 2013 in Germany prices for potatoes and butter, for example, had strongly increased, services of physicians and credit institutions as well as items such as tomatoes, cucumbers and liquid gas had become cheaper.
Destatis is publishing this list of “noticeable” price changes every month. Here is the November chart:
The comparison illustrates that there is a lot of short-term movement in prices of individual groups. The composition is changing from month to month as is the range of changes. In October, increases ranged from 16 to 29 percent, in November from 15 to 27 percent. Falling prices ranged from 9 to 19 percent in October and from 9 to 33 percent in November. Apparently, in some sectors prices are strongly rising, but they do not manifest in the overall index as they are (over-) compensated by falling prices elsewhere.
Wholesale prices confirm this impression. According to Destatis “the index of selling prices in wholesale trade was down 2.2% in November 2013 on November 2012”. But a closer look reveals that wholesale prices for fruits, vegetables and potatoes had risen by 6.4%, milk, milk products, eggs, cooking oils and dietary fats even by 7.6%. A key driver of the overall decline had been oil – an effect Frances Coppola, for instance, recently proposed to omit to get a more accurate picture of UK core inflation.
Like all aggregates, indices usually do not tell the whole story. Not only that some prices rise as others fall. Official price statistics capture only a fraction of goods and services in an economy, prices enter official statistics weighted (which means that they may matter less for aggregate changes than people think) and the weighting patterns of expenditures are revised infrequently. All this may explain the deviation of individual perceptions from official results.
The question is, what matters more for an economy in a particular circumstance, and what is more likely to propagate and to contribute to some overall tendency, the rising or the falling prices? Do potatoes and butter matter more for German consumers, and are more likely to trigger reactions such as wage claims and rising prices elsewhere, than services of physicians and credit institutions? Does olive oil matter more than heating oil in this respect? And when do the dynamics of rising prices cross the line and become inflation? In the following, I will argue that this is a question about processes to which official price statistics, and empirical economic studies based on them, have little to contribute.
Don’t get me wrong. This is not about the H-word. Here, I refer to the German example, but I think my argument holds more generally, too. It says that the current state of both the economy and our economic knowledge simply does not allow to decide whether the general level of prices has begun to rise, and is rising persistently (which is the definition of inflation), and when the data eventually give a clearer picture it may be too late for monetary policy to react adequately. Market reactions to the US Federal Reserve’s first timid signaling of its intention not to take back but just to “taper” or end its monthly purchases of long-term assets known as Quantitative Easing (QE) indicate that the scope of central banks in this respect may be smaller than assumed.
Even if measured correctly and adjusted timely, price indices would be poor early-warning signs of inflation. Price dynamics are complex. In order to come to conclusions about the state of the underlying processes it is of not much use to simply add up increases of prices for individual groups such as households or producers weighted by their respective proportion of total expenditure. Official price indices are useful instruments of economic analysis but in the current debate their importance often seems highly overrated.
Prices in an economy are constantly influencing one another across sectors and markets. They enter production processes as cost factors, which are key determinants of producer prices, and – usually with a mark-up – end up in consumer prices. They affect the demand for goods and services whose rise or fall in turn necessarily influences demand and prices elsewhere as long as budgets are kept unchanged. Price increases arouse needs and desires which manifest in higher wages – another cost factor in production and a crucial determinant of household spending and related price changes.
Furthermore and widely neglected, prices influence the choices of economic agents between both goods AND financial assets linking consumer and investor decisions in real and financial markets in ways no economic model can capture adequately at our current state of knowledge. If prices of both goods and money decline, if loans become cheaper and the interest rates savers get become less attractive, people tend to buy more things but also more bonds and equities whose rising prices offer higher returns. If then the price process is reversed, the effects again may be felt in both real and financial markets.
One important aspect of price dynamics is time scales. Some prices change more frequently than others. At one end of the spectrum is the variability of prices of exchange-traded commodities and financial instruments. At the other are prices of real investments and projects bound by very long-term contracts.
Short- and long-term influences constantly interact with the result that at times they compensate or weaken one another and at times reinforce and accelerate price movements. In principle, in this environment no item and no price change is too small or unimportant to exert a lasting effect in the interplay with others. The result may or may not enter official price statistics – and if it does, it may be difficult to interpret:
One characteristic of complex dynamics is that the relation between cause and effect is not always apparent. There are both strong changes without visible cause and dramatic events whose effects fizzle out without noticeable influence:
Take the introduction of the euro and the immediate price rises which followed. Those had all ingredients for a strong inflationary impulse. Immediately after currency conversion, many products and services became more expensive. This was most obvious for services such as hairdressers, restaurants, dry cleaners and cinemas which had to print new price lists and grasped the opportunity. Euro –Teuro (“teuer” is the German word for expensive and “Teuro” rhymes with the German pronunciation of Euro) was a widespread slogan in those first years. The explanations found for the actual moderate development that followed, in my view, were never wholly convincing.
