ArticlePDF Available

The Swedish ICT miracle - Myth or reality?

July 2005
Information Economics and Policy 17(3):275-301

July 2005
17(3):275-301

DOI:10.1016/j.infoecopol.2004.06.004

Source
RePEc

Authors:

Ericsson

This article investigates the productivity development in Sweden in the 1990s. The results show that much of the recorded Swedish surge in labor productivity was due to the spectacular growth of the Radio, television and communication equipment (RTC) industry. However, this article shows that the productivity growth of the RTC industry is very sensitive to value added price deflators. Unlike Sweden, the US uses hedonic price indexes for semiconductors and microprocessors which are important intermediate inputs in the RTC industry. Estimates based on the US intermediate input price deflators for semiconductors and microprocessors suggest that the productivity growth of the Swedish RTC industry during the 1990s is an artefact. This implies that the productivity growth of total manufacturing has also been overestimated. The results for Sweden are also interesting for other countries such as Finland, Ireland and South Korea where ICT producing industries have contributed substantially to labor productivity growth.

Labor productivity levels in manufacturing, value added (in thousands of Euros) per person engaged in Germany, Sweden and the US 1980-2001

Scatter diagram of the difference in productivity level between Sweden and the US in 1993 and the average Swedish labor productivity growth rate 1993–2000

Scatter diagram of the difference in productivity level between Sweden and Germany in 1993 and the average Swedish labor productivity growth rate 1993–2000

presents the estimates of the unit value ratios for Germany and the US. Food

Number of matches, coverage ratios and unit value ratios for the manufacturing industry in Germany and the US in 1997

Figures - uploaded by Harald Edquist

Content may be subject to copyright.

Content uploaded by Harald Edquist

Content may be subject to copyright.

SSE/EFI Working Paper Series in Economics and Finance

No. 556

February 2004

The Swedish ICT Miracle – Myth or Reality*

By

Harald Edquist‡

Abstract: This paper investigates the relative labor productivity level for total manufacturing in Germany,

Sweden and the US for the period 1980–2001. The paper also presents estimates of labor productivity

levels for 18 different manufacturing industries for the period 1993–2000. The results show that the

Swedish manufacturing productivity caught up with German and US productivity in the 1990s, overtaking

the German level in 1995 and coming very close to the US level by the end of the 1990s. It has been argued

that much of the Swedish surge in labor productivity during the second half of the 1990s was due to the

spectacular growth of the Radio, television and communication equipment (RTC) (ISIC 32) industry.

However, this paper shows that since 1998 Swedish RTC productivity has been declining relative to the

corresponding industry in Germany and the US. Moreover, it is shown that the productivity growth of the

ICT-producing industries is very sensitive to the value added price deflators that areused to calculate real

value added growth rates. Unlike Sweden, the US uses hedonic price indexes for semiconductors and

microprocessors. These electronic components are important intermediate inputs in the RTC industry.

Therefore estimates based on the US intermediate input price deflators for semiconductors and

microprocessors suggest that the productivity growth of the Swedish RTC industry during the 1990s is

partly a statistical artefact. This implies that the productivity growth of total manufacturing also has been

overestimated.

JEL-codes: O10; O30; O47; O57

Keywords: Information and communication technology (ICT), Productivity, Technological change,

Value added deflators

* I am especially grateful to Robert Inklaar, without whose help this paper never would have been written. I

would also like to thank Bart van Ark, Edwin Stuivenwold, Marcel Timmer and Gerard Ypma at the

Groningen Growth and Development Centre and Magnus Henrekson at the Stockholm School of

Economics for providing valuable comments and data for this paper. Moreover, I am grateful to Göran

Svensson, Bertil Klang and Per Ericson at Statistics Sweden and Arvid Wallgren and other participants at

the seminar at the Swedish Central Bank for insightful comments. Financial support from the Jan

Wallander and Tom Hedelius Foundation and the Carl Silvén Foundation is also gratefully acknowledged.

‡ Department of Economics, Stockholm School of Economics, PO Box 6501, SE-113 83 Stockholm,

Sweden. Phone: +46 8 736 92 59, Fax: +46 8 31 32 07, E-mail: Harald.Edquist@hhs.se

1

Table of Contents

1. INTRODUCTION .................................................................................................................................... 2

2. LABOR PRODUCTIVITY LEVELS IN MANUFACTURING .......................................................... 4

2.1 Currency conversion....................................................................................................................... 4

2.2 Unit value ratio data....................................................................................................................... 7

2.3 Productivity level results................................................................................................................. 8

2.3.1 Unit value ratio results ............................................................................................................................. 8

2.3.2 Productivity level benchmark results for 1997......................................................................................... 9

3. EXTENDING LABOR PRODUCTIVITY LEVELS BY GROWTH RATES ................................. 10

3.1 Time series data ............................................................................................................................ 10

3.1.1 Data description ..................................................................................................................................... 10

3.1.2 Price deflators ........................................................................................................................................ 11

3.2 Productivity level results for the manufacturing industry............................................................. 12

3.2.1 Total manufacturing ............................................................................................................................... 12

3.2.2 Industry level.......................................................................................................................................... 13

4. ICT DEFLATORS AND RELATIVE LABOR PRODUCTIVITY ................................................... 15

4.1 ICT deflators ................................................................................................................................. 15

4.2 A detailed investigation of the Radio television and communication equipment industry ............ 17

5. CONCLUSIONS..................................................................................................................................... 25

6. REFERENCES ....................................................................................................................................... 27

7. APPENDIX ............................................................................................................................................. 48

7.1 Appendix 1: Unit value ratios ....................................................................................................... 48

7.2 Appendix 2: ICT deflators............................................................................................................. 50

2

1. Introduction

During the 1990s productivity research increasingly came into focus. Comparisons of

productivity across countries and industries are important for evaluating economic

performance. Moreover, particular attention has been paid to productivity comparisons in

industries with rapid technological change and falling prices such as the Information and

Communication Technology (ICT) producing industry.

Comparing productivity in industries producing homogenous products is an easy task.

For example, in the crude oil industry, output is arrived at by a mere counting of barrels

of oil produced. However, measuring productivity in industries where technology

changes rapidly is a totally different matter. According to “Moore’s law” microprocessors

are halved in price and double in capacity every 18 months. A computer based on the

latest technology might be obsolete within a year or two. Is it then reasonable to compare

productivity in industries with rapidly changing technology and prices across countries?

Nordhaus (1997) argues that capturing the impact of new technologies on living

standards is beyond the practical capability of official Statistical Agencies. The essential

difficulty is that high-tech goods and services consumed today may not even have existed

a decade ago. Moreover, if they did, the quality of the goods that we consume today is

much higher compared to the quality of “the same” good a decade ago.

The increase in productivity growth in the US economy since 1995 (see Council of

Economic Advisers 2003) has resulted in an intense debate on the impact of ICT

technology on productivity in different countries. In Sweden, ICT technology created an

economic boom at the end of the 1990s. In 2000 Stockholm was named the Internet

capital of Europe by the Newsweek Magazine. According to Newsweek the Stockholm

phenomenon could be explained by “the looming marriage of the Internet and the third-

generation mobile telephony in Europe” (Newsweek 2000). Figures from Statistics

Sweden also supported the spectacular development of the Swedish Radio, television and

communication equipment (RTC) (ISIC 32) industry. For the period 1996–2000 the labor

productivity growth in RTC was approximately 35 percent per year.

3

Four years later, it is evident that much of the Swedish Internet era of the late 1990s was

a transient hype, partly created by media. However, it has been very difficult to explain

the fundamental fact that productivity growth in Swedish manufacturing and particularly

in the RTC industry increased so rapidly during the last years of the 1990s. Did the

increased productivity growth in manufacturing and RTC of the late 1990s reflect some

fundamental changes in the economy or was it largely a statistical artefact?

There have been a number of studies examining productivity development in Sweden

during the 1990s. Most of them investigate productivity growth in Sweden compared to

other countries (see Lundgren and Wiberg (2000), Edquist and Henrekson (2001, 2002),

Lind (2002, 2003) and Apel and Lindström (2003)). So far much of this research has

been focused on Swedish productivity growth, often in comparisons with productivity

growth in other countries. The results have emphasized the spectacular growth and the

increasing importance of the Swedish RTC industry. A common claim is that without the

spectacular growth of the RTC industry the productivity growth in total manufacturing

during the second half of the 1990s would have been much lower (Lind 2003). Moreover,

the productivity performance of the total manufacturing industry during the 1990s has

often been described as the “ICT miracle”.

Much research has been carried out about Swedish productivity growth. However, the

research on comparative productivity levels has been limited. Moreover, it has not been

clarified to what extent the use of country specific value added price deflators have

affected the growth in the RTC industry. The following questions have remained

unanswered: How big is the gap in productivity level for different manufacturing

industries between Sweden and other countries? Which industries have been catching up

during the 1990s? What impact does the use of different value added price deflators and

quality adjustments have on productivity growth and relative productivity levels in the

ICT-producing industries?1

1 For a definition of the ICT producing industries see OECD (2002).

4

The purpose of this paper is to answer the questions stated above. In sections 2 and 3 I

present estimates of labor productivity levels for Swedish manufacturing relative to the

corresponding levels in Germany and the US in 1980–2001. Moreover, I also provide

estimates of labor productivity levels for 18 manufacturing industries at the 2-digit ISIC2

level for the period 1993–2000. The method used for comparing productivity levels is

based on the industry-of-origin approach.3 In short, the industry-of-origin approach

converts output by industry to a common currency with a producer price-based and

industry specific Purchasing Power Parity, which is called Unit Value Ratio (UVR).4 In

section 4, the impact of value added price deflators for the ICT-producing industry is

investigated. Section 4 also compares the intermediate input and gross output price

deflators for ISIC 32 in Sweden and the US. Section 5 concludes.

2. Labor productivity levels in manufacturing

2.1 Currency conversion

In order to compare labor productivity levels between countries with different currencies,

it is necessary to convert the value added of different countries into a common currency.

Since price levels in different industries can vary substantially across countries, it is also

necessary to find a conversion method that is industry specific (Scarpetta et al. 2000).

The conversion can be made in a number of ways. One possibility is simply to use the

existing exchange rate between the two countries. However, this implies several

disadvantages. For example, the exchange rate is only based on traded goods, it is not

industry specific, it is affected by exchange rate policies and currency market fluctuations

and it does not adjust for international price differences (Monnikhof and van Ark 2002).

An alternative to the exchange rate is to use Purchasing Power Parities (PPPs). PPPs are

obtained from the expenditure side and reflect the relative price levels for private

2 ISIC stands for International Standard for Industry Classification and it is an UN based classification

standard (see United Nations Statistics Division (2000))

3 The industry-of-origin approach has been developed by the ICOP (International Comparisons of Output

and Productivity) group at the University of Groningen since 1983 (see van Ark and Pilat 1993).

4 The methodology of unit value ratios is discussed in more detail in section 2.1.

5

consumption, investment and government expenditure (van Ark and Timmer 2002). PPPs

are constructed by gathering expenditure prices for a large sample of products in each

country. The ratio between the expenditure prices for the same products in the two

countries are then used to construct the PPPs. Finally, the ratios of expenditure prices for

each product group are aggregated to a country specific PPP.

While PPPs are successfully used for comparisons of GDP and labor productivity at the

aggregate level, there are a number of problems associated with the use of PPPs for

industry level comparisons. One problem is that expenditure PPPs only apply to final

output, so that intermediate output is not covered by PPPs. According to Monnikhof and

van Ark (2002) intermediate products account for around one third of the value in

manufacturing. Another drawback with using the expenditure PPPs for comparisons on

the industry level is that they include margins, indirect taxes and subsidies. They also

include import prices, while export prices are excluded (van Ark and Pilat 1993).

According to van Ark and Timmer (2002) there are two alternatives to construct reliable

industry level PPPs. The first approach is to transform expenditure PPPs to industry

groups by “peeling off” indirect taxes and transport and distribution margins and thereby

create producer price level PPPs.5 The second approach is the industry-of-origin approach

that will be used in this paper. The industry-of-origin approach converts the currency by

using output data instead of expenditure data. The conversion is made by calculating unit

value ratios (UVRs).

Unit values (UV) are computed by dividing the ex factory value of output for a product

category by the produced quantities. The information is most often based on production

censuses or industrial surveys. In practice, products or product groups that are similar in

both countries are matched against each other. Unit values for the two countries are then

divided in order to obtain a product unit value ratio (UVR). Each product UVR indicates

the relative producer price of the matched product in the two countries. Product UVRs

5 This method was pioneered by Jorgenson and associates. For a more detailed description of the method

see van Ark and Timmer (2002).

6

are aggregated step by step to higher levels; from the product level to the industry level

and finally to the total manufacturing level.6

The industry-of-origin methodology also has some drawbacks. According to van Ark

(1996) there are three major problems7 with the UVR-method that affect the

comparability of the estimates across countries:

• In many sectors and industries UVRs are based on a limited sample of items. For

example, in manufacturing where the average percentage of output covered by

unit value ratios is between 15 and 45 percent, it is usually assumed that UVRs

for matched items within a manufacturing industry are representative for non-

matched items.

• Comparisons of unit values are affected by differences in product mix. Often

output values are only calculated for product groups instead of specific products.

This leads to problems on a disaggregated level because of the lack of harmonized

product coding systems between different countries.

