CITIES AND REGIONS: HOW THEY

AFFECT NATIONAL PRODUCTIVITY?

Global Forum on ProductivityINTERNATIONAL CONFERENCE PRODUCTIVITY AND

INCLUSIVE GROWTHPanel 2A - Cities & agglomeration – How do they affect

productivity?Santiago, 5-6 December 2016

Joaquim Oliveira Martins

OECD Public Governance Directorate

Urbanisation and development

3

Urbanisation goes along with development, but it is only a necessary condition

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

Rea

l GD

P p

er C

apit

a (

as %

of

US

GD

P/C

apit

a)

Level of Urbanization

Brazil

China

Colombia

Japan

Peru

Thailand

Korea

Rest ofthe World

4

Urbanisation and income convergence fit approximately an exponential relation…

5

… but this relation is much less pronounced for Latin America than for Asian countries

y = 4.4701x - 0.5571R² = 0.8369

y = 2.5795x - 0.1196R² = 0.4867

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

0 0.2 0.4 0.6 0.8 1Log

of th

e re

al G

DP

per

Capi

ta (

as %

of U

S G

DP/

Capi

ta)

Level of Urbanization

Log of the real GDP per Capita (as % of US GDP/Capita) and Level of Urbanization

Latin America & Caribbean, East Asia

East Asia

Latin America andthe Carribeans

Linear (East Asia)

Linear (LatinAmerica and theCarribeans )

What makes cities more productive?

Reviews by Rosenthal and Strange (2004), Duranton and Puga (2004) and Puga (2010); concepts already present in Marshall (1890):

I. Sharing facilities, inputs, gains from specialisationfirms may face lower costs for specialised non-traded inputs that are shared locally in a geographical cluster.

II. Thicker labour markets: labour market pooling; better matchinggain from reduced labour acquisition and training costs in thick local labour markets with abundant specialised labour force

III. Knowledge spillovers: learning about and spreading new ideasface-to-face contact can enable tacit knowledge spillovers through increases in the intensity of the interactions with other firms or individuals

Sources of agglomeration economies

7 7

City productivity increases with city sizeeven after controlling for sorting

8

Doubling the size of a city ≈ 3-5% productivity increase

Cit

y p

rod

uct

ivit

y (

no

rma

lise

d)

Horizontal administrative fragmentation is common as cities outgrow their historic boundaries (more than 10 local governments in 75% of OECD Metropolitan Areas; more than 100 in 22%)

A larger number of local governments may be positive:

• Provide more choice in the provision of public services, more tailored solutions and better accountability (Tiebout, 1956).

• Large literature that finds no scale effects for specific public services (Ostrom, 2010) or governmental expenditure (Kalb, 2010).

But it may also have a potential negative impact:

• Policies, investment and services require city-wide coordination (e.g. Cheshire and Gordon, 1996): e.g. transport; land use; ease of doing business; economic promotion; environmental regulation, etc.

The system of metropolitan governance may affect the productivity & inclusion of cities

9 9

Administrative fragmentation is correlated with lower city productivity

10 10

Administrative fragmentation is correlated with higher segregation of people

11

Hypothesis: Fragmented metropolitan governance can allow for segregation at the level of local units.

-.05

0

.05

.1.1

5

Ine

qu

alit

y b

etw

een

loca

l ju

risd

ictio

ns,

(C

om

po

ne

nt p

lus

resi

dua

l)

0 .2 .4 .6 .8 1

Administrative fragmentation

Controlling for country fixed effects and other city characteristics (i.e. income , population, spatial structure), higher administrative fragmentation is associated to higher spatial segregation by income in different municipalities

• Urban sprawl creates negative externalities in Metropolitan areas (MAs)

• Cooperation is a way to internalize the externalities when making policy decisions

• Sprawl decreased in MAs with a governance body, while increased in those without

Metropolitan governance bodies can reduce urban sprawl

Difference significant at the 99%-level after controlling for log-population levels and country specific trends.

