Transcript down - Asia KLEMS

Improving Service Sector Productivity:
Implications from Studies in Japan
August 2015
Masayuki Morikawa (RIETI)
Today’s Topics
 Policy initiatives to improve service sector
productivity in Japan and the backgrounds
 Viewpoints to analyze service sector
productivity: “Simultaneity of production and
consumption”
 Some caution: Difficulty in measuring
service quality & possible underestimation of
productivity
2
The “Third Arrow” of Abenomics
•
•
•
The Japanese Government (the “Third Arrow” of Abenomics) set an ambitious target of 2%
annual real GDP growth rate.
During the last 20 years, mean actual GDP growth rate (0.87%) coincides with the mean
potential growth rate (0.84%). The current potential growth rate is estimated to be around 0.5%.
It is obvious that significantly enhancing growth potential is needed to achieve the target.
3
Decline in Workforce of Japan (Projection)
•
•
•
Japan is experiencing rapid aging of population and low fertility rate. As a result, workforce in
Japan is projected to decline.
Under the status quo assumption (no policy changes), total workforce continues to decrease by
0.8% a year (2012-2030).
This means the contribution of labor inputs to potential growth rate will be negative 0.5-0.6%.
(Source) Projection by the Japan Institute of Labor Policy and Training (JILPT). The percentage figures are annualized rates.
4
Growth Accounting (JIP Database)
•
•
•
Potential growth rate is determined by the contributions of labor, capital deepening, and total
factor productivity (TFP).
According to the growth accounting for the Japanese economy, in parallel with the decline in
labor input, the TFP growth has dropped significantly since 1990s, although some recovery of
the TFP growth can be found in 2000s.
Under the declining trend in workforce, to achieve 2% real GDP growth, TFP growth rate
should be around 2%. It is not an easy task, and improving service sector productivity is a key.
5
(Note) The contribution of TFP of this figure includes quality changes in labor and capital.
Service Sector Productivity: Recent Policy Developments
 In Japan, improving service sector productivity has been frequently
advocated policy agenda since 1970s after the end of high-growth era.
Recent growth strategies picked up this issue again.
 “Japan Revitalization Strategy 2014”
 The Strategy states that “Japanese companies have lower productivity than their
Western counterparts. Particularly, the productivity of the nonmanufacturing
sector including the services industry is seriously low, dragging down the
Japanese economy.”
 “In order to raise productivity in all industries including services and win amid
severe international competition, Japanese companies must achieve global-level
earnings and productivity.”
 “Service Industries’ Challenge Program” (2015, Cabinet Office)
 The Program sets a target of labor productivity growth rate of the service sector
to be 2.0% by 2020 (currently 0.8%).
 Policies across industries: further use of ICT, globalization, upgrading human
resources, making cities compact, etc.
 Industry specific policies: hotel and lodging, transportation, restaurants,
healthcare, nursing care, wholesale & retail.
6
International Comparison of TFP Growth
•
•
According to the EUKLEMS Database, TFP growth rate of service sector is generally lower
than that of manufacturing sector. Improving service sector productivity is expected in every
country.
Different from the general perception, TFP growth performance of Japan’s manufacturing sector
(excluding ICT products) is not necessarily better than other advanced countries in this period.
ＴＦＰ, 1990-2008
Japan
U.S.A.
Europe
All industries
0.1%
0.3%
0.7%
ICT Industries
5.1%
5.3%
3.0%
Manufacturing
-0.4%
0.8%
1.6%
0.2%
0.2%
0.2%
-0.2%
-0.6%
0.1%
Market services
Non-market services
(Source) EUKLEMS Database. The figures for Europe are the simple averages of UK, Germany, and France.
7
Comparison of TFP “Level” between the US and Japan
•
•
According to estimation by Jorgenson, Nomura, and Samuels (2015), TFP level of Japan’s
service sector relative to the U.S. is very different by industries.
The productivity level of wholesale and retail is around 70% of the U.S. counterpart,
suggesting a room to improve the productivity performance.
(Source) Jorgenson, Dale W., Koji Nomura, and Jon D. Samuels (2015), “A Half Century of Trans-Pacific Competition: Price Level Indices and
Productivity Gaps for Japanese and U.S. Industries,” RIETI Discussion Paper, 15-E-054.
8
Characteristics of Services:
Simultaneity of Production and Consumption
•
•
•
Typical services have distinct characteristics that services are produced and consumed at the
same time and the same place.