• The unit value ratios also have to be adjusted to differences in product quality

across countries. However, it is even more serious in international comparisons

since the frequency of “unique products” that are only available in one country, is

higher than for comparisons over time.

Despite these caveats the industry-of-origin methodology appears to be the preferred

method for comparing productivity levels across countries. Nonetheless, it is important to

keep in mind that the industry-of-origin methodology has limitations and that results for

industries with low coverage ratios must be interpreted with caution.

6 For a detailed and more formal explanation of the industry-of-origin approach see appendix 1.

7 Another problem that is not discussed by van Ark (1996) is that UVRs are often used in a single deflation

procedure, which means that intermediate products are not included in the estimation of UVRs.

7

2.2 Unit value ratio data

The unit value ratios presented in this paper are based on two bilateral investigations for

the year 1997. The first investigation compares the unit value ratios between Germany

and Sweden and the second compares the unit value ratios between Germany and the US.

This allows for comparisons of Sweden and the US by using Germany as a link. The unit

value ratios between Sweden and Germany are based on data from the Eurostat Prodcom-

database (Europroms 2001). The unit value ratios between Germany and the US have

been calculated by Inklaar et al. (2003a) and are based on the Eurostat Prodcom-database

and the US manufacturing census for 1997.

Before aggregating the UVRs, outliers were removed from the Prodcom-database.8 For

the comparison between Germany and Sweden products with deviation more then 200%

and less than 75% of the EU average9 were removed. For the comparisons between

Germany and the US products with deviations more that 100 percent and less than 50%

of the EU average were removed. The reason for allowing a larger boundary for Germany

and Sweden is that Sweden is a smaller country with an economy characterized by a high

degree of specialization.10 Moreover, some product groups were deleted since it was

obvious that the product groups were not comparable across countries.11

The quantity of the Swedish product group Radio transmission apparatus with reception

apparatus (Prodcom 32201170) is missing. Since this product group has significant

importance for the RTC industry (ISIC 32) an estimation of the quantity has been made.

Table 1 shows the values of gross output and quantity for the Radio transmission

apparatus with reception apparatus (Prodcom12 32201170) divided into three different

subgroups. Quantity data only exists for the subgroup Transmission apparatus,

8 To remove outliers is a standard procedure in calculations of unit value ratios.

9 The average of the EU is based on at least four EU countries.

10 If a larger boundary is not used for Sweden and Germany, a very large number of product groups would

be removed since Sweden has a very specialized economy compared to the EU average.

11 For example, the product group Other machines and appliances for testing materials (Prodcom

33206259) was dropped since it was obvious that it contained different products that were not comparable

between Sweden and Germany.

12 Prodcom is a classification code for industry products at the 8-digit industry level.

8

incorporating reception apparatus, for cellular networks "mobile telephones" (CN13

85252091). It is therefore assumed that the Radio transmission apparatus with reception

apparatus (Prodcom 32201170) has the same gross output/quantity ratio as this subgroup.

This assumption appears to suggest that apples should be compared with oranges.

However, the intuition behind this assumption is that the production value for Radio

transmission apparatus with reception apparatus does not differ very much whether it is

used for radio-telephony, radio-broadcasting, television or cellular networks. This view is

supported by officials at the Swedish company Ericsson that is the largest supplier of

Radio, transmission apparatus with reception apparatus (Prodcom 32201170). According

to specialists14 at Ericsson the prices and technical specifications are approximately the

same for the two largest subgroups15 in table 1 (i.e. CN 85252091 and CN 85252099).

2.3 Productivity level results

2.3.1 Unit value ratio results

Table 2 and 3 present the results for the calculations of the unit value ratios. The unit

value ratios for 18 manufacturing industries in Sweden and Germany are shown in table

2. In total there were 802 matches between product groups in manufacturing. Food

products (ISIC 15–16), Wood and products of wood and cork (ISIC 20), Paper products

(ISIC 21) and Radio, television and communication equipment (ISIC 32) are the

industries with the highest coverage ratios. Office accounting and computing machinery

(ISIC 30), Medical, precision and optical instrument (ISIC 33) and Other transport

equipment (ISIC 35) have low coverage ratios. Medical precision and optical instrument

(ISIC 33) has the highest UVR with 15.83 SEK/EUR, while Office accounting and

computing machinery has the lowest with 5.83 SEK/EUR. For manufacturing the Fisher16

13 CN stands for Combined Nomenclature and is a classification code for industry products that is used by

Statistics Sweden. The CN code is compatible with the Prodcom classification code.

14 Interview with Olle Zimmerman 2004-01-13.

15 The radio, transmission apparatus with reception apparatus for civil aircrafts is such a small part of the

total production value of the industry that the assumed price has a very small effect on the total gross

output/quantity ratio for the total Radio, transmission with reception apparatus (Prodcom 32301170).

16 The Fisher exchange rate is derived by taking the square root of the product of the Paasche exchange rate

and the Laspeyres exchange rate.

9

exchange rate is 9.41 SEK/EUR which is higher than the average exchange rate of 8.65

SEK/EUR in 1997.

Inklaar et al. (2003a) also provide estimates of unit value ratios for manufacturing in

Sweden and Germany. Their results are based on 250 matches compared to 802 for the

study presented here. Moreover, the coverage ratios for Sweden are higher for all

industries except Chemicals (ISIC 22).17 The unit value ratios estimates for different

industries do not differ much between the results in this article and those by Inklaar et al.

However, there is a large difference for Medical precision and optical instrument (ISIC

33). In table 2 the UVR is 15.83 SEK/EUR for Medical precision and optical instrument,

while it is only 7.18 SEK/EUR in Inklaar et al. One reason for the difference is that

Inklaar et al’s UVR estimates are based on 3 matches, while the results in table 2 are

based on 16 matches.

Table 3 presents the estimates of the unit value ratios for Germany and the US. Food

products (ISIC 15–16), Textile, clothing, leather and footwear (ISIC 17–19), Paper

products (ISIC 21) and Basic metals (ISIC 27) have high coverage ratios. The lowest

coverage ratios are found for Printing and publishing (ISIC 22), Fabricated metal

products (ISIC 28) and Other transport equipment (ISIC 35). Printing and publishing

(ISIC 22) has the highest UVR with 2.12 Dollar/EUR and Textile clothing, leather and

footwear (ISIC 17–19) has the lowest with 0.66 Dollar/EUR. The Fisher exchange rate

for the whole manufacturing industry is 1.11 Dollar/EUR.

2.3.2 Productivity level benchmark results for 1997

Table 4 reports the labor productivity levels18 for the benchmark year 1997. The

differences in labor productivity level among industries within the same country depend

heavily on the capital intensity among industries. Therefore the interesting results are the

differences in relative productivity in the same industry across countries. According to

table 4 the Swedish Chemicals (ISIC 24) industry had the highest labor productivity level

17 For Chemicals the difference in coverage ratio is 5 percentage points between the result in table 2 and

Inklaar et al (2003a).

18 In this paper labor productivity is defined as value added per number of persons engaged.

10

relative to Germany and the US. Paper products (ISIC 21) also had very high levels of

labor productivity relative to Germany and the US. The Swedish manufacturing recycling

(ISIC 36–37) industry had the lowest labor productivity level relative to Germany, while

manufacturing recycling (ISIC 36–37) and RTC (ISIC 32) had the lowest productivity

level relative to the US. The highest labor productivity level for Germany relative to the

US was found for Printing and publishing (ISIC 22). The highest labor productivity level

in the US relative to Germany was found for RTC.

3. Extending labor productivity levels by growth rates

3.1 Time series data

3.1.1 Data description

The results of the relative productivity level for the benchmark year (1997) can be

extended to other years by using labor productivity growth rates (based on value added in

fixed prices). Labor productivity growth rates are calculated by using time series with

value added, value added deflators19 and employment. The labor productivity growth

rates are then used to calculate the change in relative productivity performance based on

the benchmark year. The Swedish time series data has been taken from the Swedish

National Accounts (Statistics Sweden 2003b). Due to changes in industrial classification

and the introduction of the new 1993 system of National Accounts (SNA), the Swedish

data only covers the period 1993–2001. This limits the estimation of the relative labor

productivity levels at a detailed industry level20 to the period 1993–2000. However, for

total manufacturing it has been possible to link time series of value added, value added

deflators and employment of the Swedish National Accounts with industry data from the

STAN database for 1980–1992 (OECD 2001b). This makes it possible to present

estimates of the productivity level in Swedish manufacturing for the period 1980–2001.

The data for Germany and the US are based on the 60-industry database (GGDC 2003).

19 A definition of value added deflators and how they are calculated can be found in section 4.2.

20 This paper presents estimates for most manufacturing industries at the 2-digit ISIC industry level.

11

Furthermore, all aggregation for the US and Germany have been based on Törnqvist

weights.21

3.1.2 Price deflators

One of the major problems with comparing productivity growth and levels across

countries is to construct similar and reliable deflators. All three countries use double

deflation22 in order to calculate the value added in fixed prices for the production side of

the economy. Double deflation means that the production value (gross output) is deflated

with an output price index23, while intermediate inputs are deflated with an input price

index. Since double deflation is used in all three countries there should not be a major

problem to compare the value added growth rates across countries. However, the value

added in fixed prices for Sweden is based on a Laspeyres volume index with moving

average based on year t–1, while value added in fixed prices for Germany and the US are

based on the Törnqvist index with moving averages based on the average of the year t–1

and t. The way these indexes are weighted influences the value added deflator. This is

further discussed in appendix 2. Appendix 2 also shows how the Swedish data is

approximated to provide estimates that closely approximate estimations based on

Törnqvist weights.

Another major problem when comparing productivity levels for different industries is the

different policies used by Statistical Offices to account for quality changes. In the US

hedonic price indexes are used extensively to account for the quality changes for the ICT-

producing industries. Sweden only uses hedonic price indexes for imports of computers,

while Germany does not use any hedonic measures (Scarpetta et al. 2000). Due to the

differences of price deflation in the ICT-producing industries I will use the US ICT-

deflators for the ICT-producing industries in Sweden and Germany. By applying the US

ICT-deflators also on Sweden and Germany one implicitly assumes that the industry

structure and price changes for the ICT-producing industry would be identical across

21 See appendix 2.

22 A thorough description of how value added price deflators are calculated and its implications for

productivity growth is made in section 4.2.

23 In this article gross output price deflators are based on producer price indexes.

12

countries. The empirical validity of these assumptions is questionable. In section 4, I

therefore analyze the effects of relaxing these assumptions on the productivity

development in RTC.

3.2 Productivity level results for the manufacturing industry

3.2.1 Total manufacturing

The labor productivity level estimates24 for total manufacturing for Germany, Sweden

and the US are presented in figure 1. The results in figure 1 indicate that the productivity

level in Swedish manufacturing was well below that of Germany and the US at the

beginning of the 1980s. During the 1980s Sweden caught up slightly with Germany,

while the productivity gap between Sweden and the US increased. During the late 1980s

and the beginning of the 1990s relative productivity levels remained unchanged.

However, from 1993 to 2001 Sweden was catching up with Germany and the US. In

1995 Sweden overtook Germany in terms of labor productivity and the productivity gap

between the two countries was increasing during the period 1995–2000. Moreover, labor

productivity gap between Sweden and the US was only 5 percent in 2001 compared to 33

percent in 1993.

The results for total manufacturing seem to correspond well with the growth patterns of

total manufacturing presented by Lind (2003). However, Inklaar et al. (2003b) present

estimates of labor productivity levels in manufacturing for EU countries and the US.

According to the results by Inklaar et al. the labor productivity level in manufacturing in

Sweden increased from 93.5 percent of the US level in 1979–81 to 99.3 percent in 1994–

1996. However, labor productivity fell to 86.6 percent for 1999–01. The fall in Swedish

labor productivity for manufacturing in the late 1990s is not supported by the results

presented here. One possible explanation is that Inklaar et al. use harmonized US

24 The labor productivity level results for total manufacturing are based on domestic deflators.

13

deflators for ICT producing industries,25 while the results in figure 1 are based on

national deflators.

3.2.2 Industry level

Estimates of labor productivity levels at the industry level (2-digit ISIC level) are less

certain than those of total manufacturing. It is important to keep in mind that the results

presented for the industry level are based on the assumption that the unit value ratios also

apply for unmatched product groups. This implies that the result for industries with low

coverage ratios must be interpreted with caution (see table 2 and 3). Nevertheless, labor

productivity level estimates for different manufacturing industries at the more

disaggregated level are important in order to understand the dynamics of productivity

changes in manufacturing. Table 5 and 6 present labor productivity estimates at the

industry level for Germany, Sweden and the US for the years 1993 and 2000. These

estimates were calculated by extending the benchmark estimates for 1997 with labor

productivity growth rates (in fixed prices).

The results in table 5 show that in 1993 Sweden had its highest labor productivity level

relative to Germany and the US in Chemicals (ISIC 24). Labor productivity in this

industry was approximately 80 percent higher than in the US and Germany. Paper

products (ISIC 21), Fabricated metal products (ISIC 28) and Office, accounting and

computing machinery (ISIC 30) were other industries where relative productivity was

high in Sweden. Electric machinery and computing (ISIC 31), Radio, television and

communication equipment (ISIC 32), Motor vehicles, trailers and semi-trailers (ISIC 34)

and Manufacturing recycling (ISIC 36–37) were industries where Swedish relative

productivity was low compared to Germany and the US.26

In 2000, Chemicals (ISIC 24) and Paper products (ISIC 21) still had the highest labor

productivity relative to Germany and the US. Radio, television and communication

equipment (ISIC 32), Motor vehicles, trailers and semi-trailers (ISIC 34) and

25 The use of different value added deflators will be discussed in detail in section 4.2.

26 It is important to keep in mind that labor productivity level results for industries with low coverage ratios

must be interpreted with caution.

