Persson: Why is inflation so low? The example of the retail sector
Introduction
Strong forces are affecting the developments in the world economy today. Globalisation and stiffer international competition are having an impact on individual countries' economies in many different ways. In Sweden, this has contributed, among other things, to a very low inflation rate in recent years. The subdued developments in the prices of imported goods partly reflect the integration of countries such as China and India into the world economy. Likewise, the surprisingly robust productivity growth can probably to some extent be attributed to the increased international competition.
Also contributing to the low inflation is the increasing establishment of low-price chains in the Swedish non-durables sector in recent years. This increases competition and helps squeeze Swedish food prices.
I intend to begin by briefly discussing global structural change in general and thereafter to concentrate primarily on the changes that we are seeing in the Swedish non-durables sector at the moment. Finally, I also intend to touch upon the current economic situation.
Globalisation raises the demands on flexibility
Globalisation means ”the growing mutual dependence between countries around the world through the increasing volumes and amounts of cross-border transactions in goods, services and capital flows”(1). This process has intensified in recent years, partly reflected in a sharp pick-up in world trade, an even faster increase in direct investment and a very steep rise in global equity and currency trade. Globalisation has given rise to rapid structural change, resulting in jobs being shed in certain sectors or being relocated to countries with lower production costs.
Sweden has, for a long time, experienced fast structural change, which contributed to the strong economic performance in the 1990s. Less than 100 years ago, around half of Sweden’s employed worked in agriculture and forestry; today, that figure is only 2 per cent. Over the same period, the proportion of the employed in the services sector has risen from roughly 20 per cent to 75 per cent, at the same time as the proportion in manufacturing increased at first from about 30 per cent to just over 40 per cent and then fell to 20 per cent. Among other things, we saw during the 1960s and 1970s that large parts of Swedish manufacturing, in the shape of the textile and shipbuilding industries, were outsourced from Sweden.
One consequence of structural change is that when old sectors or companies fail to survive, scope for new employment is created in sectors where the country enjoys a comparative advantage or in domestic services that are not exposed to competition to the same extent as the goods sector. For a long time, the growth of the public sector compensated for the drop in employment in manufacturing and created new jobs at an even faster rate than the old ones disappeared.
If such a change process is to proceed quickly and smoothly, it requires a lot in terms of the ability to adjust and the flexibility of people and companies. It also requires that policy facilitates the ability to adapt through the design of regulatory systems. The labour market has to function in a satisfactory way and it must be feasible to make the time difference between closure and new establishment as short as possible. In other words, it is a question of reducing the frictional unemployment that arises when industries or companies are wound up and people need to move or retrain in order to find work in a new location or in a new company or industry. The fact that employment is so low in Sweden today, given the state of the economy, may be a sign that the labour market is not flexible enough to meet the demands imposed by tougher competition or that the economy is not sufficiently dynamic.
Among other things, support is needed in the form of education and retraining to mitigate the effects of structural change on the labour market. Likewise, changes are necessary to facilitate innovation and entrepreneurship in Sweden so that new companies are created. It is also important to lay foundations for Swedish business to be competitive in foreign markets and to be able to attract capital, as well as for the economy as a whole to be able to attract labour to Sweden.
Globalisation has contributed to the low inflation
In the past decade or so, inflation has gradually fallen, both in Sweden and globally. Swedish inflation is very low today compared with the inflation rate we had in the 1980s and in the early 1990s (Chart 1). The same is true of inflation in the OECD area. One explanation is the shift that has occurred in economic policy and the establishment in many countries of independent central banks with the objective of price stability. But the deceleration in inflation has also been facilitated by globalisation, which, among other things, has resulted in higher competition from low-wage countries, forced old industrial countries to boost productivity and furthered the prospects for cheaper imports. This trend is reinforced by the opportunities that are created by the new information technology.
The low inflation is a result of several factors
If we consider the most recent years, inflation has been so low that it has undershot the Riksbank’s inflation target of 2 per cent. In addition to the shift in economic policy and the globalisation process, other factors, of course, have played a part, including the Riksbank’s rate cuts, which lower mortgage interest costs. The effects of sharp but temporary rises in energy prices have also contributed, e.g. in 2001 and 2003. But even when excluding these transitory effects, inflation has been very low.