While manufacturing firms use inventory as a measure to smooth production, it is generally
impossible in the service industries.
As a result, capital utilization rate is an important determinant of “measured” productivity
in the service sector. For example, room occupancy rate (hotels & accommodations) and
load factor (taxi, railroad, airline) are the key performance indicators of the industries.
9
Short Distance to Customers
• The spatial boundary of the market for services is relatively narrow, because of the
characteristics of simultaneous production and consumption.
• According to a detailed firm-level transaction data, the mean distance to customers is different
by industry. Manufacturing firms sell their products to distant firms, but service firms sell their
services to nearly located firms.
(Note) Calculated from the TSR database of more than 275 thousands firms located in Japan. Domestic B2B transactions only.
I thank Dr. Yukiko Saito (RIETI) for providing the calculation result.
10
Less International Competition
• Compared with manufacturing firms, service firms face less international competition.
According to our survey, nearly 40 % of firms in the non-manufacturing industry state that
their businesses are unrelated to international competition.
(Source) “Survey on the Outlook of the Japanese Economy and Economic Policy” (RIETI). Calculated from a survey for more
than three thousands firms.
11
Spatial Simultaneity: Density and Productivity
•
•
•
We estimate the TFP of service establishments using micro data and analyze the effects of
population density.
Agglomeration economy is stronger for service and retail industries than for manufacturing
industry. The productivity of service industry increases more than 10%, when the population
density of the city doubles.
The result suggests that making cities “compact” is an effective policy to enhance productivity
of service industries, especially under the trend decrease in national population.
Compact city (Toyama city)
(Note) The bars indicate the effects of doubling population density on TFP.
(Source) Morikawa, Masayuki (2011), "Economies of Density and Productivity in Service Industries: An Analysis of Personal-Service Industries Based on
Establishment-Level Data," Review of Economics and Statistics, 93(1), 179-192.
12
Urban Wage Premium by Industry
•
•
•
The previous analysis rests upon a small number of personal service industries.
We estimate elasticity of wages with respect to city population density by industry. Wage data
has advantages that it covers the whole private sector and that we can control skill-mix of
workers such as education and tenure.
We found significant agglomeration wage premium for both manufacturing and service sector
workers, but the size of the premiums are relatively large in service industries.
(Note) Estimated from micro data of the Basic Survey on Wage Structure. Age, gender, education, tenure, and firm size are controlled.
“Standard worker” is those who continue working with a firm after graduation from school.
(Source) Morikawa, Masayuki (2011), “Urban Density, Human Capital, and Productivity: An Empirical Analysis Using Wage Data,”
RIETI Discussion Paper, 11-e-060.
13
Demand Fluctuation and Productivity
•
•
According to a study on personal services industries, demand fluctuation is negatively related
to the “measured” TFP at the establishment level. One standard deviation larger fluctuation is
associated with 8-14% lower TFP.
This result suggests that smoothing demand for services contribute to the productivity of the
service sector. Introduction and diffusion of flexible working arrangements and dispersing
annual leave are the examples to smooth demand for services.
(Note) The service industries included in the estimation are movie theater, golf course, tennis court, bowling allay, fitness club, and golf driving range.
(Source) Morikawa, Masayuki (2012), “Demand Fluctuations and Productivity of Service Industries,” Economics Letters, 117(1), 256-258.
14
Sales Volatility, Non-standard Employment, and TFP
•
•
•
Among firms facing with volatile sales, employing non-standard workers positively contribute
to their TFP. This relationship is observed both in manufacturing and service industries.
This result suggests a trade-off between the productivity enhancement and the security of
employment.
A desirable policy mix is to establish safety net and to provide opportunity of training for nonstandard workers on the one hand, and to make labor market flexible on the other hand.
(Note) The figure indicate the effects of 10 % points increase in temporary agency workers on the TFP of firms whose sales volatility is one standard
deviation higher than the mean. The data come from the Basic Survey of Japanese Business Structure and Activities.
(Source) Morikawa, Masayuki (2010), “Volatility, Nonstandard Employment, and Productivity,” RIETI Discussion Paper, 10-E-025.
15
ＩCＴ and Service Sector Productivity: Capacity Utilization
•
In the service industries, the effect of using ICT on productivity at the firm level is
often generated through its positive effect on capacity utilization. Some studies
empirically confirm this relationship.
– Hubbard (2003: AER): Use of advanced on-board computers (OBCs) have increased
capacity utilization among adopting trucks by 13 percent. Better matches between truck
capacity and demands boost capacity utilization and productivity in the industry.