The counterexample is the first oil crisis in 1973 and the resulting widespread stagflation. This came at a time when the globalization and sophistication of financial flows – one of the key determinants explaining why the transmission mechanism of national monetary policies these days is disrupted – still played a far lesser role. But the extent of price reactions took observers by surprise and there is still no unanimity on the role of monetary policy.
Shocks and gradual changes
The diagnosis of the beginnings of inflation is complicated by the variety of sources of rising prices, and by the different ways in which people react to them, which, as we will see shortly, are not always captured properly in official price statistics.
Studies of the impact of price increases on an economy usually tend to focus on “external shocks” such as changes of oil prices which cannot be influenced by national policies. But, in daily life we experience other, less dramatic sources of price rises as well which contribute to the overall dynamics.
One is upgrading. Often products and services bought today are rarely the same as those a year ago. The new smartphone has a more advanced technology which may or may not justify its higher price. The garage is offering a new combination of new and old services at the next check of the car which makes it a little more expensive than before.
The problem with upgrading is that usually old and new prices are not comparable. The old product and its price may be no longer available and the new one is adding to the overall cost of living – contributing to the process of price formation in the economy by triggering a decline in demand and prices elsewhere and/or rising wage claims.
The opposite source of hidden price rises is downgrading. There are changes in quality which make products cheaper to produce but their prices remain unchanged. New items may be a little simpler or a little shabbier than older ones. Packages may be smaller or have a little less content. Again, the resulting price difference may not be captured by price statistics. But people have to pay it, and react to it, causing adjustments of volumes and prices elsewhere in the economy.
Reduced spending and substitution
On the “price taker” side, people try to mitigate the effects of rising prices, or circumvent them, by reducing spending. This works best for goods and services with a high price elasticity of demand. One example is luxury goods: It is easier for people to forego a sailing boat than a car needed for daily commuting. Examples for a price inelastic demand are cigarettes and other drugs where a rising price usually triggers a small or no reaction. Reduced spending tends to lower price pressures and may even result in declining prices which would betray the initial inflationary impression. Nevertheless, prices did rise and the damage is done. Desires are awakened to find means and ways to regain the former living standard which may lead to new inflationary impulses through higher wages and earnings and demand and supply adjustments elsewhere.
Reduced spending may also affect complementary goods. If people can no longer afford theatre and restaurant visits their demand for taxi rides and fancy clothes may decline as well even if initially their prices remained unchanged. As a consequence these items may become cheaper. If this shows up in price statistics the resulting impression, again, may be misleading as the initial source of this “deflationary” effect, too, was a price increase.
A similar argument holds for price effects of substitution. Switching to cheaper alternatives as a reaction to rising prices works best for goods which can be easily replaced by others. Examples are shifts of demand to alternatives to oil as a source of energy with rising oil prices or to cheaper meals with higher food prices. Again, if prices rise less or even fall again as a result of the reduced demand the picture may be misleading if the initial price increases have already begun to raise production cost or lower living standards, and adjustment processes throughout the economy with inflationary consequences have set in.
CPI adaption and updates – the German example
Official statistics try to account for these phenomena and other changes in the economy as the example of the construction of the German CPI may illustrate. Here, a distinction must be made between the weighting pattern of expenditures and the basket of goods and services included. The weighting pattern is kept unchanged over five years. The basket is updated constantly.
For the weighting pattern, every five years, approximately 60,000 households are asked to record their income and expenditures on approximately 600 individual types of goods (“from coffee pods to cinema tickets”) for a few months. From these records, a rough “expenditures profile” is calculated, which shows the proportion of expenditure of private households on food, clothing, services and other categories. The result is then “fine-tuned” by using additional information from other sources. For example, “the expenditures for tobacco and alcohol are regularly cited too low in household surveys. Since tobacco and alcoholic beverages are subject to a special excise tax in Germany, the purchased quantities can be calculated very reliably from tax revenues and rates. The expenditure sums for tobacco and alcoholic beverages ascertained from the household surveys are corrected accordingly.” (Destatis)
The following figure shows the six largest expenditure categories.