14

Manufacturing recycling (ISIC 36–37) had the lowest labor productivity levels relative to

the US, while Electrical machinery and computing (ISIC 31), Medical, precision and

optical instrument (ISIC 33) and Other transport equipment (ISIC 36–37) had the lowest

productivity level relative to Germany.

Tables 4–6 show that the relative labor productivity level for Swedish manufacturing

industries throughout the period 1993–2000 was high for Chemicals (ISIC 21) and Paper

products (ISIC 21). However, it was not these industries that had the highest growth

rates throughout the period. As documented by Edquist and Henrekson (2001) it was the

ICT-producing industries that experienced the highest growth during the latter part of the

1990s in Germany, Sweden and the US.

Tables 4–6 also indicate that there was a relative increase in the labor productivity level

of the Swedish RTC industry relative to Germany and the US for the period 1993–1997.

However, from 1997 to 2000 the German RTC industry caught up with and forged ahead

of its Swedish counterpart in terms of labor productivity level. The same pattern can be

found for the US RTC industry, even though relative labor productivity was higher in the

US throughout the period 1993–2000. Lind (2003) argues that RTC has been crucial for

economic growth in Swedish manufacturing. The results in table 4–6 do not imply that

the growth rate was low in the Swedish RTC industry for the period 1997–2000. Instead

the results indicate that for the period 1997–2000 labor productivity growth for this

industry was higher in both Germany and the US compared to Sweden. From 1997–2000

Sweden lost much of its labor productivity edge in RTC compared to Germany and the

US.

There is a well known hypothesis that productivity growth rates vary inversely with

productivity level. This has to do with the level of technology embodied in a country’s

capital stock. When a leader in technology invests in new capital the accompanying

productivity increase is limited by the advance of knowledge between the time when the

old capital was installed and the time it is replaced (Abramovitz 1986). However, a

lagging country has the opportunity to embark on a catching-up process by borrowing

superior techniques from the more advanced economies. This implies that the larger the

15

gap between leader and follower the greater the follower’s potential for productivity

growth.

This catching-up hypothesis can also be applied to industries. Figures 2 and 3 show the

correlation of the difference in the Swedish productivity level relative to Germany and

the US in 1993 and the average Swedish labor productivity growth rate for industries at

the 2-digit ISIC level. The results in figures 2 and 3 indicate a negative correlation.

However, the correlation evidence is not very strong. One possible reason to that there is

no strong correlation is that the period investigated is very short. Many articles that

investigate the catching up hypothesis use time periods of at least 25 years. There are also

several other reasons why there is no catching up at the industry level. According to

Gerschenkron (1962) different countries have different productive and organizational

structures of industry. For example, Hansson and Henrekson (1994) found that

competition could explain that the Swedish tradadables sector was catching up 1970–85,

but not the nontradables sector.

4. ICT deflators and relative labor productivity

4.1 ICT deflators

The results for the ICT-producing industries presented in tables 4–6 are based on the US

ICT deflators (see section 3.1.2). Applying the US deflators for all three countries

automatically assumes that the industry structure of the Swedish and German ICT-

producing industries are identical to the US and that the price decline for all products

would be the same in all three countries. These assumptions are not empirically valid. In

this section, I will therefore try to relax these assumptions by comparing deflators for the

three countries. An interesting question then is what effect the use of different value

added deflators has on measured productivity?

16

Table 7 shows the deflators based on the calculations from each country’s statistical

office.27 The approximation of the Swedish deflators to the German and US deflators are

described in appendix 2. Table 7 shows that the US deflators for Office, accounting and

computing machinery (ISIC 30) are m7uch more negative than those for Sweden and

Germany. Interestingly, the German deflators are more negative than the Swedish ones

for Office, accounting and computing machinery. One reason to this could be that the

structure of the industry is very different in the two countries. For example, the US

Office, accounting and computing machinery industry could be producing more

semiconductors and microprocessors, while the corresponding industry in Sweden

produces other types of computer equipment. For RTC (ISIC 32) the Swedish deflators

are more negative than both the US and German deflators for all years except for 1998

when the US deflator is slightly more negative than the Swedish one. The deflators for

Electric machinery and computing (ISIC 31) and for Medical, precision and optical

instruments (ISIC 33) do not differ as much as the other two ICT-producing industries in

the three countries.

What effects does the use of different deflators have for the estimates of relative labor

productivity? Table 8 presents the result for relative labor productivity for Sweden and

Germany when different ICT-deflators are used. The results indicate that the use of

different deflators have large impacts on labor productivity levels for Office, accounting

and computing machinery (ISIC 30) and for Radio, television and communication

equipment (ISIC 32). If the US deflators are used for the German Office, accounting and

computing machinery industry, while the Swedish deflators are applied to the same

industry in Sweden, this results in a substantial decline in the relative labor productivity

level for the Swedish Office, accounting and computing machinery industry. According

to table 8 the labor productivity level for Office, accounting and computing machinery

went from being 18 times higher than the German level in 1993 to becoming only one

half of the German labor productivity level in 2000. There is no empirical evidence that

can justify these results. Nevertheless, the results clearly show how sensitive productivity

calculations are to large differences in value added deflators over a longer time period. In

the other two cases (see table 8), the productivity level in the Swedish Office, accounting

27 An exact description of how value added price deflators are calculated is presented in section 4.2.

17

and computing machinery industry remains higher relative to the same industry in

Germany for the period 1993–2000.

For RTC the Swedish labor productivity level increases throughout the period 1993–2000

relative to Germany when country specific deflators are used. When the US deflators are

applied for Germany and the country-specific ones for Sweden, the result shows that

Swedish relative labor productivity increased for the period 1993–1997. After 1998 there

is a decline in the Swedish relative labor productivity level and in 2000 the higher

productivity level in Sweden has almost disappeared. When the US deflators are applied

on both countries, there is a similar decline in the Swedish relative labor productivity

level after 1998. For the year 2000 the relative labor productivity level is only 93 percent

of the German labor productivity level. On the other hand, if country specific deflators

are used for both countries the productivity level increases throughout the period 1993–

2000.28

The results presented in table 8 show that the use of different deflators for the ICT-

producing industries has a large influence on the relative labor productivity level between

Sweden and Germany. Nonetheless, the results in table 8 strongly suggest that the labor

productivity level for the Swedish RTC (ISIC 32) industry relative to Germany has

decreased since 1998. This does not imply that the productivity growth rate for this

industry has been slow in Sweden since 1998, but rather that the Swedish RTC industry

has lagged in labor productivity relative to Germany since 1998.

4.2 A detailed investigation of the Radio television and

communication equipment industry

During the period 1993–2000 labor productivity growth in the Swedish RTC industry

was 47 percent per year. Figures 4–6 illustrate the development of the RTC industry in

Germany, Sweden and the US. Figure 4 shows that gross output in the Swedish RTC

industry as a share of gross output in manufacturing, increased from 4 percent in 1993 to

28 To compare labor productivity growth rates for Germany and Sweden with country specific value added

deflators is very problematic, since Sweden uses other types of quality adjustments than Germany (see

section 3.1.2).

18

12 percent in 2000. The corresponding figures for Germany and the US were

approximately 2 and 6 percent 1993–2001. As illustrated by figure 5 the value added in

the Swedish RTC industry as a share of value added in manufacturing also increased

considerably during the 1990s. However, the corresponding share for the US RTC

industry was higher in 1993–2000. Figure 6 shows that the number of persons engaged in

the Swedish RTC industry as a share of total manufacturing increased from around 4

percent in 1993 to 6 percent in 2000. The number of persons engaged in RTC related

service industries such as data-consulting and data-services also increased considerably

during the 1990s (Johansson 2004).

Figures 4–6 show that the Swedish RTC industry became increasingly important for the

Swedish economy during the 1990s. It is therefore crucial that the productivity

development in the Swedish RTC industry is correctly measured. Table 8 showed that the

use of different deflators for the RTC industry can have enormous effects on productivity

growth measures. By using US deflators also for the German and Swedish ICT-producing

industries one implicitly assumes that the structure of the ICT-producing industries is the

same in all three countries and that the price fluctuations of output and intermediate input

prices are identical. In this section, I investigate what happens with the deflators for the

Swedish and the US RTC industry when these assumptions are relaxed.

When comparing ICT deflators across countries it is crucial to understand how the value

added in different countries is deflated. Both the Swedish and the US National Accounts

are based on double deflation to arrive at a value added in fixed prices (see section 3.1.2).

Double deflation implies that the values of gross output and intermediate input are

deflated separately with an output price index and an intermediate input price index,

respectively. These two series are then used to arrive at value added in fixed prices. More

specifically, value added in fixed prices can be defined as an average of the price change

in gross output )

ln

ln

(t

P

Output

∂

∂

and the price change of intermediate inputs )

ln

ln

(t

P

Input

∂

∂

. The

price change of intermediate inputs is weighted by the share of intermediate inputs in

gross outputs )( QP

MP

Output

Input and the entire expression is multiplied by the inverted share of

19

value-added in gross output )( VAP

QP

VA

Output (OECD 2001a). The exact relation for the value

added price deflator and intermediate input and output prices is shown in the following

expression:















−=

dt

Pd

QP

MP

dt

Pd

VAP

QP

dt

Pd Input

Output

InputOutput

VA

Output

VA lnln

ln (4.1)

Equation 4.1 shows that the price change in intermediate inputs has a large influence on

the value added price deflator if the proportion of intermediate input as a share of total

output is high.

Figure 7 shows the gross output and intermediate input price deflators for RTC in

Sweden and the US. According to figure 7 the US gross output and the intermediate input

prices decreased more rapidly than the corresponding gross output and input prices for

Sweden. The average price deflator for the Swedish intermediate inputs was zero, while

the average price deflator for the US intermediate inputs was –0.05 for the period 1994–

2001. For the output prices the average price deflator for Sweden was –0.10 and for the

US –0.18. For which products have the price deflator for the intermediate input prices

and for the output prices decreased more in the US compared to Sweden?

To answer this question I investigate the price deflators for RTC at a more disaggregated

industry level. At the 3-digit ISIC industry level, RTC consists of the following three

industries: Electronic valves and tubes29 (ISIC 321), Telecommunication equipment

(ISIC 322), Radio and television receivers (ISIC 323). Figures 8–10 compare the gross

output price deflator for these three industries in Sweden and the US for the period 1994–

2001. For Sweden there exists two price indexes for the three industries. One price index

is published by the Department of Prices and Consumption and the other is based on the

National Accounts.30 The difference between the two price indexes is that the price index

29 By and large, Electronic valves and tubes (ISIC 321) consists of the production of semiconductors and

microprocessors.

30 The Department of Prices and Consumptions and the National Accounts are both Departments at

Statistics Sweden.

20

published by the Department of Prices and Consumption is based on a product mix that is

lagged two years, while the price index in the National Accounts is not.31 Moreover, the

output price index in the National Accounts is an industry index, which means that it

includes both goods and services, while the index published by the Department of Prices

and Consumption is a product index which only represents goods.

Figure 8 shows that the US gross output price deflator for Electronic valves and tubes

(ISIC 321) was much more negative than the corresponding Swedish deflator throughout

the period 1994–2001. Figure 9 shows that the Swedish gross output price deflator for

Telecommunication equipment (ISIC 322) differs considerably for the years 1997–2001

depending on which price index that is used. For the years 1997–2000 the difference is

approximately 10 percent per year. According to the Department of National Accounts

these differences are due to the fact that the Department of Prices and Consumption uses

a product mix that is lagged two years. However, it is difficult to accept that this would

explain the whole difference of approximately 10 percentage points per year 1997–2000

between the two output price indexes.32 According to the price index published by the

Department of Prices and Consumption the Swedish gross output prices for

Telecommunication equipment (ISIC 322) has declined less than the corresponding US

deflator 1997–2000. However, the price index in the National Accounts suggests that the

Swedish price deflator has been approximately the same as the US deflator. Figure 10

indicates that for the period 1994–2001 the Swedish gross output price deflator for Radio

and television receivers (ISIC 323) has been more negative than the corresponding US

deflator.

Intermediate input price deflators for Sweden are not available at the 3-digit ISIC

industry level.33 Figure 11 shows the US intermediate input price deflators for Electronic

valves and tubes (ISIC 321), Telecommunication equipment (ISIC 322) and Radio and

31 On February 9th 2004 the Department of National Accounts at Statistics Sweden decided to release their

output price indexes for the RTC industry at the 3-digit level. The output price indexes published by the

Department of National Accounts had not been public at the 3-digit level and they were released after a

close investigation of an earlier draft of this paper by officials at Statistics Sweden.

32 One explanation to the large difference between the two indexes could be that the Swedish

telecommunication company Ericsson decided to outsource the manufacturing of cell phones abroad during

this period.