– Dana and Orlov (2014: AEJ Micro): Internet penetration dramatically increased airline
capacity utilization rate in the U.S. A doubling in Internet penetration increases load
factors by 6 .1 percent.
•
In Japan, QB house – a new haircut company – is a good example using ICT to
increase capacity utilization rate.
16
Tourists from Overseas and Occupancy Rate of Accommodations
•
•
•
Recent increase in foreign tourists to Japan has contributed to the rise in room occupancy rate
of hotel and accommodation facilities.
Since foreign tourists’ seasonal and weekly patterns of lodging are different from Japanese
natives’, increase in tourists from abroad has an effect of smoothing peak demand.
In addition to the direct effect of the number of guests, this demand smoothing effect
significantly contributes to the (measured) productivity of hotels and lodging sector.
17
(Note) Estimated from the Accommodation Survey (Japan Tourism Agency).
Difficulty in TFP Measurement
•
•
•
However, it should be noted that accurate measurement of service sector productivity is very difficult.
Since long-term productivity decline (“technological deterioration”) is unlikely, negative TFP seems
to be a result of measurement error. Specialists in productivity analysis suggest a “thought
experiment” that the productivity trend is be replaced by zero for industries that were observed to be
negative (Corrado and Slifman, 1999).
According to the JIP Database, negative long-run (40 years) TFP growth is often found in service
industries. If we assume TFP of these industries to be flat (=zero productivity growth), the difference
between manufacturing and service sectors disappears. This pattern is not specific to Japan.
(Note) Calculated from the JIP database 2014 for the years between 1970 and 2011. The corrected TFP growth rates are calculated by replacing negative
TFPs by zero.
18
Difficult to Evaluate Quality of Services
•
•
•
Measuring quality of services is notoriously difficult. This difficulty causes underestimation (or
overestimation) of the true productivity growth rate.
International comparison of productivity level in the service sector is also very difficult,
because international price comparison of services often do not capture cross-country
differences in the quality of services.
For example, accurate operation of public transportation and hospitality in accommodations are
not taken into accounted in calculating productivity.
New Hokuriku Shinkansen
Japanese Ryokan
19
Possible Underestimation: Quality Change of Services
•
•
In the U.S., upward bias of CPI (= underestimation of quality upgrading) has long been
discussed. Significant upward biases have been found in services.
The situation is similar in Japan. Quality adjustment in constructing price index is difficult for
services than goods as documented by the BOJ (42% vs. 65%).
Quality Change & New-item Bias in the U.S. CPI (%)
Boskin Report Lebow and
(1996)
Rudd (2003)
Airfares
0.0
0.5
Medical care services
3.0
2.5
Telecommunication
1.5
0.8
Personal financial services
2.0
1.0
(Source) Lebow, David E. and Jeremy B. Rudd (2003), “Measurement
Error in the Consumer Price Index: Where Do We Stand?” Journal of
Economic Literature, 41(1), 159-201.
(Source) Research and Statistics Department, Bank of
Japan (2009).
20
A Comparison of Service Quality in Japan and the U.S.
•
•
•
We conduct a survey to compare the quality of services in Japan and the U.S. The survey
asked Japanese and American people those who have experience to live in both countries.
According to the survey result, both Japanese and U.S. respondents evaluate that the quality
of services in Japan is generally higher than the U.S.
The result suggests the TFP levels of these services in Japan may be underestimated by 525%.
(Source) Japan Productivity Center (2009), ”Comparison of Quality in Services between Japan and the U.S.”.
21
Conclusion
• The “Third Arrow” of Abenomics stresses the importance of improving
service sector productivity.
• In order to plan effective policy measures, we should understand the nature
of services ― “simultaneous production and consumption.” As a result,
capacity utilization rate is a key to improve “measured” productivity of
service industries.
• In addition, spatial boundary of the market of services is generally narrow:
Inter-regional and international competition is often limited. Liberalization
and facilitation of services trade is important.
• Service sector productivity is deeply linked to the economic and social
institutions. Fundamental laws and regulations, e.g., urban planning
regulations and labor market rules and regulations, greatly affect the
productivity performance. Overall structural reform is required.
• However, we should be careful that it is difficult to measure service
productivity accurately. Efforts to collect good data and the progress of
estimation methods are necessary. The Asia KLEMS is an important
initiative to move forward to this direction.
22
Thank you for your attention!
Please visit RIETI’s website.
http://www.rieti.go.jp/en/