German CPI: The six largest expenditure categories, comparison of the weighting patterns 2005 and 2010
Source: Federal Statistical Office, Regular Adaptation of the Consumer Price Index 2013, Figure 1
There has not been much change over the years. According to the data, households in Germany tend to spend roughly 30 percent of total expenditure for housing, 13 percent for transport and another 10 percent for food. These numbers explain the public attention paid to price changes in these categories.
The “basket” is the actual choice of goods and services for which prices are surveyed on a monthly basis.
The basket includes more than 300,000 individual prices. Here, product descriptions are broad in order to provide for the necessary flexibility in an environment where the variety and characteristics of items are changing constantly. One example Destatis mentions in this context is dishwashing agents. They can be cleaning agents in liquid form, powder form and tablets.
Sellers are asked always to include the highest-selling variety in the survey. The results may be deceptive as the following example illustrates. Assume cleaning agents in liquid form are initially the highest-selling variety in the basket and, getting more expensive, are bought less with people switching to cheaper tablets. As a consequence, tablets become highest-selling and replace the liquid form in the survey. This means, the rising price gets excluded and replaced by a lower one. Generally, if the highest-selling variety changes as a consequence of substitution or a general decline in the demand for goods in reaction to rising prices the results signal “deflation” and an underlying inflationary tendency may go unnoticed.
The basket is constantly revised taking into account model alterations of products and the fact that goods are no longer available. To quote Destatis:
“On a monthly basis the staff handling the price survey are faced with some of these items from the sample no longer being on offer and having to be replaced with new product variants. Also, when consumption significance of items drops excessively they are replaced. …
The proportion of monthly replacements varies greatly between the individual types of goods and can also fluctuate highly from month to month within one type of goods. Processed foods such as rice or toast bread usually have only very low monthly replacement rates, while the replacement rates for technical goods often are in the double digit range. Every month up to 10 % of all types of items and services in the basket are replaced.”
Again, as far as the reason for the “significant drop” in consumption is a rising price, an inflationary impulse may remain unrecorded.
Of course, all these effects will not stay unnoticed forever. For instance, the growing number of food banks and their clients in Germany and elsewhere is a brutal reminder that substitution is only one way in which rising food prices may show up. But when the signals become clearer it may be too late for monetary policy to act adequately.
To sum up: Currently, official price indices in major economies show no signs of broadly rising prices whatsoever. With regard to the long-lasting strong monetary growth in these economies, the question of the reliability of these statistics is of utmost importance. The example of the German consumer price index illustrates several shortcomings:
- Indices refer to special groups of goods and services and economic agents and sectors such as households, producers, imports or construction. This makes them poor early-warning signs of emerging inflation dynamics in a complex environment where prices influence one another across sectors and interact on different time scales.
- Indices are averages. They may include strong price movements in either direction some of which may have a greater inflationary potential in triggering adjustments throughout the economy than others.
- Indices capture only a fraction of goods and services. In a complex environment, the argument that some of those neglected would matter less than those included is not valid as it is the interplay of factors small and large which determines the final outcome.
- Although weighting patterns remain constant over a long time, the baskets of goods included are adjusted frequently. But the way in which adjustments are made is possibly blocking an early view on the emergence of inflationary impulses.
To stress it again: This text does not state that inflation is an immediate danger. It is only warning observers and policy makers to rely too strongly on an indicator which does not account properly for the underlying dynamics of price formation. Whether it is justified to trust the significance of official price indices for judging the inflationary potential of monetary policy in the current economic environment depends on the importance attributed to
(a) the existence of complex interdependencies of prices across sectors and agents and
(b) the extent of demand and substitution effects that are dampened or go unnoticed as a result of official adjustments.
The lesson here is caution. Debates about monetary policy and deflation leave the impression that there are great uncertainties with respect to two aspects: (1) How does money get into the economy and (2) how does money get into prices? Currently, both appear to be very open questions. The only point on which observers seem to agree is that most of the money created these days is not flowing into production and goods markets.
As traditional views of the transmission mechanism of monetary policy do no longer hold the problem is not that inflation is looming on the horizon but that
a) the dynamics of price interaction and propagation are poorly understood and
b) this lack of understanding of a complex interacting system – in which regarding the multitude of financial markets and channels monetary policy is often restricted to the role of an extra – makes the process uncontrollable.
In a world of so many unknowns it appears foolish and irresponsible to pour ever more liquidity into an economy which apparently has no productive uses of it at the moment. Other ways must be found for stimulating growth and fighting unemployment. Playing around with a “low inflation” target confident that official statistics present an accurate and timely view of the dangers involved, and that an “overshooting” can be identified and brought to a halt in due course, certainly is a very bad idea.