The consumer price index, CPI, is the inflation measure on which the Riksbank’s target is focused. The CPI measures the price of a basket of goods and services for household consumption. By breaking down the CPI into prices of goods and services, excluding electricity, oil and mortgage interest rates, it becomes evident that falling goods prices have been the main contributor to the low price pressures (Chart 2). But services prices, too, are currently rising somewhat less than motivated by the generally lower productivity growth in the services sector.
Developments in goods prices are firstly influenced in part by price developments for the goods that we produce ourselves in Sweden, and in part by the price level and the price developments in the countries from which we import. Finally, movements in the exchange rate are important. The rate of increase in producer prices of domestic goods has been low in recent years. When we have investigated this in more detail we have concluded that it is a combination of low price increases for intermediate goods, subdued wage inflation and robust productivity growth in manufacturing that has caused lower domestic producer prices.
As regards producer prices of imported goods, these have dropped since the end of 2002. This partly reflects the more or less continuous strengthening of the krona between 2001 and 2004, which has had an impact on the prices with a slight lag. Moreover, consumer price inflation in the international market was muted during the last economic slowdown. We have also seen a rise in the share of imports from low-cost countries, which has also contributed to the low rate of price increases for consumer goods. The effect of this shift towards cheaper import countries is not fully captured, however, by the method that we so far have used to forecast import prices. This could be one reason why our forecasts have not completely managed to predict the fall in import prices. This is a problem that we are working on at the moment.
The distribution sector also important for inflation
The prices of consumer goods have fallen more than the aggregate producer prices of imported and domestic goods. That is because consumer goods prices, in a second step, are also affected by what happens in the distribution sector, i.e. the part of the business sector that distributes goods from producers to consumers. It is difficult to find comprehensive data for the distribution sector, but we have seen that unit labour costs in the sector that includes transportation fell in 2004. In the wholesale and retail trade, unit labour costs have increased at a slower and slower rate in recent years, not rising at all in 2004, and profit margins, moreover, have continued to shrink(2).
The retail trade is a significant part of the distribution sector and is half composed of non-durable goods, which largely comprise the food sector, while the other half consists of durables such as electronic goods and clothing.
Competition in the retail sector has increased considerably in recent years, to a great extent due to globalisation. For several years we have been able to buy home electronics at increasingly cheaper prices. Likewise, particularly in the recent period, we have seen greater downward price pressure in the Swedish non-durables sector, which has long been dominated by a few major players. We are seeing the establishment of more and more low-price chains, both foreign and Swedish chains, as well as massive advertising campaigns in which one large operator after the other announces price cuts. This development can to a large degree be seen as an effect of tougher competition and of the price consciousness that globalisation has entailed. I now intend to mainly concentrate on what is happening in the food sector and to put this into a slightly broader perspective.
Developments in food prices as a component of the CPI
Food is an important component of the CPI, comprising just over 13 per cent. That means that falling food prices can have a significant impact on the CPI. In the past two years, developments in consumer food prices have been well below the inflation rate in terms of both the CPI and the measure we use to look at underlying inflation, UND1X, as well as in terms of domestic inflation, UND1X (Chart 3).
Food prices are greatly affected by weather conditions, e.g. the fluctuations in inflation around the turn of the year 2001/2002 were largely attributable to the effects on fruit and vegetable prices of the floods in Europe. But temporary factors also play a part, such as mad cow disease, which first caused meat prices to rise and then to fall back in 2003. However, it is reasonable to assume that the recently low, and periodically negative, rate of price increases for food is due in some measure to stiffer competition in the non-durables sector.
Increased competition and new trends in the retail food sector
For a long time three large food chains have dominated the Swedish non-durables sector, and still do so. All three have integrated their businesses, meaning that their retail network and wholesale activities are part of the same company, thus making them active in both the wholesale and retail markets. As a result, the Swedish non-durables sector is strongly concentrated to a few players; together, they accounted for roughly 90 per cent of the non-durables market in 2004 (3).
It is not obvious what effects such a concentrated market will have on food prices. On the one hand, one can argue that fewer players should result in higher prices (market-power hypothesis), but on the other hand economies of scale should give rise to lower prices (efficiency-market hypothesis). So the fact that there are only three major participants in the Swedish market does not have to be a problem from the point of view of competition as long as they are exposed to pressures from potential competitors. That presupposes, though, that it is possible for new players to enter the market. In Sweden, the municipal planning process for local site development has long been an obstacle and may well have hindered competition. The fact that food prices in Sweden are higher than in many other countries could be an indication that the first effect is the dominant one.