33 Statistics Sweden does not publish input price deflators for the 3-digit ISIC level.

21

television receivers (ISIC 323) 1991–2001. For the period 1991–1995 the intermediate

input price deflators for all three industries were close to zero. However, for the period

1996–2001 the price deflators have become more negative in all three industries. The

decrease has been more rapid for Electronic valves and tubes (ISIC 321) and

Telecommunication equipment (ISIC 322) compared to Radio and television receivers

(ISIC 323).

One possible explanation to the larger decrease in the intermediate input and output price

deflators in the US (see figure 7) is that the US systematically uses hedonic adjustments

for semiconductors and microprocessors. This implies that the improved quality in

semiconductors and microprocessors is considered when the price changes are estimated.

Since the invention of the transistor in 1948 there has been an extraordinary increase in

the capacity of semiconductors. According to “Moore’s” law microprocessors are halved

in price and double in capacity every 18 months. In Sweden hedonic price adjustments

are not used to take the quality improvements of semiconductors and microprocessors

into account. This could be the reason why the gross output Swedish price deflators for

Electronic valves and tubes have not decreased as much as in the US (see figure 8).

Since semiconductors are important intermediate inputs in Telecommunication

equipment (ISIC 322) and Radio and television receivers (ISIC 323), it is likely that the

use of hedonic price adjustments for semiconductors also influences the input deflators

for these industries. The fact that Sweden is not using hedonic adjustments for

semiconductors and the lack of Swedish price data for intermediate inputs at the 3-digit

ISIC level for RTC cause problems for accurately comparing price deflators between

Sweden and the US.

Triplett (1996) has shown that if the output price decline in the semiconductor producing

industry is underestimated this means that the intermediate input price decline in

computers is also underestimated. Thus, if the output price decline in the semiconductor

producing industry is overestimated, the intermediate input price decline in computers

would be overestimated. This means that if all intermediate inputs where produced

domestically, the measured productivity for the computer industry would be correct

22

despite the incorrect measurement of prices in the semiconductor producing industry.

Though, the measured productivity would be incorrect for less aggregated industries

within the computer industry such as the semiconductor industry. If the findings by

Triplett are applied on the RTC industry this means that if all semiconductors that are

used in the RTC industry also were produced domestically by the RTC industry the

productivity for the whole RTC industry would be unaffected if the price decline of

semiconductors were underestimated. However, the reasoning by Triplett is only correct

as long as all semiconductors are produced domestically.

Figure 12 shows the value of imports of Electronic valves and tubes as a share of the

total value of production and imports. According to figure 12 approximately 75 percent

of the Electronic components that were used in Swedish RTC industry were imported in

1995–2001. Hence, Triplett’s results do not hold for the Swedish RTC industry. If the

estimated prices of semiconductors are incorrect, the effect on intermediate inputs is

much larger since approximately 75 percent of the electronic components that are used as

intermediate inputs in the RTC industry are imported. How would the Swedish value

added price deflators change if hedonic price adjustments were made also for

semiconductors in Sweden? In order to give an accurate answer to this question it would

be necessary to have price data at a very detailed product level for Sweden and the US.

This data is not available for Sweden due to secrecy. Nevertheless, table 9 and 10 provide

estimates of how value added deflators would change if hedonic price indexes also were

used for semiconductors in Sweden.

Table 9 and 10 shows the recalculation of the Swedish value added deflators under the

assumption that the Swedish intermediate input prices for Electronic valves and tubes

(ISIC 321), Telecommunication equipment (ISIC 322) and Radio and television receivers

(ISIC 323) are the same as for the corresponding industries in the US. The intuition

behind this assumption is that price changes of all intermediate inputs except

semiconductors would be the same in the US and Sweden. It is true that prices vary

between different markets, however a large part of the intermediate inputs in the RTC

industry is purchased globally at world market prices. Moreover, it is also assumed that

the Swedish gross output price deflators for Electronic valves and tubes (ISIC 321) are

23

equal to the corresponding industry in the US. The intuition behind this assumption is

that if hedonic prices were implemented in Sweden for semiconductors and

microprocessors the price decline in the semiconductor producing industry would equal

that in the US. This is a plausible assumption since semiconductors are often priced and

purchased at world market prices (Triplett 1996).

Neither Sweden nor the US use hedonic price indexes for estimating gross output price

deflators for Telecommunication equipment (ISIC 322) and Radio and television

receivers (ISIC 323). Therefore, the calculations in table 9 and 10 for these industries are

based on domestic price indexes for Sweden. Gross output price deflators for

Telecommunication equipment (ISIC 322) and Radio and television receivers (ISIC 323)

in table 9 are based on the price indexes by the Department of Prices and Consumption,

while the price deflators in table 10 are based on the price indexes in the National

Accounts. Finally, the prices are weighted by the specific industry structure of the

Swedish RTC industry (measured as shares of production in gross output and

intermediate inputs at factor costs).

Not surprisingly, the results of the recalculated deflators presented in table 9 and 10

differ widely from the results of the official value added deflators presented in table 7.

The largest difference can be noticed for the period 1997–2000. The recalculated value

added price deflators in table 9 are even positive for the years 1997, 1998 and 2000. The

recalculated value added deflators in table 10 are all negative, but less negative than the

value added deflators in table 7. The reason for the large difference between the deflators

in table 7, 9 and 10 is that the method to calculate the value added price deflator is very

sensitive to the development of the intermediate input34 price deflators. The reason why

Sweden is much more sensitive to price changes in intermediate inputs than the US is

because the intermediate input/gross output ratio for the Swedish RTC industry is much

larger compared to the US.

Figure 13 shows the intermediate input/gross output ratio for the Swedish and the US

RTC industry 1993–2001. During the period investigated the Swedish intermediate

34 Semiconductors and microprocessors are important intermediate inputs in RTC.

24

input/gross output ratio has been constantly higher than the US. Since 1998 the Swedish

ratio has increased dramatically and in 2001 intermediate inputs exceeded the total gross

outputs. Hence, value added in current prices was negative. This development is due to

the increased outsourcing by the Swedish telecommunication company Ericsson. In

Sweden a very large part of the total output of RTC is produced by Ericsson. This implies

that the bulk of intermediate input prices that are reported to Statistics Sweden are

determined by the pricing of one single firm. Semiconductors are often purchased and

priced on the world market. However, if semiconductors or other intermediate inputs are

produced by Ericsson abroad and then imported and used in the Swedish RTC industry,

there is a risk that the internal pricing by Ericsson would not reflect world market prices

of semiconductors and other inputs. It is unclear to what extent Ericsson produces its own

intermediate inputs abroad. However, if a large share of Ericsson’s inputs are produced

abroad by Ericsson and imported, there is a possibility that price changes of

semiconductors and other inputs would be measured incorrectly. This would result in

incorrect productivity estimates for RTC in Sweden.

The value added deflators presented in table 9 and 10 have a great impact on how the

productivity growth in the Swedish RTC industry is measured. Figure 14 shows the labor

productivity growth in the RTC industry 1994–2000 with the official value added price

deflators (see table 7) and the recalculated deflators (see table 9 and 10). The results

show that the productivity growth differs widely depending on which deflators that are

being used. The price deflators based on the price indexes published by the Department

for Prices and Consumption (see table 10) give the largest difference in productivity

growth compared to the official deflators. However, the difference in productivity growth

is also large when the deflators based on the price indexes in the National Accounts are

used instead of the official deflators. The annual productivity growth becomes 20 percent

instead of 35 percent 1997–2000 if the recalculated deflators based on the price indexes

in the National Accounts are used instead of the official deflators.

The use of different deflators also has implications for the growth in total manufacturing.

Figure 15 shows the growth rate of total manufacturing with official and recalculated

deflators. For the period 1997–2000 the growth rates of total manufacturing would be

25

considerably smaller if the recalculated deflators are used. The effect on productivity

growth in manufacturing is smaller if the recalculated deflators based on the price

indexes in the National Accounts are used instead of the deflators based on the price

indexes published by the Department of Prices and Consumption. However, in 1998 the

productivity growth in manufacturing would be about one third lower with the

recalculated value added deflators based on the price indexes in the National Accounts.

The relative productivity development in Sweden is also affected by the use of different

value added deflators. Figure 16 shows the relative productivity with the recalculated

deflators based on the price indexes published by the Department of Prices and

Consumption for the period 1993–2000. The conclusion is that Sweden has only been

growing at the same rate as in the US. The catching up effect in the end of the 1990s (see

table 1) has been eroded.

5. Conclusions

I have used the industry-of-origin methodology to investigate the development of labor

productivity levels in Swedish manufacturing relative to manufacturing in Germany and

the US. The results show that Swedish manufacturing productivity caught up with levels

in Germany and the US during the 1990s. In 1995 Sweden overtook Germany in terms of

labor productivity level and continued to catch up with the US throughout the period

1995–2000. Moreover, Chemicals (ISIC 24) and Paper products (ISIC 21) had the

highest relative labor productivity compared to Germany and the US in 1993–2000.

Evidence of the increasing importance of the RTC industry for total manufacturing in

Sweden during the 1990s was also presented. For RTC, labor productivity increased

substantially in Sweden relative to Germany and the US in 1993–1998. However, for the

period 1998–2000 labor productivity of the Swedish RTC industry declined relative to

Germany and the US. This suggests that the productivity growth of RTC was slower in

Sweden than in the US and Germany 1998–2000.

The results of the labor productivity levels for Office accounting and computing

machinery and RTC turn out to be very sensitive to the choice of value added price

26

deflators. Value added price deflators are used by Statistical Offices to take price and

quality changes into account. Moreover, value added price deflators differ widely among

industries and countries. The Swedish value added price deflators for RTC was

considerably more negative compared to the German and US deflators throughout the

period 1993–2000.35

One explanation to why value added price deflators are more negative in Sweden than in

the US is that the US Statistical Agencies systematically use hedonic adjustments for

semiconductors and microprocessors, while Statistics Sweden is not. Hedonic price

indexes take the improved quality in semiconductors and microprocessors into

consideration when the price changes are estimated. Moreover, semiconductors and

microprocessors are important inputs in the Swedish RTC industry. Calculations of the

Swedish value added deflators based on the US price development for semiconductors

and microprocessors, show that the productivity growth in the RTC industry becomes

considerably lower. This suggest that the spectacular labor productivity growth exceeding

35 percent per year in 1996–2000 for the Swedish RTC industry is partly an artefact.

Moreover, the results show that it is dangerous to draw conclusions from international

productivity comparisons in industries characterized by rapidly changing technology.

The overestimation of labor productivity growth for Swedish RTC also has important

effects for productivity growth in total manufacturing. If the recalculated value added

deflators for RTC are used in order to calculate labor productivity growth rates for total

manufacturing, the productivity performance is less impressive than what is suggested by

official data. Using the revised estimates Sweden caught up with German and US labor

productivity levels during the first half of the 1990s. However, for the period 1997–2000

the labor productivity level was lower than suggested by official data. From a policy

perspective this is an important result, because it shows that the productivity growth

miracle in Swedish manufacturing during the late 1990s is partly an artefact.

35 Except for the year 1998.

27

6. References

Abramovitz, Moses (1986), “Catching Up, Forging Ahead, and Falling Behind”, Journal

of Economic History, Vol. 46, No. 2, pp. 385–406.

Apel, Mikael and Lindström, Tomas (2003), “Informationsteknologins betydelse för den

svenska produktivtetsutvecklingen – ännu en pusselbit”, Ekonomisk Debatt, Vol. 31, No.

5, pp. 29–37.

van Ark, Bart and Pilat, Dirk (1993), “Productivity Levels in Germany, Japan, and the

United States: Differences and Causes”, Brookings Papers on Economic Activity:

Microeconomics 2, Washington D.C.

van Ark, Bart (1996), “Issues in Measuring and International Comparison Issues of

Productivity – An Overview”, OECD Expert Workshop on Productivity: International

Comparison and Measurement Issues, OECD, Paris.

van Ark, Bart and Timmer, Marcel (2002), “Measuring Productivity Levels – A reader”,

OECD Working Paper, Paris.

Council of Economic Advisers (2003), Economic Report of the President, United States

Government Printing Office, Washington D.C.

Edquist, Harald and Henrekson, Magnus (2001), “Solowparadoxen och den nya

ekonomin”, Ekonomisk Debatt, Vol. 29, No. 6, pp. 409-419.

Edquist, Harald and Henrekson, Magnus (2002), “Kommer IKT-revolutionen även att att

lyfta europas ekonomier?” in Litan, Robert E. and Rivlin, Alice M. (2002), Bortom

dot.com-företagen, SNS Förlag, Stockholm.

Europroms (2001), European production and market statistics: EU-15

Prodcom/Combined Nomenclature, Eurostat, Luxemburg.

Gerschenkron, Alexander (1962), Economic Backwardness in Historical Perspective,

Harvard University Press, Cambridge (Massachusetts).

Groningen Growth and Development Centre (GGDC) (2003), 60-industry database,

Groningen. Available online: http://www.ggdc.net

IMF (2003), PPI Manual, Washington DC. Available online: http://www.imf.org

Hansson, Pär and Henrekson, Magnus (1994), “Catching up in industrialized countries: a

disaggregated study”, Journal of International Trade & Economic Development, Vol. 3,

No.2, pp. 129–145.

