Specialization
Versus Diversification: The Microeconomics of Regional Development and the
Spatial Propagation of Macroeconomic Shocks in Europe
Research
Training Network
Research
Spillovers
and the microeconomics of industrial districts
Forni
(Modena/CEPR) and Paba (Modena) investigated the nature and directions of
inter-industry dynamic linkages across Italian manufacturing sectors. They
found that diversity matters for growth, but each industry needs its own
diversity. They provided some evidence of clustering of industries based
on dynamic externalities. They found that many spillovers occur within
input-output relationships. Spillovers often originate in downstream
sectors favouring the growth of upstream industries and their importance
does not depend on the technological intensity of the industry. They have revised
this work for publication in the Journal of Industrial.
Bertocchi
(Modena/CEPR), Forni and Paba are further exploring the role of
technological spillovers and the geographic scope of localization within
the Italian manufacturing industry. From a preliminary investigation, they
obtained two main results. First, they found evidence of both MAR
(specialization) and Jacobs (diversity) externalities, particularly
whenever geographic agglomerations are characterized by a large incidence
of firms of small size. This suggests a potential propulsive role of small
firms - when they operate under specific conditions - which does not
merely rely on their ability to create more jobs internally. Their second
result is that the simultaneous presence of both MAR and Jacobs
externalities occurs independently of the technological intensity of
sectors, as long as they are characterized by a small average firm size.
Murat
(Modena) has worked on industrial districts and migration. Her research
has so far focused on the causes underlying the low levels of migration
rates in Italy and other European countries, where regional unemployment
is high. The recent literature focuses on factors affecting labour supply:
i.e. workers’ decisions to migrate and reservation wages. The research
emphasises the role of labour demand. It is argued that the shift from
standardized methods of production to flexible production systems, which
occurred in the last two decades, has increased the demand for skilled
workers. In addition, as it happens in Italian industrial districts, part
of the human capital required is region specific. It belongs to native
workers, but can only be acquired at a positive cost by migrants.
Empirical evidence is provided; a theoretical model showing that this
change in labour demand can explain the reduction of internal migration
flows has been developed.
Pistoresi
(Modena) and Giannone (Young Researcher, ECARES) are planning joint
research on the dynamic analysis of spillovers using sectoral industrial
data for Europe. They are developing the econometric framework introduced
by Forni (Modena), Hallin (ECARES), Lippi (ECARES and CEPR/Rome) and
Reichlin (ECARES) for this purpose.
In
“Space, Factors and Spillovers”, Lamorgese and Ottaviano (Sassari/CEPR)
tested the theoretical result that technology spillovers are
geographically more bounded than demand externalities by using a two step
procedure. They firstly had to recover the common and the idiosyncratic
components of total personal income of U.S Cities using a dynamic factor
model approach and then used spatial dependence and the pattern and timing
of diffusion. The dynamic factor model approached used has been developed
by the Modena (Forni) and ECARES team (Hallin, Lippi,Reichlin).
The
analysis of spillovers require the development of spatial econometric
methods. The network is benefiting from the expertise of Croux (ECARES)
who is doing frontier research on the topic. In particular, Croux and
Dehon (ECARES) have developed estimators of multiple correlation
coefficients which are locally robust. This research is going to appear in
Statistical Papers. Croux has also worked with co-authors outside the
network on related topics (see publications below).
Moreover, Rodrigues (Sassari) has worked under the supervision of
Reichlin (ECARES) and now is collaborating with Guiso (Sassari).
In
“Classifying inter-dependent time series in the frequency domain”,
Jorge Rodrigues (Sassari, Young Researcher) proposed a methodology for
classifying a panel of time series into an unspecified number of clusters,
whose structure evolves over time. He defined the similarity measure
in terms of a frequency domain-based correlogram (the coherence spectrum).
A simulation exercise shows that the method performs well relatively to
conventional clustering techniques. Moreover Croux and Wasmer (both ECARES)
applied new methods in the analysis of spatial data and showed that the
use of these methods can uncover features generally ignored by standard
analysis. In particular they analysed the segregation of the labour force
and the spatial propagation of employment shocks. The analysis was applied
to data from 260 communes of Wallonia.
Wasmer
(ECARES) and Zenou (Southampton/CEPR) developed a model in which workers'
search efficiency is negatively affected by access to jobs. They proved
that there exists a unique and stable market equilibrium in which both
land and labour markets are solved simultaneously. They found that,
despite inefficient search in the segregated city equilibrium, the welfare
difference between the two equilibria is not so large due to differences
in commuting costs.