Since 2002, however, we have seen a number of changes in the retail food sector. New foreign participants have entered the market and previously local players have branched out, shops have become bigger and the proportion of own brands has increased.
- Several foreign-owned low-price chains have set up both outside cities and in city centres. In addition, Swedish low-price chains are increasing their number of establishments. An ”unscientific” random sample that we have taken for a number of common food products – with the aim of representing the most important goods categories in the CPI – points to large price differences between low-price chains and traditional food shops in the Stockholm region; differences of up to 30 per cent. The prices at low-price chains outside the city centre were a further 10-20 per cent cheaper than those of the traditional chains in the city centre. Other surveys also indicate large price differences. The stiffer competition is being facilitated by a greater ambition on the part of politicians to increase the number of shop sites for low-price chains. For example, in Stockholm there are plans to raise the number of new sites considerably.
- Another trend we are seeing is that shops are becoming bigger and the share of sales accounted for by hypermarkets is rising. The total shop space is increasing at the same time as small food shops are being driven out of business. That entails a drop in costs and enables lower prices, partly in order to respond to the new chains’ lower prices.
- Food chains are also concentrating more and more on private labels, which is clearly evident on their shelves in terms of the greater share of products that bear the shops’ own brands as a complement to the different suppliers' brands. The purpose of private labels is to strengthen the chain’s brand name and to reduce costs through lower marketing expenses and fewer middlemen. In Switzerland and the UK, where this trend has been going on the longest, 30-40 per cent of the shops’ products are private labels(4). Sweden has long been one of the countries with the lowest share of private labels, around 10 per cent. According to a limited survey by the Swedish Competition Authority of typical low-price goods, the prices of supplier brands have developed in line with the CPI while the prices of private labels have fallen sharply.
Increased competition and lower prices
Food prices in the CPI have dropped more or less since the start of last year. Tougher competition increases the pressure on food retailers to cut prices in order to keep customers, even though other factors, such as service and quality, are important to consumers in their choice of shop. Lower prices are made possible by reduced profit margins, lower taxes or lower costs resulting from, for instance, lower wages, higher productivity or a more limited product range.
VAT rates in the Swedish non-durables sector have not changed during the current decade. However, wage costs have increased at a slower rate, which has led to lower unit labour costs in the wholesale and retail markets, despite a decline in productivity growth in recent years. Moreover, companies' profit margins have continued to narrow. So it seems that the stiffer competition has contributed to lower food prices primarily through smaller profit margins and low wage increases.
How much lower can Swedish food prices go?
One interesting question to ask is how large the scope is for lower Swedish food prices in the future. Current theory has it that the prices of identical products should be the same regardless of country or region if transportation costs and other trade barriers are excluded. According to the theory, price differences lead to cross-border trade and eventually to equal prices. Food prices in Sweden are substantially higher than the EU average, and large regional differences exist in Sweden. One possible way to investigate the scope for price cuts would be to take as a starting point the difference in Swedish food prices in relation to the prices of other countries and adjust for the other factors that affect the prices and that are different between the countries. Alternatively, one can look at the regional differences in food prices in Sweden.
Last year, Eurostat published two price comparisons, the first in cooperation with the OECD in order to measure purchasing power in different countries, while the second investigated price levels in capital cities. According to these surveys, Swedish food prices are 17 per cent above the average for the EU15, and 16 and 19 per cent, respectively, above those in Germany and the UK, while Stockholm’s food price level is some 14 per cent above the average for the EU15, and 17 and 18 per cent, respectively, above the level in Berlin and London.
Before we can draw any conclusions from such surveys, however, we have to adjust these figures for transportation costs as well as differences in taxes and wage costs. Transportation accounts for only a few per cent of the final food price, and the differences in transportation costs between EU countries in turn are so small that they do not have any great significance for the price differences. Nor are transportation costs likely to be lower in Finland, for example, which has a lower price level than Sweden.
The differences in VAT on food, however, are considerable. Swedish VAT on food is 12 per cent, and the general VAT rate is 25 per cent, while the rate is between 0 and 10 per cent in several comparable EU countries and just over 19 per cent on average in the EU15 (5). If we adjust for the respective VAT rates, the differences in food prices between Sweden and other European countries turn out to be smaller, but the majority of comparable countries still have up to 10 per cent lower prices. In comparisons between Stockholm and other European capitals, too, the price difference averages roughly 10 per cent when VAT is excluded.