Inklaar, Robert, Stokes, Lucy, Stuivenwold, Edwin, Timmer, Marcel and Ypma, Gerard

(2003a), “Chapter VII Data Sources and Methodology” in O’Mahony, Mary and van Ark,

28

Bart (eds), EU Productivity and Competitiveness: A Sectoral Perspective. Can Europe

Resume the Catching-up Process? European Commission, Luxemburg.

Inklaar, Robert, O’Mahony, Mary, Robinson, Catherine and Timmer, Marcel (2003b),

“Chapter III Productivity and Competitiveness in the EU and the US” in O’Mahony,

Mary and van Ark, Bart (eds), EU Productivity and Competitiveness: A Sectoral

Perspective. Can Europe Resume the Catching-up Process? European Commission,

Luxemburg.

Johansson, Dan (2004), “Is small beautiful? The case of Swedish IT industry”

Entrepreneurship & Regional Development, forthcoming.

Lind, Daniel (2002), “Tillväxtens drivkrafter – Produktion och användande av

informationsteknologi i svensk ekonomi”, Ekonomisk Debatt, Vol. 30, No. 7, pp. 611-

619.

Lind, Daniel (2003), ”Svensk industriproduktivitet i ett internationellt perspektiv under

fyra decennier – vad kan vi lära oss av 1990-talet?”, Ekonomisk Debatt, Vol. 31, No. 5,

pp. 39–48.

Lundgren, Kurt and Wiberg, Anders (2000), “Solowparadoxen eller den nya ekonomin?”,

Ekonomisk Debatt, Vol. 28, No. 8, pp. 747-757.

MathWorld (2004), ”MathWorld – A Wolfram Web Resource”, Wolfram Research.

Available online: http://mathworld.wolfram.com

Monnikhof, Erik and van Ark, Bart (2002), “New Estimates of Labor productivity in the

Manufacturing Sectors of Czech Republic, Hungary and Poland”. Research

Memorandum GD-50, Groningen Growth and Development Centre.

Newsweek (2000), “Shining Stockholm”, February 7.

Nordhaus, William D. (1997) “Do Real-Output and Real-Wage Measures Capture

Reality? The History of Lighting Suggests Not,” in Bresnahan, Timothy F. and Gordon,

Robert J. (eds), The Economics of New Goods, University of Chicago Press, Chicago.

OECD (2001a), OECD Productivity Manual: A Guide to the Measurement of Industry-

Level and Aggregate Productivity Growth, Directorate for Science, Technology and

Industry, Paris.

OECD (2001b), STAN Database, Paris

OECD (2002), ”Measuring the Information Economy”, OECD Working Paper, Paris.

OECD (2003), Structural Statistics for industry and Services Database, SourceOECD,

Paris.

29

Scarpetta, Stefano, Bassanini, Andrea, Pilat, Dirk and Schreyer, Paul (2000), “Economic

Growth in the OECD Area: Recent Trends at the Aggregate and Sectoral Level,”

Economics Depatement, Working Paper No. 248, OECD, Paris.

Statistics Sweden (2003a), Industrins varuproduktion (IVP), Stockholm. Available

online: http://www.scb.se

Statistics Sweden (2003b), National accounts 1993–2001, Stockholm. Available online:

http://www.scb.se

Statistics Sweden (2003c), Swedish Statistical Database; prices and consumption,

Stockholm. Available online: http://www.scb.se

Triplett (1996), “High-Tech Industry and Hedonic Price Indices”, in OECD, Industry

Productivity; International Comparisons and Measurement Issues, OECD, Paris.

United Nations Statistics Division (2000), International Standard Industrial Classification

of All Economic Activities, Third Revision, (ISIC, Rev.3). Available [online]:

http://esa.un.org/unsd/cr/family2.asp?Cl=2

30

Table 1 Values of gross output (in thousands of SEK) and quantity (number of

radio transmission apparatus) for the Swedish Radio transmission

apparatus with reception apparatus product group in 1997

Code Gross output Quantity

Radio-telegraphic and radio-telephonic

transmission apparatus, incorporating reception

apparatus, for civil aircraft

CN 85252010 1307438 n.a.

Transmission apparatus, incorporating reception

apparatus, for cellular networks “mobile

telephones”

CN 85252091 31779377 1270537

6

Transmission apparatus for radio-telephony,

radio-telegraphy, radio-broadcasting or television,

incorporating reception apparatus

CN 85252099 48538126 n.a.

Radio transmission apparatus with reception

apparatus

Prodcom 32301170 81624940 n.a.

Sources: Europroms (2001) and Statistics Sweden (2003a).

Notes: n.a. = not available. CN stands for combined nomenclature and is a classification code for industry

products that is used by Statistics Sweden. The CN code is compatible with the Prodcom classification

code.

31

Table 2 Number of matches, coverage ratios and unit value ratios for the

manufacturing industry in Sweden and Germany in 1997

Industry ISIC Number

of

matches

Percentage of output

matched

Unit value ratios SEK/EUR

Sweden Germany Laspeyres Paasche Fisher

Food products 15–16 188 79 64 10.19 9.38 9.78

Textile, clothing, leather and

footwear

17–19 91 29 2 11.10 6.99 8.81

Wood and products of wood

and cork

20 26 71 36 7.97 7.97 7.97

Paper products 21 49 54 48 9.34 7.23 8.22

Printing and publishing 22 14 39 37 12.58 10.16 11.31

Chemicals 24 78 14 16 10.89 6.20 9.59

Rubber and plastic products 25 24 23 11 9.28 9.92 9.59

Non-metallic mineral products 26 40 35 42 10.49 6.02 7.95

Basic metals 27 62 29 27 13.17 9.57 11.22

Fabricated metal products 28 33 11 10 7.76 5.30 6.41

Machinery and equipment 29 88 18 12 11.42 5.02 7.57

Office, accounting and

computing machinery

30 4 14 6 7.22 4.71 5.83

Electrical machinery and

computing

31 34 34 13 17.38 7.00 11.03

Radio, television and

communication equipment

32 10 74 23 10.06 9.33 9.69

Medical precision and optical

instruments

33 16 8 10 18.67 13.41 15.83

Motor vehicles, trailers and

semi-trailers

34 15 15 17 10.49 11.80 11.12

Other transport equipment 35 6 12 8 12.76 10.37 11.50

Manufacturing, recycling n.e.c 36 24 28 37 7.67 6.70 7.16

Total Manufacturing 802 37 25 10.75 8.24 9.41

Source: Europroms (2001) and own calculations.

Note: For an exact definition of the Laspeyres, Paasche and Fisher index see MathWorld (2004).

32

Table 3 Number of matches, coverage ratios and unit value ratios for the

manufacturing industry in Germany and the US in 1997

Industry ISIC Number

of

matches

Percentage of output

matched

Unit value ratios Dollar/EUR

US Germany Laspeyres Paasche Fisher

Food products 15–16 132 65 62 1.09 1.36 1.22

Textile, clothing, leather and

footwear

17–19 76 44 62 0.62 0.71 0.66

Wood and products of wood

and cork

20 13 52 31 0.93 1.08 1.00

Paper products 21 18 61 48 1.14 1.22 1.18

Printing and publishing 22 1 0.2 1 2.12 2.12 2.12

Chemicals 24 59 13 18 1.10 1.04 1.07

Rubber and plastic products 25 4 7 23 0.98 1.11 1.04

Non-metallic mineral products 26 23 22 29 1.26 1.42 1.34

Basic metals 27 43 71 70 1.12 1.25 1.18

Fabricated metal product 28 11 7 4 1.24 1.35 1.30

Machinery and equipment 29 53 14 15 0.95 1.04 0.99

Office, accounting and

computing machinery

30 6 38 44 1.09 1.24 1.16

Electrical machinery and

computing

31 18 15 42 0.78 1.22 0.98

Radio, television and

communication equipment

32 17 17 9 0.84 0.96 0.90

Medical precision and optical

instruments

33 16 14 3 1.52 1.72 1.62

Motor vehicles, trailers and

semi-trailers

34 5 39 29 0.87 0.90 0.88

Other transport equipment 35 1 6 3 1.88 1.88 1.88

Manufacturing, recycling n.e.c 36 20 24 16 1.01 1.14 1.08

Total Manufacturing 516 28 28 1.09 1.13 1.11

Source: Inklaar et al. (2003a).

Note: For an exact definition of the Laspeyres, Paasche and Fisher index see MathWorld (2004).

33

Table 4 Labor productivity and relative labor productivity levels (value added

in thousands of EUR per person engaged) for manufacturing in

Germany, Sweden and the US 1997 (Germany = 100)

Industry ISIC Germany Sweden US

Levels

Relative

levels

Levels Relative

levels

Levels Relative

levels

Food products 15–16 36.6 100 52.3 143 56.5 154

Textile, clothing, leather and

footwear

17–19 33.8 100 36.2 107 51.3 152

Wood and products of wood

and cork

20 41.6 100 55.6 134 41.6 100

Paper products 21 53.8 100 79.8 148 45.6 85

Printing and publishing 22 46.0 100 43.0 94 21.7 47

Chemicals 24 71.3 100 112.9 158 66.6 93

Rubber and plastic products 25 48.3 100 46.6 97 46.4 96

Non-metallic mineral products 26 51.2 100 57.7 113 38.2 75

Basic metals 27 52.1 100 55.7 107 44.1 85

Fabricated metal products 28 41.9 100 61.8 147 32.2 77

Machinery and equipment 29 49.5 100 65.4 132 50.0 101

Office, accounting and

computing machinery

30 68.4 100 98.7 144 59.0 86

Electrical machinery and

computing

31 51.5 100 38.6 75 65.5 127

Radio, television and

communication equipment

32 50.3 100 68.3 136 120.8 240

Medical precision and optical

instruments

33 38.9 100 34.7 89 36.8 95

Motor vehicles, trailers and

semi-trailers

34 59.9 100 48.3 81 109.2 182

Other transport equipment 35 50.5 100 37.9 75 33.4 66

Manufacturing, recycling n.e.c 36-37 37.0 100 25.7 69 64.1 173

Total manufacturing 15-37 48.5 100 53.8 111 62.4 129

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

34

Table 5 Labor productivity and relative labor productivity levels (value added

in thousands of EUR per person engaged) for manufacturing in

Germany, Sweden and the US 1993 (Germany = 100)

Industry ISIC Germany Sweden US

Levels Relative

levels

Levels Relative

levels

Levels Relative

levels

Food products 15–16 33.8 100 41.6 123 51.9 153

Textile, clothing, leather

and footwear

17–19 31.5 100 32.0 102 48.4 154

Wood and products of wood

and cork

20 33.8 100 41,7 123 35.4 105

Paper products 21 44.9 100 77.4 173 45.5 101

Printing and publishing 22 42.8 100 32.8 77 21.1 49

Chemicals 24 54.3 100 97.2 179 55.7 102

Rubber and plastic products 25 40.6 100 35.5 87 39.4 97

Non-metallic mineral

products

26 45.5 100 53.7 118 38.2 84

Basic metals 27 35.1 100 40.2 114 33.8 96

Fabricated metal products 28 37.8 100 53.5 141 29.8 79

Machinery and equipment 29 39.8 100 51.4 129 40.4 102

Office, accounting and

computing machinery

30 4.2 100 5.6 136 3.6 86

Electrical machinery and

computing

31 53.3 100 34.8 65 69.6 131

Radio, television and

communication equipment

32 8.3 100 5.9 70 18.6 223

Medical precision and

optical instruments

33 43.5 100 35.1 81 44.5 102

Motor vehicles, trailers and

semi-trailers

34 50.0 100 30.5 61 110.6 221

Other transport equipment 35 35.4 100 40.3 114 35.0 99

Manufacturing, recycling

n.e.c

36-37 38.9 100 17.5 45 58.5 150

Total manufacturing 15-37 41.4 100 38.4 93 51.1 121

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

Note: Calculations for the ICT producing industries are based on the US ICT deflators.

35

Table 6 Labor productivity and relative labor productivity levels (value added

in thousands of Euros per person engaged) for manufacturing in

Germany, Sweden and the US 2000 (Germany = 100)

Industry ISIC Germany Sweden US

Levels Relative

levels

Levels Relative

levels

Levels Relative

levels

Food products 15–16 37.8 100 54.4 144 48.3 128

Textile, clothing, leather and

footwear

17–19 36.4 100 39.5 108 55.1 151

Wood and products of wood

and cork

20 41.8 100 66.3 160 43.8 105

Paper products 21 56.9 100 90.8 159 40.6 71

Printing and publishing 22 50.7 100 44.0 87 23.4 46

Chemicals 24 75.8 100 139.9 185 70.1 93

Rubber and plastic products 25 47.9 100 50.7 106 46.6 97

Non-metallic mineral

products

26 53.0 100 65.4 123 35.8 68

Basic metals 27 52.7 100 59.4 113 54.2 103

Fabricated metal products 28 43.5 100 68.0 156 31.8 73

Machinery and equipment 29 50.0 100 72.6 145 50.3 100

Office, accounting and

computing machinery

30 219.1 100 319.8 146 188.9 86

Electrical machinery and

computing

31 56.1 100 44.7 80 66.7 119

Radio, television and

communication equipment

32 160.5 100 150.0 93 355.8 222

Medical precision and

optical instruments

33 37.4 100 27.5 74 35.0 94

Motor vehicles, trailers and

semi-trailers

34 48.5 100 72.8 150 126.0 260

Other transport equipment 35 52.8 100 39.8 75 35.8 68

Manufacturing, recycling

n.e.c

36-37 38.2 100 32.4 85 68.7 180

Total manufacturing 15-37 50.4 100 67.5 136 72.6 147

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

Note: Calculations for the ICT producing industries are based on the US ICT deflators.