Perelman,
Rodrigues (Young Researcher, Sassari) and Wasmer used a methodology to
identify dimensions which can be interpreted as segments of a dualistic
labour market. This method, applied to two samples, Belgian and US workers
respectively, showed that the first segment is composed of adult, male,
qualified workers while the second segment by female, young people and
non-qualified workers.
Understanding
the propagation of macroeconomic shocks
On
this topic, the network is working on different topics.
(a)
Econometric
methodology (which
mainly involves the Modena and the ECARES teams plus two researchers from
Sassari (Lamorgese and Rodriguez).
Giannone
(Young Researcher, ECARES) has investigated the problem of testing for
linear restrictions in this framework. This method is used in applications
described below. “Testing for
linear restrictions with a large panel of time series “.
In this paper, he proposed a procedure to test for over-identifying
restrictions in large-scale dynamic factor models. The properties of the
test for large cross-sections and time dimensions were analysed and a
simulation study was performed to examine the finite sample properties
(b)
Interplay between public policy and economic geography
Pistoresi
and Strozzi (Modena) concentrated their research on the empirical analysis
of the actual functioning of the different wage bargaining systems in
Europe. In particular, they studied how centralised and decentralised wage
bargaining agreements face national and sector-specific shocks to labour
productivity at the different frequencies of the cycle. Special
attention was devoted to the Italian basic metal industry. They are
currently revising this work for the Giornale degli Economisti e Annali di Economia.
(c) Geography and growth
Martin
published two articles written in collaboration with Ottaviano (CEPR). The
first article "Growth and agglomeration" will be published in
the International Economic Review.
This paper presents a model in which growth and geographic agglomeration
of economic activities are mutually self-reinforcing processes. Economic agglomeration in one region spurs growth because it
reduces the cost of innovation in that region through a pecuniary
externality. Growth fosters agglomeration because, as the sector at the
origin of innovation expands, new firms tend to locate closer to this
sector. The second paper,
"Global Economic Divergence, Trade and Industrialisation: The
Geography of Growth Takeoffs", is published in another leading
journal, the Journal of Economic
Growth. In this article the authors take a step towards formalising
the theoretical interconnections among four post-industrial revolution
phenomena- the industrialisation and growth take-off of rich northern
nations, massive global income divergence, and rapid trade expansion.
(d)
Informational
spillovers.
Schivardi
(Sassari) showed how cyclical
aggregate shocks can stimulate structural reallocation activities, which
in turn amplify the effect of the shock. He analysed the informational
aspects of restructuring activities and their interplay with aggregate
shocks. He developed a model in which production units are uncertain about
the value of staying in the market and learn about it over time in a
Bayesian fashion.
Guiso
and Schivardi (Sassari) tested the theoretical proposition that
information on relevant state variables spills over from one firm’s
decision to another. They tested three independent implications of
information spillovers (IS) models and the results were remarkably
consistent with the theoretical predictions.
Guiso,
Schivardi and Pistaferri (Stanford/CEPR), used a long panel of matched
employer-employee data to test the theoretical predictions of
principal-agent models of wage determination in a general context where
all types of workers, and not only CEOS, are present. They showed that the
amount of insurance varies by type of worker and firm in ways that are
consistent with principal-agent models but are hard to reconcile with
competitive labour market models, with or without frictions.
Lamorgese
(Sassari) showed why the empirical results in urban economics cannot
perfectly explain the growth performance for the bulk of the U.S cities.
(e)
Risk
sharing, specialization and diversification
The
UPF team (Ciccone and Canova) looked at the benefits of specialisation
economies in Europe and compared it to the estimates available for the
USA. In cases where they did not find estimates for the USA that were
comparable to European estimates, they also estimated specialisation
economies in the USA directly. With regard to its work on Europe, they
followed two strategies: firstly, by generating estimates for
specialization economies across as many European countries as possible as
long as there was comparable data of sufficient quality; and secondly,
developing very detailed data, and a matching high-quality empirical
approach, for special cases where this was possible.
Yosha
(Tel Aviv) with Sebnem Kalemli-Ozcan and Bent Sorensen (CEPR) provided
empirical evidence that risk sharing enhances specialisation in production
and calculated an index of regional specialization for European Community
(EC) and non-EC OECD countries, US states, Canadian provinces, Japanese
prefectures, and regions of Italy, Spain, and the United Kingdom.