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

1,4

With GovernanceBody

Without GovernanceBody

Change in Urban Sprawl

12

Metro governance bodies can increase the well-being of citizens

55%

60%

65%

70%

75%

80%

With TransportAuthorities

Without TransportAuthorities

Share of Citizens Satisfied with Public Transport

• Public Transport projects usually cut through many jurisdictions

• Cooperation is required for effective implementation and coordination of services

• Citizens are more satisfied in MAs that have metro authorities for public transport

Based on European Urban Audit perception survey. Difference significant at 95% level.

13

How cities affect other regions?

Proximity & connectedness to cities benefits surrounding regions

15

Productivity trends by type of region

16

Rural remote regions present a higher variation in productivity growth rates than other types of regions

Annual average labour productivity

growth, 2000-12 Standard deviation

Coefficient of variation

Predominantly urban

1.01% 1.02% 1.019

Intermediate 1.07% 1.09% 1.024

Predominantly rural close to

cities 1.36% 1.32%

0.972

Predominantly rural remote

0.70% 1.15% 1.641

Note: Labour productivity is defined as real GDP per employee. GDP is measured at PPP constant 2010 US Dollars, using SNA2008 classification; employment is measured at place of work. The coefficient of variation represents the ratio of the standard deviation over the mean.

Source: OECD Regional Outlook 2016

Only Labour productivity of remote rural areas has declined

88%

89%

90%

91%

92%

93%

94%

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Total RURAL

RURAL CLOSE TO CITIES

RURAL REMOTE

Productivity levels of Predominantly Urban regions = 100

The productivity gap between frontier and lagging regions has increased

Notes: Average of top 10% and bottom 10% TL2 regions, selected for each year. Top and bottom regions are the aggregation of regions with the highest and lowest GDP per worker and representing 10% of national employment. 19 countries with data included.

Averages of top 10%

(frontier), bottom

75%, and bottom

10% (lagging) regional GDP per worker,

TL2 regions

50 000

60 000

70 000

80 000

90 000

100 000

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

USD PPP per employee

Frontier regions Lagging regions 75% of regions

1.6% per year

1.3% per year

1.3% per year

Contribution of the different regional productivity patterns to OECD growth

Type of regions Employment share in

2000

GDP share in 2000

Annual avg. GDP growth,

2000-13GDP growth contribution

Frontier 16.1% 20.1% 1.7% 21.9%

Catching up 20.3% 18.2% 2.2% 25.3%

Stableproductivity gap

38.9% 39.1% 1.3% 30.4%

Diverging 24.6% 22.6% 1.6% 22.4%

OECD average 1.6%Note: Frontier regions are fixed for the 2000-13 period. In four countries the values for 2000 or 2013 were extrapolated from growth rates over a shorter time period as data for 2000 or 2013 were not available. The countries are FIN (2000-12), HUN (2000-12), NLD (2001-13) and KOR (2004-13).

Regional policy can improve the

performance of cities & regions

The OECD Regional Development policy paradigm

Compensating lagging regions does not work:

• Creates dependency, not development

• Richer regions may become reluctant to support lagging regions

OECD promotes ‘place-based’ policies focusing on:

• Use of regional specific assets (or create absolute advantages to stimulate competition & experimentation across regions)

• Create complementarities among sectoral policies at the regional (or local) level

• Use of multi-level governance mechanisms for aligning objectives & implementation

• Invest using an integrated strategy tailored to different places

• Adopt effective co-ordination instruments across levels of government

• Co-ordinate across SNGs to invest at the relevant scale

Pillar 1

Co-ordinate across governments and

policy areas

• Assess upfront long term impacts and risks

• Encourage stakeholder involvement throughout investment cycle

• Mobilise private actors and financing institutions

• Reinforce the expertise of public officials & institutions

• Focus on results and promote learning

Pillar 2

Strengthen capacities and promote policy

learning across levels of government

• Develop a fiscal framework adapted to the objectives pursued

• Require sound, transparent financial management

• Promote transparency and strategic use of procurement

• Strive for quality and consistency in regulatory systems across levels of government

Pillar 3

Ensure sound framework conditions at all levels of

government

The OECD Recommendation on the Governance of Public Investment

22

The OECD Governance Models for Rural-Urban partnerships

Explicit rurban partnerships

Rennes (France)