36

Table 7 Value added deflators for the ICT producing industries (ISIC 30–33)

1994–2001

1994 1995 1996 1997 1998 1999 2000 2001

Germany

Office, accounting and

computing machinery

–0.17 -0.05 –0.01 –0.06 –0.06 –0.09 –0.13

Electric machinery and

computing

–0.002 –0.001 0.02 –0.01 0.004 0.01 –0.02

Radio, television and

communication equipment

–0.02 –0.01 –0.004 –0.003 –0.04 –0.04 –0.07

Medical, precision and

optical instruments

0.01 0.02 0.03 0.03 0.01 0.03 –0.008

Sweden

Office, accounting and

computing machinery

0.04 0.02 0.06 –0.01 0.02 0.01 –0.004 0.21

Electric machinery and

computing

0.05 0.07 0.12 0.008 –0.03 –0.07 –0.05 0.021

Radio, television and

communication equipment

–0.41 –0.51 –0.40 –0.30 –0.39 –0.39 –0.51

Medical, precision and

optical instruments

0.06 0.02 0.05 0.007 –0.03 –0.03 –0.09 0.06

US

Office, accounting and

computing machinery

–0.23 –0.29 –0.50 –0.56 –0.56 –0.51 –0.23 –0.31

Electric machinery and

computing

0.006 0.01 0.03 0.02 0.03 0.01 –0.01 0.02

Radio, television and

communication equipment

–0.14 –0.41 –0.35 –0.26 –0.41 –0.35 –0.41 –0.35

Medical, precision and

optical instruments

0.04 0.07 0.14 0.08 0.13 0.07 0.06 0.11

Sources: GGDC (2003), Statistics Sweden (2003b) and own calculations.

Note: n.a. = not available.

37

Table 8 Relative productivity level in Sweden and Germany with different

ICT deflators (Germany=100) 1993–2000

Sweden = Swedish deflators

Germany = German deflators

1993 1994 1995 1996 1997 1998 1999 2000

Office, accounting and

computing machinery

207 166 160 173 144 149 130 110

Electric machinery and

computing

83 84 88 82 75 87 86 91

Radio, television and

communication equipment

7 14 35 78 136 182 219 228

Medical, precision and optical

instruments

86 84 85 79 89 100 99 84

Sweden = Swedish deflators

Germany = US deflators

Office, accounting and

computing machinery

1842 1370 954 533 144 106 67 52

Electric machinery and

computing

78 80 84 80 75 89 88 94

Radio, television and

communication equipment

29 49 79 113 136 136 129 104

Medical, precision and optical

instruments

67 68 72 74 89 112 115 110

Sweden = US deflators

Germany = US deflators

Office, accounting and

computing machinery

136 137 144 200 144 155 151 146

Electric machinery and

computing

65 69 78 81 75 84 77 80

Radio, television and

communication equipment

70 73 95 121 136 137 126 93

Medical, precision and optical

instruments

81 84 85 81 89 97 90 74

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

38

Table 9 Recalculation of the Swedish value added price deflators for the

Radio, television and communication industry (ISIC 32)

1994 1995 1996 1997 1998 1999 2000

Gross output price deflator (1)

Electronic valves and tubes (US) –0.12 –0.33 –0.33 –0.23 –0.39 –0.28 –0.32

Telecommunication equipment (SWE)‡ –0.06 –0.10 –0.12 –0.00 –0.01 –0.05 –0.01

Radio and television receivers (SWE) ‡ –0.03 –0.06 –0.12 –0.04 –0.09 –0.13 –0.15

Shares of gross output, measured as

production at factor costs (2)

Electronic valves and tubes 0.07 0.05 0.04 0.04 0.07 0.05 0.05†

Telecommunication equipment 0.89 0.91 0.90 0.92 0.89 0.90 0.90†

Radio and television receivers 0.04 0.04 0.06 0.05 0.04 0.05 0.05†

Gross output price deflator

(3) = (1)*(2)

Radio, television and communication

equipment industry (ISIC 32)

–0.06 –0.11 –0.13 –0.01 –0.03 –0.06 –0.03

Intermediate input price deflator (4)

Electronic valves and tubes (US) 0.01 0.01 –0.07 –0.06 –0.10 –0.05 –0.03

Telecommunication equipment (US) 0.00 0.00 –0.08 –0.07 –0.12 –0.05 –0.05

Radio and television receivers (US) 0.02 0.02 –0.02 –0.02 –0.03 –0.01 –0.00

Shares of intermediate input, measured

as production at factor costs (5)

Electronic valves and tubes 0.05 0.03 0.02 0.04 0.05 0.05 0.05†

Telecommunication equipment 0.90 0.93 0.92 0.90 0.91 0.91 0.91†

Radio and television receivers 0.04 0.04 0.06 0.06 0.04 0.04 0.04†

Intermediate input price deflator

(6) = (4)*(5)

Radio, television and communication

industry equipment (ISIC 32)

–0.002 0.001 –0.08 –0.07 –0.11 –0.05 –0.04

Gross output/value added (7) ‡‡ 3.51 3.90 4.04 3.93 3.92 4.38 5.98

Intermediate input/gross output (8) ‡‡ 0.71 0.74 0.75 0.75 0.74 0.77 0.83

New value added deflators ††

(9) = (7)*[(3)–(8)*(6)]

–0.21 –0.44 –0.30 0.15 0.19 –0.11 0.02

Sources: GGDC unpublished data, Statistics Sweden (2003b), Statistics Sweden (2003c) and OECD

(2003).

Notes: ‡Gross output deflators for Telecommunication equipment and Radio and television receivers are

based on producer price indexes published by the Department of Prices and Consumption. ‡‡Results for

gross output/value added and intermediate input/gross output are average for period t and t–1. †Shares of

gross outputs and intermediate inputs for the year 2000 are assumed to be the same as for 1999. This is due

to the lack of data for the year 2000. ††The new value added deflators is derived from the formula in

equation 4.1.

39

Table 10 Recalculation of the Swedish value added price deflators for the

Radio, television and communication industry (ISIC 32)

1994 1995 1996 1997 1998 1999 2000

Gross output price deflator (1)

Electronic valves and tubes (US) –0.12 –0.33 –0.33 –0.23 –0.39 –0.28 –0.32

Telecommunication equipment (SWE)‡ –0.10 –0.13 –0.15 –0.08 –0.09 –0.09 –0.08

Radio and television receivers (SWE)‡ –0.02 –0.05 –0.11 –0.04 –0.09 –0.12 –0.14

Shares of gross output, measured as

production at factor costs (2)

Electronic valves and tubes 0.07 0.05 0.04 0.04 0.07 0.05 0.05†

Telecommunication equipment 0.89 0.91 0.90 0.92 0.89 0.90 0.90†

Radio and television receivers 0.04 0.04 0.06 0.05 0.04 0.05 0.05†

Gross output price deflator

(3) = (1)*(2)

Radio, television and communication

equipment industry (ISIC 32)

–0.10 –0.14 –0.16 –0.08 –0.11 –0.11 –0.10

Intermediate input price deflator (4)

Electronic valves and tubes (US) 0.01 0.01 –0.07 –0.06 –0.10 –0.05 –0.03

Telecommunication equipment (US) 0.00 0.00 –0.08 –0.07 –0.12 –0.05 –0.05

Radio and television receivers (US) 0.02 0.02 –0.02 –0.02 –0.03 –0.01 –0.00

Shares of intermediate input, measured

as production at factor costs (5)

Electronic valves and tubes 0.05 0.03 0.02 0.04 0.05 0.05 0.05†

Telecommunication equipment 0.90 0.93 0.92 0.90 0.91 0.91 0.91†

Radio and television receivers 0.04 0.04 0.06 0.06 0.04 0.04 0.04†

Intermediate input price deflator

(6) = (4)*(5)

Radio, television and communication

equipment industry (ISIC 32)

–0.002 0.001 –0.08 –0.07 –0.11 –0.05 –0.04

Gross output/value added (7)‡‡ 3.51 3.90 4.04 3.93 3.92 4.38 5.98

Intermediate input/gross output (8)‡‡ 0.71 0.74 0.75 0.75 0.74 0.77 0.83

New value added deflators††

(9) = (7)*[(3)–(8)*(6)]

–0.34 –0.53 –0.41 –0.12 –0.10 –0.29 –0.38

Sources: GGDC unpublished data, Statistics Sweden (2003b), Statistics Sweden (2003c) and OECD

(2003).

Notes: ‡Gross output deflators for Telecommunication equipment and Radio and television receivers are

based on producer price indexes in the National Accounts. ‡‡Results for gross output/value added and

intermediate input/gross output are average for period t and t–1. †Shares of gross outputs and intermediate

inputs for the year 2000 are assumed to be the same as for 1999. This is due to the lack of data for the year

2000. ††The new value added deflators is derived from the formula in equation 4.1.

40

Figure 1 Labor productivity levels in manufacturing, value added (in

thousands of Euros) per person engaged in Germany, Sweden and the

US 1980-2001

0

10

20

30

40

50

60

70

80

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

US

Sweden

Germany

Sources: GGDC (2003), Europroms (2001), OECD (2001b), Statistics Sweden (2003b) and own

calculations.

Note: Calculations are based on official value added deflators. The calculations for Germany before 1991

are based on figures for West Germany.

Figure 2 Scatter diagram of the difference in productivity level between

Sweden and the US in 1993 and the average Swedish labor

productivity growth rate 1993–2000

-200

-150

-100

-50

0

50

100

-0,10 0,00 0,10 0,20 0,30 0,40 0,50

Differe n ce in p roductivity level 1993

Average growth rate

Correlation coefficient: -0.61

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

41

Figure 3 Scatter diagram of the difference in productivity level between

Sweden and Germany in 1993 and the average Swedish labor

productivity growth rate 1993–2000

-80

-60

-40

-20

0

20

40

60

80

100

-0,10 0,00 0,10 0,20 0,30 0,40 0,50

Difference in productivity level 1993

Average growth rate

Correlation coefficient: -0.34

Sources: GGDC (2003), Europroms (2001), Statistics Sweden (2003b) and own calculations.

Figure 4 Gross output in the Radio, television and communication equipment

as a share of gross output in total manufacturing (current prices)

1993–2001

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

1993 1994 1995 1996 1997 1998 1999 2000 2001

Sweden

Germany

US

Sources: OECD (2001b) and Statistics Sweden (2003b).

42

Figure 5 Value added in Radio, television and communication equipment as a

share of the value added in total manufacturing (current prices) 1993–

2001

0,00

0,01

0,02

0,03

0,04

0,05

0,06

0,07

0,08

0,09

0,10

1993 1994 1995 1996 1997 1998 1999 2000 2001

German

y

US

Sweden

Sources: GGDC (2003) and Statistics Sweden (2003b).

Figure 6 Persons engaged in Radio, television and communication equipment

as a share of the persons engaged in total manufacturing (current

prices) 1993–2001

0,00

0,01

0,02

0,03

0,04

0,05

0,06

0,07

1993 1994 1995 1996 1997 1998 1999 2000 2001

Germany

US

Sweden

Sources: GGDC (2003) and Statistics Sweden (2003b).

43

Figure 7 Gross output and input price deflators for the Radio, television and

communication equipment industry (ISIC 32) 1994–2001

-0,30

-0,25

-0,20

-0,15

-0,10

-0,05

0,00

0,05

0,10

1994 1995 1996 1997 1998 1999 2000 2001

Gross output price deflator SWE

Intermediate input deflator SWE

Gross output price deflator US

Intermediate input deflator US

Sources: GGDC unpublished data and Statistics Sweden (2003b).

Figure 8 Gross output price deflators for the Electronic valves and tubes (ISIC

321) industry 1994–2001 (percent)

-0,50

-0,40

-0,30

-0,20

-0,10

0,00

0,10

0,20

1994 1995 1996 1997 1998 1999 2000 2001

Sweden, price index publis hed by the

Department of Prices and Consumption

US

Sweden, price index in the

National Accounts

Sources: GGDC unpublished data and Statistics Sweden (2003c).

44

Figure 9 Gross output price deflators for the Telecommunication equipment

(ISIC 322) industry 1994–2001 (percent)

-0,30

-0,25

-0,20

-0,15

-0,10

-0,05

0,00

1994 1995 1996 1997 1998 1999 2000 2001

S we den , pr i ce i nde x publ i s he d by th e

Department of Prices and Consumption

US

Sweden, price index in the

National Ac counts

Sources: GGDC unpublished data and Statistics Sweden (2003c).

Figure 10 Gross output price deflators for the Radio and television receivers

(ISIC 323) industry 1994–2001 (percent)

-0,16

-0,12

-0,08

-0,04

0,00

1994 1995 1996 1997 1998 1999 2000 2001

Sweden, price index published by the

Department of Prices and Consumption

US Sweden, price index in the

National Accounts

Sources: GGDC unpublished data and Statistics Sweden (2003c).

45

Figure 11 Input price deflators for Electronic valves and tubes (ISIC 321),

Telecommunication equipment (ISIC 322) and Radio and television

receivers (ISIC 323) in the US 1991–2001

-0,14

-0,12

-0,10

-0,08

-0,06

-0,04

-0,02

0,00

0,02

0,04

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Electronic v alves and

tubes (321) US

Telecommunication

equipment (322) US

Radio and television

receivers (323) US

Source: GGDC unpublished data.