The
issue of risk sharing has been analysed looking at metropolitan areas as
relevant units by Lamorgese (Sassari). He has formalised one mechanism
through which the operation of labour pooling economies dampens volatility
in the local labour market, improves the incentives of producers to
perform R\&D and smoothes the life cycle fluctuations of a city.
In addition, he has investigated how much of risk-sharing operates
among cities within and across states in the U.S.
Giannone
(ECARES, Young Researcher) analysed features of the propagation mechanism
of technology and aggregate demand shocks across European countries.
Giannone
(ECARES), Reichlin (ECARES) and Sala (Young Researcher, Tel Aviv) are
studying the dynamic of “common” European policy shocks across the
single European economies. They are using a large cross section of time
series from the largest Euro economies. The econometric model is a
development of Forni, Hallin, Lippi and Reichlin (2000) which allows for
the structural identification of shocks and impulse response functions.
This research is still in progress.
Research
at the CEPR node has focused on the propagation and dynamics of
macroeconomic shocks across regions. More specifically, Fatás and Mihov
studied the role that fiscal policy can play in an integrated economic
area such as the European Union. Their study made use of both
international and intra-national (regional) data from Europe and the US to
uncover the stabilising effects of fiscal policy. The results showed that
there is a strong stabilising effect of fiscal policy on regional output.
This effect is much larger in magnitude than what previous studies have
shown. A second issue addressed in their research is how labour mobility
and regional migration affects the volatility and co-movements of regional
business cycles. Fatás
provided a comprehensive study of the dynamics of labour mobility among
European regions and US states.
Forni
(Modena) and Reichlin (ECARES) have studied the evolution of the
integration of European economies from the real side and the financial
side. They found that while financial markets have become progressively
more integrated, real economies have not. They analysed these findings
evaluating risk-sharing versus specialisation factors. Their work is very
much related to that of Sorensen (CEPR) and Yosha (Tel Aviv) and Lamorgese
(Sassari). This research is still in progress.
In
addition, Yosha and Sorensen sought to provide empirical evidence that
more international portfolio holdings and more foreign direct investment
are associated with greater international risk sharing.
Lucrezia
Reichlin (ECARES) organised the first conference held in Brussels on 20/21
October 2000, at which representatives from all teams and some young
researchers were present. A full list of participants can be found in
Annex I.
Philippe
Martin (CERAS) co-organized with Antonio Ciccone (UPF) the first
conference related to the RTN: the general topic was “Macro and micro
aspects of economic geography”. This conference will take place in
Barcelona on 23/24 November 2001. The programme can be found in Annex II.
Data
sets and stylised facts
Some
of the work of the network has been devoted to the construction of new
detailed data sets for the analysis of specialisation and externalities at
the local level and macro data sets for the analysis of national and
sectoral cycles in Europe.
The
new data sets and major descriptive findings are listed below:
Barrios
(Young Researcher, CORE), Bertinelli (CORE), Strobl (University College
Dublin) and Teixeira (CORE) constructed a comparable (and exhaustive)
plant-level data and common industrial classification for the
manufacturing industries of Belgium, Ireland and Portugal to study
patterns of agglomeration.
They
find that patterns of location strongly differ across industries and
moreover, contrary to common believes, high tech sectors are not the most
agglomerated ones, pointing hence to the burden of history. Second, the
Irish and Portuguese manufacturing location is analysed on a 14-year
period of time. Again, the evolution has been very divergent according to
industries, with upward/downward sloping agglomeration patterns.
Rita
Almeida (Young Researcher, UPF) has finished developing her regional data
set for Portugal. The data set has very detailed information on the
characteristics and location of firms. Its detail is unmatched within
Europe.
Almeida
has used this data to analyse the prevalence of specialisation versus
diversification spillovers at the regional level (see below).
Antonio Ciccone (UPF) and Federico Cingano (UPF/Bank of Italy) have developed an extensive data set on experience wage-premia in Italian industrial districts.
The objective is to examine whether local spillovers related to specialisation or diversification of the industrial structure translate into greater experience wage-premia (see below).
Forni
and Paba (Modena) have developed a
data set for the analysis of local spillovers. The primary sources are the
ISTAT Industry Censuses of 1971 and 1991. Original data include over three
million data points, i.e. employment and the number of local plants for
the mentioned years, 101 three-digit NACE-ATECO industries and 8,086
municipalities. These data have been carefully reclassified by ISTAT in
order to harmonise the 1971 and the 1991 sector definitions. They have
spatially aggregated the data into 955 larger areas, whose boundaries have
been identified according to ISTAT-IRPET (1986). These areas are called
local labor systems and are constructed by using data on the residence and
the workplace of workers from the 1981 ISTAT Population Census. The main
idea is to cluster municipalities in such a way as to get areas which are
both small and "self-contained", in the sense that many of the
workers living in the area have their workplace within the area.