Payroll taxes in Sweden are clearly higher than in many other countries; employers’ payroll taxes and social security contributions comprise around 15 per cent of GDP while the median in the EU15 is roughly 7 per cent. Given that wage costs make up a large portion of the food sector’s costs, this difference in itself could explain the difference in prices.
Detailed studies show that taxes, labour costs and other economic factors (e.g. GDP per capita and population density) may account for around half of the differences in food prices in Sweden and the EU(6) .The other half appears to be largely attributable to lower competition pressures in Sweden than in other countries.
That the price differences in relation to EU countries are due in large measure to the relative price competition is reinforced by the fact that there are also large price differences within Sweden, where the differences in wages and other costs between regions are small and the taxes the same. Studies(7) show that the prices in the south-east and west of Sweden are around 8 and 5 per cent lower, respectively, than in the eastern part of the country. The prices in Greater Stockholm are some 6 per cent higher than those in the Greater Gothenburg area. This reflects the differences in shop structure in various parts of the country, with a much larger market share for low-price chains and other chain stores than for the three major players in southern and western Sweden than in eastern Sweden.
Ongoing price squeeze in the non-durables sector
So far we can say that the retail food sector is seeing a price squeeze following the arrival of new foreign players, and the current operators have begun to compete through low-price subsidiaries or jointly owned companies. This price squeeze appears to have been made possible by shrinking profit margins and low unit labour costs. The potential for price reductions is difficult to estimate, but comparisons with other countries as well as between different regions in Sweden indicate that there is scope for lower food prices in Sweden.
For some time now the price squeeze has become increasingly noticeable to people in general in so far as we have more low-price shops to choose between and there is a wider range of cheap products in the traditional shops, but also through greater downward pressure on the prices of “ordinary” goods. It is difficult, however, to ascertain how long the price squeeze will continue and when it will have reached its full impact on the recorded CPI.
Concluding remarks
Globalisation and tougher competition have entailed effects both on price developments and on the labour market in Sweden and requires much in terms of the ability to adapt in the economy and in society on the whole. If the structural changes do not generate new employment quickly enough, the competition risks driving companies out of business and reducing the number of employees without this reduction being compensated for in other areas. We can also see how productivity in Sweden has grown stronger than what is usual at this stage of the business cycle.
In the current debate there have been claims for some time that the Riksbank is responsible for the fact that employment has not risen faster, by not having lowered the policy rate earlier and by more, which the presently low inflation would have permitted. Like other forecasters, the Riksbank underestimated the effects on inflation of the increasingly strong global competition and productivity growth. It cannot be ruled out that employment could have been somewhat higher today if we had been able to better foresee the developments, but we are talking about small differences and far from the figures that have been mentioned in the debate. It is important to remember that the repo rate is unusually low today, which means that monetary policy is already expansionary and thus not an obstacle to consumption and investment. The main reasons behind the weak employment are rather to be found in supply factors that the Riksbank cannot influence; among other things, the incentives to start businesses and hire staff as well as the willingness of people to accept employment, move and look for new jobs.
In the recent period, the Swedish economy has mainly been driven by firm export demand. This has generated healthy profits for companies and higher incomes for employees, and we are now at a stage where domestic demand is expected to take over as the driving force in the economy and employment to rise. But there is a risk that an improvement in the labour market will be held back by structural factors, i.e. that production and the labour market will not adapt quickly enough to changed conditions abroad, which could impair the prospects for a pick-up in domestic demand.
In addition, we have recently seen signs of somewhat weaker economic activity than we had previously expected. Should this tendency prove to be more persistent, there is a risk of an economic downturn, with a softer inflation outlook as a consequence. I have previously noted a risk that rising interest rates and falling house prices could lead to unnecessary strains in the real economy in the long run, with a drop in demand as a result. This remains a risk, but in a situation with clearly weakening economic activity it may prove necessary to conduct monetary policy in a more expansionary direction. However, we are currently in a position where the uncertainty regarding what course the developments will take is still considerable. Therefore, we will as usual be monitoring the situation very closely up to the next monetary policy meeting.
-----
Notes
1) IMF’s definition.
2) Source: National Institute of Economic Research.
3) Source: Fri Köpenskap.
4) Source: Gfk.
5) Source: International Tax Comparisons 2004, the Swedish Taxpayers Association.
6) See, e.g., the Swedish Competition Authority 2002a, 2002b and 2000:2.
7) Swedish Competition Authority, in cooperation with the Pensioners’ National Organisation.