Figure 12 Imports of Electronic valves and tubes (ISIC 321) as a share of total

production and imports of Electronic valves and tubes in Sweden

1995–2001 (in current prices)

0,69

0,7

0,71

0,72

0,73

0,74

0,75

0,76

1995 1996 1997 1998 1999 2000 2001

Source: Statistic Sweden (2003b)

46

Figure 13 Intermediate input/gross output ratio for the Swedish and US Radio,

television and communication equipment industry 1993–2001 (in

current prices)

0,00

0,20

0,40

0,60

0,80

1,00

1,20

1993 1994 1995 1996 1997 1998 1999 2000 2001

Sweden

US

Sources: Statistics Sweden (2003b) and OECD (2003).

Figure 14 Labor productivity growth rates in the Radio television and

communication equipment industry with official and recalculated

deflators

-0,400

-0,200

0,000

0,200

0,400

0,600

0,800

1994 1995 1996 1997 1998 1999 2000

Official deflators

Recalculat ed deflators based on price indexes by the Depart ment of P rices and Consumption

Recalculat ed deflators based on output price indexes in National Accounts

Sources: GGDC unpublished data and OECD (2003), Statistics Sweden (2003b) and Statistics Sweden

(2003c).

47

Figure 15 Labor productivity growth rates in the total manufacturing industry

with official and recalculated deflators

0,000

0,020

0,040

0,060

0,080

0,100

0,120

0,140

0,160

1994 1995 1996 1997 1998 1999 2000

Official deflato rs

Recalculated deflators based on price indexes by th e Depar tm ent of Prices and Consumption

Recalculated deflators based on output price indexes in National Accounts

Sources: GGDC unpublished data and OECD (2003), Statistics Sweden (2003b) and Statistics Sweden

(2003c).

Figure 16 Labor productivity levels in the manufacturing industry, value added

(in thousands of EUR) per person engaged in Germany, Sweden and

the US 1980-2000

0,00

20,00

40,00

60,00

80,00

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

US

Sweden, official deflators

Sweden, recalculated deflators based on

price indexes by the Department of Prices

and Consumpti on

Germany

Sources: GGDC (2003), Europroms (2001), OECD (2001b), Statistics Sweden (2003b) and own

calculations.

Note: Recalculated deflators are used for the period 1994–2000.

48

Figure 17 Value added price deflators for the Swedish Radio, television and

communication equipment industry (ISIC 32) for different types of

index formulas

-0,80

-0,70

-0,60

-0,50

-0,40

-0,30

-0,20

-0,10

0,00

1994 1995 1996 1997 1998 1999 2000

Chained Laspeyres, arithmetic c hange

Chained Laspeyres, log change

Törnqvist aggregation, log change

Sources: Statistics Sweden (2003b) and own calculations.

Note: See Appendix 2 for further details.

7. Appendix

7.1 Appendix 1: Unit value ratios

The UVR-based method was first introduced in the late 1950s, but has been further

refined by the ICOP (International Comparisons of Output and Productivity) group at the

University of Groningen under the direction of Angus Maddison and Bart van Ark (van

Ark and Timmer 2002).

Industry UVRs are based on two alternative indexes: the Laspeyres index that is using the

quantity weights of the base country and the Paasche index that uses the quantity weight

of the other country. As a first step, unit values (uv) are derived by dividing ex-factory

output values (o) by produced quantities (q) for each product i in each economy:

49

i

i

i

q

o

uv = (7.1)

The unit value can be thought of as an average price, averaged throughout the year for all

producers and across a group of nearly similar products. In a bilateral comparison broadly

defined products with similar characteristics are matched. For each matched product, the

ratio of the unit values in both countries is taken. This unit value ratio (UVR) is given by:

B

i

A

i

xu

iuv

uv

UVR = (7.2)

where, A and B are the countries being compared, B being the base country. The product

UVR indicates the relative producer price of the matched product in the two countries.

The product UVRs are used to derive an aggregate UVR for manufacturing branches and

total manufacturing. The most simple aggregation method is to weight each product UVR

by its share in total manufacturing gross output.

BA

ij

I

i

ij

BA

jUVRwUVR

j

∑

=

=

1

(7.3)

with i= 1,…, Ij the matched products in industry j; wij = oij / oj the output share of the ith

commodity in industry j; and oj = ∑=

j

I

iij

o

1the total matched value of output in industry j.

In bilateral comparisons the weights of the base country (B) or the other country (A) can

be used, which provide a Laspeyres and a Paasche type UVR respectively.36 As the

quantity weights are consistent with those that are used to derive the unit values, the

weights and units are consistent. The same procedure is repeated for the final aggregation

step from industry level to the level of total manufacturing.

36 In this paper, calculations are based on the average of the Laspeyres and Paasche indexes, i.e. the Fisher

index.

50

In a comparison between two different countries, it is not possible to match all products

in an industry. This is due to missing data of gross output value and quantity, difficulties

in finding corresponding products and the existence of country specific products. The

composition of production tends to differ much more across countries than the

composition of expenditure (van Ark and Timmer 2002).

7.2 Appendix 2: ICT deflators

Even though the Swedish and the US National Accounts are based on double deflation

there are still differences in the way value added is measured. One important difference is

that the US uses a Törnqvist price index to derive a Törnqvist value added volume index

while Sweden uses a chained Paasche price index to derive a chained Laspeyres volume

index, where the year t–1 is used as the base year.

A Törnqvist volume index is a weighted geometric average of the quantity relatives using

arithmetic averages of the value shares in the two periods as weights.

2

)(

1

0

0

)/(

i

t

iss

n

i

i

t

iT qqQ

+

=

∏

= (7.5)

where 0

i

s denotes the share of the value of product i in the total output of goods and

services in period 0: that is, 0000 /iiii qpqp ∑.

A Laspeyres volume index is a weighted arithmetic average of quantity relatives using

the values of the earlier period as weights.

0

1

1

00

1

0

)/( i

o

i

n

i

t

i

n

i

ii

n

i

t

ii

Lsqq

qp

qp

Q∑

∑

∑

=

=

=

≡= (7.6)

where 0

i

s denotes the share of the value of product i in the total output of goods and

services in period 0: that is, 0000 /iiii qpqp ∑.

51

The rational for using a certain index formula is based on theoretical arguments that will

not be discussed in this paper.37 However, from the definitions above there appear to be

two major differences between the chained Laspeyres index and the Törnqvist index. One

difference is that the Laspeyres index is based on the arithmetic average, while the

Törnqvist index is based on the geometric average. Moreover, the Törnqvist price index

uses the average of the two periods t and t–1 as weights while the Laspeyres index only

uses the period t–1 as weights.

The logarithm of the Törnqvist index can be expressed in the following way:













+= ∑

=

0

0

1

ln)(

2

1

ln

i

t

i

i

t

i

n

i

Tq

q

ssQ (7.8)

In order to approximate the Swedish data based on the Laspeyres index to the Törnqvist

index, I use the logarithmic change of the values derived by the Laspeyres volume index.

This gives the log change between two years instead of the arithmetic change. Moreover,

I also use the average of the Swedish value added and intermediate input weights for the

period t–1 and t. Since I do not have access to the weights of every product for the

intermediate input and output it is not possible to change the weights for each product.

Nonetheless, for the total gross output/value added ratio as well as for the intermediate

input/gross output ratio it is possible to use the average weights of the two years t-1 and t

(see section 4.2).

Figure 17 shows the different results from calculating the value added deflator for the

Radio, television and communication equipment by using arithmetic mean and weights

with year t–1 as the base year, log change and weights with the year t–1 as a base, and

log change with the average of the years t and t–1 as base years. Since the latter is the

closest approximation to the Törnqvist price index it will be used for all calculations of

value added price deflators. Moreover, one of the reasons that the value added price

deflators based on arithmetic mean differs widely from those based on logarithmic

37 For a thorough discussion of the theoretical reasons to use certain index formula, see IMF (2003).

52

change is that there are extremely high growth rates of production value and intermediate

input 1993–2000. If the growth rates had been lower than 10 percent per year the

difference would have been negligible.

Building a Bridge: Knowledge Sharing Flows into Entrepreneurial Ecosystems

Article

Full-text available

Aug 2022

Entrepreneurial ecosystems remain under-theorised and conceptually fragmented, making it challenging to comprehend their disposition and performance in the business process. Accordingly, in this research, we explored how knowledge sharing flows through entrepreneurial ecosystems to make analyses and trials to assess new ventures' creation, continuity, and development opportunities. We carried out a systematic literature review on the Web of Science database. The analysis was carried out in two stages: (i) content analysis using NVivo software and (ii) statistical processing and clustering with the support of VOSviewer and Bibliometrix software. Moreover, we reviewed entrepreneurial literature and proposed conceptual model mapping relations through all main actors and knowledge flow in ecosystems. Our findings suggest the knowledge path in the near field sharing mechanisms resulting in a new conception of traditional structures and relations used to judge and decide how to assess opportunities for new ventures' opening, maintenance, and growth. This study contributes to entrepreneurial literature, demonstrating knowledge sharing flow through entrepreneurial ecosystems, considering an embracing, dynamic, and multilevel approach. Furthermore, it highlights political and social contributions to include new emergent perspectives: resource scarcity and structural and institutional gaps. This representation is the first knowledge management model applied to different economies and areas, respecting their singularities.

Effects of ICT Investment and Usage on Economic Growth in MENA Countries: Does Governance Matter?

Thesis

Full-text available

Mar 2022

Mohammed Abu Alfoul

Information and Communications Technology (ICT) is considered an important component in improving the efficiency of various economic activities. However, this is often interrelated with a country’s quality of governance, particularly in developing countries, such as the Middle East and North Africa (MENA). The MENA countries have relatively well-developed ICT coverage, but they suffer from poor quality of governance. Therefore, this region provides a setting to examine new research questions, such as ‘Do ICT investment and usage affect economic growth in MENA countries?’ and ‘Does the impact of ICT investment and usage on economic growth depend on governance quality?’. The main objective of this study is to investigate the long- and short-run impact of ICT investment and usage on economic growth on the one hand, and the moderating role of the quality of governance on the association between these variables on the other hand. The theoretical foundations of this study can be found in endogenous growth theory as proposed by Barro (1996b) and Romer (1990). This study contributes to the ICT literature by emphasising the countries’ quality of governance in the association between technology and economic growth. More importantly, this study is the first to use all the Worldwide Governance Indicators (WGI) developed by the World Bank as moderator variables in the association between ICT investment, usage, and economic growth. This study applies the panel ARDL method and uses data for 16 MENA countries between 1995 and 2018. The results suggest that ICT usage alone does play a significant role in contributing to better economic growth, whereas ICT investment has an insignificant impact on economic growth in MENA region. Interestingly, improvements in quality of governance increase the effectiveness of ICT investment and usage in the economic growth of the MENA countries. The results are of importance for policymakers interested in improving the effectiveness of ICT’s contribution to economic growth, by exposing the potential impacts of the MENA region’s governance indicators on ICT investment and usage. It is important to mention that the MENA countries’ policymakers face the challenge of slow economic growth, and they need to formulate policies aimed at increasing ICT investment and usage. Thus, they also need to develop policies that enhance governance quality, as without effective governance, no significant improvements in economic growth can be expected from ICT investment and usage in the MENA region. The MENA countries’ policymakers should guide ICT investment and usage to achieve greater labour productivity to increase and accelerate economic growth.

The ICT Sector in Economic Development of the Countries of Eastern Europe: a Comparative Analysis

Article

Full-text available

Jan 2022

The purpose of the article is to study the impact of the ICT sector on economic development of countries based on the comparative analysis of this sector development in some Eastern European countries. Within the article, economic development of the outlined countries in 2010-2019 was studied and analyzed. The analysis of the impact of the ICT sector on the GDP formation allowed to single out certain groups of countries under this indicator and to identify the characteristics that are inherent to them. Using the correlation-regression analysis made it possible to analyze the ICT impact on economies development and the Czech Republic. Authors paid a special attention to the study of the influence of various factors on the ICT sector development. Accordingly, an analytical study of the dependence of the ICT sector weight in the GDP of the countries on the following parameters: enterprises that employ ICT specialists; enterprises that provided training to develop/upgrade ICT skills of their personnel; percentage of the ICT personnel in total employment; using Internet for Internet banking, % of individuals; enterprises who have ERP software package to share information between different functional areas; enterprises selling online (at least 1% of turnover), % of enterprises; online purchase in the last 12 months, % of individuals; enterprises having received orders via computer mediated networks, % of enterprises. Within the article, the features of the ICT sector development in the COVID-19 context are examined, and it is analyzed how the pandemic has affected the development of this sector in long and short terms. The study showed that the ICT sector today already plays a key role in the development of the national economies. Countries where the sector is developing faster show better performance and economic development.