Eric
Bartelsman, Stefano Scarpetta & Fabiano Schivardi (both Sassari) in
“Comparative Analysis of Firm Demographics and Survival: Micro-level
Evidence for the OECD countries” constructed a dataset of sectoral
indicators of firm dynamics for 10 OECD countries using information from
business registers. The patterns of firm entry, exit, survival and
employment growth are described and analysed across countries, sectors,
and over time.
Altissimo,
Bassanetti, Cristadoro and Veronese (Bank of Italy) have updated and
completed the data set on the euro economy which contains 1000 monthly
time series for on macro variables for different sectors and countries
since 1986. This data set is used to construct the euro economy business
cycle indicator (EuroCOIN) published by the CEPR and has been used in
various papers listed below. Forni (Modena) and Reichlin (ECARES) have
supervised this project.
Hedva
Ber, Asher Blass and Oved Yosha (Tel-Aviv) have computed, from firm-level
data for Israel 1991-8, a year-by-year flow of funds chart. This has been
constructed by combining information from reports on the use of funds that
publicly traded companies are required to provide since 1990 with
information from standard financial statements. This has allowed them to
compute firm-level year-by-year capital expenditure and inventory
investment taking into account factors that cannot be controlled for it if
only financial statements data are used.
Location
and productivity growth
A.
Agglomeration, specialization and productivity growth.
(i) Specialization, localization and productivity growth
Federico
Cingano & Fabiano Schivardi (Sassari) in “Sources
of Local Productivity Growth” use balance-sheet based TFP indicators
(as opposed to employment-based proxies) to estimate the effects of
alternative sources of dynamic spillovers on manufacturing productivity
growth at the local geographical level. Contrary to previous empirical
work they find that industrial specialisation affects TFP growth rates at
the city-industry level, while they do not find evidence that either the
degree of local competition or productive variety (known as Porter and
Jacobs-like external effects, respectively) do impact on subsequent
productivity growth at the local level. Employment based regressions yield
nearly the opposite results, in line with the previous empirical works.
They argue that the existing literature, inferring evidence on alternative
sources of agglomeration economies based on employment growth regressions,
might suffer from serious identification problems.
Almeida
(Young Researcher, UPF) (in her paper “Local
Characteristics and Growth in Portuguese Regions”) finds that, in
Portugal, specialisation is the key variable determining regional
spillovers and hence employment growth at the regional level.
Diversification spillovers play a much smaller role (especially compared
with similar work done for the US).
Paba
(Modena) is now trying to estimate how firms’ size and sectoral
specialization affect economic growth in European regions. Preliminary
results indicate that small firms positively affect the growth performance
of regions while sectoral specialization does not matter
Ghio
(Young Researcher, Modena), with co-authors Luchini, Mendy and Rychen,
studied geographical spillovers and regional industrial growth with a
special focus on France. They develop a method to measure spatial
interactions between French regional industries and show that the
probability for a French region to be linked to another one depends on
industry specific structural indicators and geographical proximity.
Part of this work is forthcoming in La
Revue Economique.
Fabio
Panetta, Fabiano Schivardi (Sassari) and Matthew Shum in “Detecting
the Informational Effects of Bank Mergers” formalise one mechanism
through which the operating of labour pooling economies dampens volatility
in the local labour market, improves the incentives of producers to
perform R\&D and smooths the life cycle fluctuations of a city. A
trade-off between diversification and specialisation is pointed out in the
case of concave R\&D technology. Assuming that innovation becomes more
difficult at each round, this paper provides a theoretical back up for and
reproduces a well-known piece of empirical evidence: upstart cities tend
to be more specialised, mature cities tend to be more diversified.
(ii)
Localization, learning and growth
In
a paper written with Thierry Verdier (CERAS), Thoenig (CERAS) investigates
how informational spillovers affect growth and innovation.
In a nutshell, to reduce informational leakages and spillovers
which can be freely acquired by outside competitors, and thereby lessen
the threat of imitation and technological leapfrogging, firms have
incentives to increase the share of tacit knowledge and non-codified
know-how embedded in their production process.