Geelong (Australia)

Nuremberg (Germany)

Central Zone of West Pomeranian Voivodeship (Poland

BrabantStad (Netherlands)

Implicit rurban partnerships

Forlì-Cesena (Italy)

Extremadura (Spain)

Castelo Branco (Portugal)

Central Finland (Jyväskylä and Saarijärvi-Viitasaari) (Finland)

Lexington (United States)

Prague/Central Bohemia (Czech Republic)

Model 1 Model 2 Model 3 Model 4

Delegated functions No delegated functions Delegated functions No delegated functions

Rennes (France) Geelong (Australia)

Nuremberg (Germany)

Central Zone of West Pomerania Voivodeship (Poland)

BrabantStad (Netherlands)

Extremadura (Spain)

Forlì-Cesena (Italy)

Lexington (United States)

Prague (Czech Republic)

Central Finland (Jyväskylä and Saarijärvi-Viitasaari) (Finland)

Castelo Branco (Portugal)

23

OECD (2013), Rural-Urban Partnerships: An Integrated Approach to Economic Development, OECD Publishing.

24

Devolution of spending at lowers level of government is associated with development

AUSAUT

BEL

CAN

CHL

CZE

DNK

EST

EU28

FINFRA

DEU

GRCHUN

ISL

IRL

ISR

ITAJPN

KOR

MEX

NDL

NZL

NOR

OECD25

OECD34

OECD9

POL

PRT SVK

SVN

ESP

SWE

CHE

TUR

GBR

USA

30%

40%

50%

60%

70%

80%

90%

100%

110%

120%

130%

0,0% 10,0% 20,0% 30,0% 40,0% 50,0% 60,0% 70,0% 80,0% 90,0%

GD

P p

er

ca

pit

a a

s a

sh

ar

e o

f U

S G

DP

pe

r c

ap

ita

(%

, b

as

ed

o

n G

DP

pe

r c

ap

ita

in

US

D P

PP

)

SNG expenditure as a % of public expenditure

25

Subnational Governments are key policy actors across the OECD

40%

63%59%

32%

20%

Greece

New Zealand Chile

Estonia

Greece

CanadaCanada

Canada

CanadaCanada

13%

27%

12%8%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Expenditure Staffexpenditure*

Investment Tax revenue Debt**

OECD Minimum Maximum Chile% of general government - 2014

*: No data for Australia**: Debt OECD definition ie including, in addition to "financial debt", insurance reserves and other accounts payable. No data for Mexico, Chile and New Zealand

?

Almost 60% of total public investment across the OECD (2014)

Source: OECD National Accounts

Subnational governments account for the bulk of Public Investment in the OECD

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Rest of the public sector (central government and social security)

Sub-national governments (states, regions and local governments)

59%

27

The singularity of Chile within the OECD: small and centralised public sector

Australia

Austria

Belgium

Canada

Germany

Mexico

Spain

Switzerland

United States

Chile

Czech Rep.

Denmark

Estonia

FinlandFrance

Greece Hungary

Iceland

IrelandIsrael

Italy

Japan

Korea

Luxembourg

Netherlands

New Zealand

Norway

Poland

Portugal

Slovak Rep

Slovenia

Sweden

Turkey

UK

OECD34

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

00% 10% 20% 30% 40% 50% 60% 70% 80% 90%

Pu

bli

ce

xp

en

dit

ur

e a

s a

% o

f G

DP

Subnational government expenditure as a % of Public expenditure

High public spending & high decentralisation

Low public spending & high decentralisation

High public spending & low decentralisation

Low public spending & low decentralisation

• Regional productivity catching-up can contribute significantly to national productivity growth

• It requires mobilisation of specific regional assets and an appropriate governance model across levels of government

• When well-designed decentralisation & regional governance should contribute to aggregate productivity

Bottom-line

OECD (2016), OECD Regional Outlook

OECD (2015) The Metropolitan Century: Understanding Urbanisation and its Consequences

OECD (2015) Governing the City

OECD (2012) Redefining Urban: a new way to measure metropolitan areas

Some references used in this presentation

29

Thank you!Gracias!

joaquim.oliveira@oecd.org