Formación profesional y procesos de innovación a nivel regional

Article

Aug 2016
REV VENEZ GERENC

Respecto al modelo formativo que debe acompañar las apremiantes transformaciones que requiere el actual modelo económico se señala el fomento de la innovación, si bien constriñéndola a la relación entre universidad/empresa. Una combinación ineludible, sin duda, pero que obvia otra relación, la que se establece entre Centros de Formación Profesional y pymes. El objetivo de este artículo es explicar esta última relación apoyándose en la teoría de los sistemas regionales de innovación. La metodología que se ha utilizado ha sido el análisis comparado de las propuestas y perspectivas sobre sistemas, modos de innovación y formación profesional a través del análisis e interpretación de textos. En este artículo, como conclusión, se resalta que los procesos de innovación que impactan positivamente en el nivel de competitividad y en la tasa de crecimiento no son generados exclusivamente por instituciones de innovación y desarrollo sino que también los Centros de formación profesional pueden realizar una importante contribución a la innovación en las pymes.

Innovation in the periphery: refresh!

Article

Nov 2023

UMA REVISÃO DA LITERATURA SOBRE ADOLESCENTES TRABALHADORES, SUA CONJUNTURA SOCIAL E ALGUNS IMPACTOS DE SUA ATIVIDADE LABORAL PRECOCE

Chapter

May 2022

O livro “As novas perspectivas das ciências sociais vol.01”, publicado pela Reflexão Acadêmica Editora, coletânea que une vinte e um capítulos, apresenta trabalhos relacionados com temas diversos da área de ciências sociais. Assim, o livro apresenta seguintes temáticas que envolvem, por exemplo, como as organizações sociais nos dias de hoje, por meio de trabalhos voluntários, conseguem acrescentar algo à sociedade, e como os colaboradores se sentem ao fazer parte de um movimento organizacional. Então, é apresentado um estudo sobre as principais abordagens e formas de manutenção da água, fonte importante para a vida de tantas espécies. Nos entrecruzamentos entre sociedade, sustentabilidade e gestão de negócios, será apresentado um estudo sobre a implantação de mudanças e culturas em uma empresa localizada em Palmas-TO, afim de que sejam adotadas políticas empresariais, governamentais, práticas de sustentabilidades, responsabilidade socioambiental, tendências e demandas globais e que sejam possíveis de serem acompanhadas, mas sem perder a competitividade nos negócios. Temas como os padrões de beleza impostos pela sociedade hoje em dia também são abordados, assim como o impacto que causam na vida de mulheres de meia-idade, e como são necessários buscar caminhos que possam ser alternativas e se encaixar harmoniosamente, entre o ideal de beleza atualmente e como pode ser levado no decorrer da vida. Os trabalhos apresentados nesta coletânea permitem, assim, analisarmos sobre questões sociais enfrentadas atualmente, e como podemos lidar com os desafios encontrados. Dessa forma agradecemos todos os autores pelo esforço e dedicação colocados em seus trabalhos e esperamos poder contribuir com a comunidade científica que se interessa por temas relacionados com as ciências sociais e que o livro auxilie nas temáticas discutidas.

Estimating effects of ICT intensity on productivity, employment and output in South Africa: an industry-level analysis

Article

Mar 2021

This article aims to estimate the effects of ICT intensity on labor productivity, employment and output of agro-processing industries. To achieve this, the ICT intensity index is applied to rank industries into ‘more ICT-intensive’ and ‘less ICT-intensive’ groups. Thereafter, the annual growth rates of labor productivity, employment and output were calculated. Ultimately, the effects of ICT intensity were examined using Pooled Mean Group estimation, the Toda and Yamamoto Granger Non-Causality Test, and the Impulse Response Function and Variance Decomposition analyses. The findings suggest that ICT intensity yields higher positive and significant effects on the growth of the more ICT-intensive industries. Evidence of a causal relationship was detected for the more ICT-intensive industries. The findings further proved that ICT intensity contributed more to the forecast error variance in the growth of the more ICT-intensive industries. Overall, this article provides evidence of ICT-led growth for industries that use ICT most intensively.

Estimating effects of information and communication technology (ICT) on the productivity of manufacturing industries in South Africa

Article

Feb 2020

This paper serves to examine whether the growth in labour productivity (LP) in South Africa’s manufacturing sector, following policy reforms after 1994, can be attributed to ICT use. To achieve this, we examine the link between ICT intensity and LP growth of 23 manufacturing industries for the period 1970–2016 and sub-periods 1970–1995 and 1996–2016. The industries are disaggregated into two groups, which are ‘more ICT-intensive’ and ‘less ICT-intensive’, using the ICT intensity index. Four dummy variable regression models are applied to test for the relationship between ICT intensity of industries and LP growth. The findings suggest that LP of the more ICT-intensive industries accelerated more than that of their counterparts. The results underscore the need for policy measures to increase ICT use with the aim of improving LP performance of industries.

Tools for digitalization of economic processes for supporting management decision-making in the region

Article

Full-text available

Aug 2019

The article proposes a solution to one of the key problems – the use of information technology in the management of territorial processes based on the development of tools for digitalization of economic processes on the basis of the software product to identify the structural elements of the economic potential of the regions of Russia, including the features of such regions as the Arctic. As a tool for the implementation of this approach, the VBA programming language built into the Excel spreadsheet is used, which allows flexible use of a standard software product for solving problems at the macroeconomic level. The aim of the study is to develop tools for digitalization of economic processes on the basis of a computer program to determine the extreme structural elements of the growth of the economic potential of the regions to support management decision-making. The systematic approach, methods of statistical and economic analysis, including horizontal and vertical analysis, allowed the authors to assess the dynamics of extreme values of the level of development of economic sectors in the regions of Russia and to develop recommendations for the development of the Federal districts and regions. The directions of further researches consist in application of multidimensional modeling of data for collecting and automatic processing of statistical information on a condition of subjects of the Arctic region.

Role of Information and Communications Technology (ICT) in Enhancing Service Sector Productivity in Palestine: An International Perspective

Article

Full-text available

Jan 2019

This article discusses the productivity of the Information and Communication Technology (ICT) sector using cross-sectional data from 793 service firms in Palestine. The authors have examined the impact of ICT growth on service sector productivity in Palestine using a set of indicators for ICT including internet usage, e-commerce, networks, websites, and use of “smart” phones. They find that using ICT (mainly Internet) in commerce (e-commerce) is one of the most important levers of labor productivity among service firms. Service firms that are less ICT-intensive are less productive than more ICT-intensive firms; moreover, the use of mobile phones for services other than send-and-receive calls, highly improves the labor productivity of service firms. Conversely, using a website and computer network does not positively affect the labor productivity. Regarding geographical differences in labor productivity, the analysis shows that firms in Jerusalem are characterized by higher productivity than firms in the West Bank, while firms in Gaza have a lower productivity compared to firms in the West Bank.

Industrial Revolution and Scientific and Technological Progress

Article

Full-text available

Nov 1997

Rainer Fremdling

This article is the original version of a chapter for the UNESCO-History of Humanity published slightly modified in 2008. Different views on the concept and spread of the industrial revolution, which took place from the late 18th century onwards, are dealt with. By way of example the revolutionary character of technological change and the search for new sources of energy is described and analysed for some major innovations and industries, namely the steam engine, the iron and steel industry and the use of electricity. Furthermore the question is discussed as to what extent sciences were related to technical progress. In this context the emergence of higher technical education at universities is put forward. Some brief remarks on the role of agriculture and the increasing economic ties among countries during the 19th century conclude this paper.

Foreign Trade–Transfer-Adaptation: The British Iron Making Technology on the Continent (Belgium and France)

Data

Full-text available

Aug 2002

Rainer Fremdling

READER ON PRODUCTIVITY LEVELS

Article

Full-text available

Jan 2006

Catching Up in Industrialized Countries: A Disaggregated Study

Article

Full-text available

Jul 1994

In this study we test whether catching up, the hypothesis that there is technological spillover from leaders to followers, is still important among industrialized countries. Since the USA is no longer the technological leader in many industries and since catching up, if it still exists, may not operate uniformly across different industries, a disaggregated study is more appropriate. A testable model is developed and a number of tests for the existence of catching up are performed. A major improvement on previous tests is that the level of technology is measured in terms of total factor productivity. The two major conclusions, which are quite robust, are that after 1970 there is no catching-up effect left in the tradables sector, while catching up is found for industries in the nontradables sector.

'Changing Gear': Productivity, ICT and Service Industries in Europe and the United States

Chapter

Oct 2003

Economic Backwardness in Historical Perspective

Article

Jul 1963

Measuring the Information Economy

Article

Jan 2002

OECD

Productivity and competitiveness in the EU and the US

Article

Jan 2003

ICT Production and ICT Use: What Role in Aggregate Productivity Growth?

Chapter

Mar 2004

This paper examines the roles of the ICT-producing sector and of key ICT-using industries in overall productivity growth in OECD countries. The ICT manufacturing sector, in particular, has been characterised by very high rates of productivity growth in many countries and provides a large contribution to labour productivity growth in Finland, Ireland and Korea. In a few countries, notably the United States and Australia, certain ICT-using services have also experienced an aboveaverage pick-up in productivity growth in the second half of the 1990s. Differences in the measurement of productivity in ICT-producing and -using industries across countries complicate the cross-country analysis.

The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?

Article

Sep 2000
J ECON PERSPECT

The growth of U.S. labor productivity rebounded in the second half of the 1990s, after nearly a quarter century of sluggish gains. We assess the contribution of information technology to this rebound, using the same neoclassical framework as in our earlier work. We find that a surge in the use of information technology capital and faster efficiency gains in the production of computers account for about two-thirds of the speed-up in productivity growth between the first and second halves of the 1990s. Thus, to answer the question posed in the title of the paper, information technology largely is the story.

Article

Full-text available

ICT Production and Productivity in Sweden and Finland, 1975-2004

February 2008 · Ekonomiska Samfundets Tidskrift

Daniel Lind

This article compares the development of labour productivity in the Swedish and the Finnish business sectors and the role of the information and communication technology (ICT) sector in this process. The results show that the Finnish productivity level has been converging towards the Swedish level, but that there is still a significant difference. This trend has coincided with the growing ... [Show full abstract] importance of the ICT sector, especially since the mid 1990s. Due to higher productivity and employment growth, the Finnish ICT sector has contributed to this convergence. This is explained by the electrical engineering industry. The Nokia effect has been stronger than the Ericsson effect.

Article

Can double deflation explain the ICT growth miracle?

November 2013 · Economics Letters

Harald Edquist

It is unclear why productivity growth in the information and communication technology (ICT) industry differs considerably among countries. This paper shows that, for Sweden, it is primarily the use of the double deflation method that has created the apparent ICE-miracle.

Article

Full-text available

Parallel development? Productivity growth following the diffusion of electric motors and ICT

March 2011 · Scandinavian Economic History Review

Harald Edquist

This paper investigates labor productivity growth and the contribution to labor productivity growth in Swedish manufacturing during the diffusion of electric motors and ICT. The paper distinguishes between technology-producing, intensive and less intensive technology-using industries during these technological breakthroughs. The results show that labor productivity growth and the overall ... [Show full abstract] contribution to labor productivity growth was considerably higher in technology-producing industries following the diffusion of ICT. Moreover, the results presented here show no evidence that industries that were early adopters of electric motors and ICT, on average, would have contributed more to productivity growth in Swedish manufacturing.

Article

Full-text available

Does Hedonic Price Indexing Change Our Interpretation of Economic History? Evidence from Swedish Ele...

May 2008 · The Economic History Review

Harald Edquist

Rapid price decreases for ICT-products in the 1990s have been largely attributed to the introduction of hedonic price indexes. Would hedonic price indexing also have large effects on measured price and productivity during other technological breakthroughs? This paper investigates the impact of hedonic and matched model methods on historical data for electric motors in Sweden 1900–35. The results ... [Show full abstract] show that during the productivity boom of the 1920s, the constant prices for electric motors decreased by 9.7 and 8.1 percent per year depending on whether hedonic or matched model price indexes were used. This indicates high productivity growth in the industry producing electric motors 1920–29. In contrast to Sweden, the US annual total factor productivity was only, according to current best estimates, 3.5 percent in Electric machinery compared to 5.3 percent in manufacturing 1920–29. However, hedonic price indexes were not used to calculate US productivity. Moreover, in comparison to the matched model, the hedonic price index on average overestimates price decreases when prices are decreasing and overestimates price increases when prices are increasing. However, the total effect of the two different price indexes remains approximately the same in 1900–35. Finally, it is shown that the price decreases for electric motors in the 1920s are not in par with the price decreases for ICT-equipment in the 1990s, even if hedonic indexing is used.

Article

Full-text available

Can Investment in Intangibles Explain the Swedish Productivity Boom in the 1990s?

January 2009 · Review of Income and Wealth

Harald Edquist

After a severe crisis in the early 1990s, the Swedish economy experienced a boom in productivity growth. According to economists there have been primarily three explanations for the fast productivity growth in 1995–2004: Market reforms, recovery from the crisis and the impact of information and communication technology (ICT). This paper offers an alternative view by recognizing that firms make ... [Show full abstract] substantial investment in intangible assets such as R&D, design, advertising etc. These investments are not classified as investment in the National Accounts, where only tangible assets are defined as investment. This paper provides estimates of investment in intangible assets and uses the growth accounting framework to analyze the Swedish productivity boom. The results show that investment in intangibles was approximately 246 bn SEK in 2004 or 9 percent of GDP. Moreover, intangible capital accounted for almost 50 percent of labor productivity growth in the Swedish business sector 1995–2004. Thus, investment in intangibles was an important source to the Swedish productivity boom in 1995–2004.