But they do so at the cost of a larger share of skilled labour in
their workforce. In this
context, openness, by intensifying international technological
competition, triggers a race to imitation and innovation. As a
consequence, it may induce firms to develop innovations of a new kind,
less imitable and endogenously more skill intensive.
Suggestive empirical evidence for France is provided where the
existence of defensive skill biased technical change is clearly
established. This paper is to
be published in the American
Economic Review.
(iii)
Human capital, localization and productivity growth.
Antonio
Ciccone (UPF) has terminated his research work on human capital
externalities within cities (“Identifying Human Capital Externalities:
Theory with an Application to US Cities” with Giovanni Peri).
Human capital externalities at the aggregate level play a central
role in applied economic theory as well as economic policy analysis. In
applied theory, human capital externalities are invoked to capture key
features of the data. In policy analysis, the strength of human capital
externalities is one of the main determinants of the optimal subsidy to
human capital. Assessing the strength of human capital externalities at the
aggregate level is therefore important for economic theory as well as
policy, and empirical research has responded with a variety of different
approaches and estimates. The
theoretical identification problem is still not fully understood however.
The difficulty is simple to explain. Empirical work finds that
workers with different levels of education are imperfect substitutes in
production. An increase in
the aggregate supply of highly educated workers will therefore tend to
increase wages of workers with low levels of education and decrease wages
of workers with high levels of education, even if wages of highly educated
workers reflect their marginal social product (and there is no need for
corrective policies). Can we
avoid mistaking these standard supply effects with (positive or negative)
human capital externalities at the aggregate level? So far the answer to
this question is unclear as existing work on the estimation of the
strength of human capital externalities at the aggregate level assumes
that workers with different human capital are perfect substitutes in
production. Perfect
substitutability simplifies the theoretical identification problem because
it implies that the aggregate supply of human capital does not affect
individual wages if there are no externalities.
All effects of the supply of human capital on individual wages can
therefore be interpreted as externalities. This yields, for example, that
average-schooling externalities at the local geographic level can be
estimated by simply including average schooling of the local workforce in
a standard Mincerian wage regression.
It can be shown however that if workers with different levels of
education are imperfect substitutes in production then this (Mincerian)
approach to human capital externalities at the aggregate level may yield
positive or negative
externalities even if wages reflect marginal social products.
The main point of the work of Ciccone and Peri is to show how human
capital externalities can be identified in the case of imperfect
substitutability. The approach is then implemented to estimate human
capital externalities at the city level.
Antonio
Ciccone (UPF) is also working on a research project on “Local geographic
spillovers, learning, and the experience premium” (Manuscript,
Universitat Pompeu Fabra and Bank of Italy, 2001 (with Federico Cingano).
The basic idea is that it takes time for workers to learn about the
local knowledge. Moreover, we
would expect that in industrial districts with more knowledge (because of
specialization and diversification spillovers) there is faster learning.
To the extent that this learning translates into productivity, it
will be picked up in more rapid wage-growth, i.e. a higher experience
wage-premium. Hence,
specialization or diversification spillovers can be estimated using data
on industrial district level experience wage-premia (data set described
above).
Murat
and Paba (Modena) have completed for the Rivista
di Politica Economica a paper where they study why internal migration
rates so low in Italy and other European countries, with a special focus
on Italian industrial districts, where part of the human capital required
is region specific. They
emphasise the role of labour demand and argue that the shift from
standardised to flexible methods of production has increased the demand
for skills and explain the reduction of internal migration flows.
(iv)
Specialization, trade and productivity growth.
Antonio
Ciccone (UPF) has also terminated his research work on the role of
regional and international specialization for average labor productivity
(“Trade and Productivity” with Francisco Alcala). Theories about trade
increasing aggregate productivity at the country or regional level are
nearly as old as economics. But
how large is this effect empirically?
Answering this question is difficult because any particular summary
measure of trade is likely to miss some aspects of how trading activities
affect countries’ productivity. Moreover,
while trade may increase aggregate productivity, the reverse also seems
likely. Empirical work therefore has to make sure to identify the effect
of trade on productivity instead of the other way round.
The summary measure of trade nearly always used in empirical work
is nominal imports plus exports relative to nominal GDP, usually referred
to as openness.
We argue that estimates of trade’s effect on average labor
productivity at the country or regional level in the existing literature
give a misleading picture of the true productivity-gains caused by trade
because of this summary measure of trade used in the empirical analysis.
Summarizing trade using nominal imports plus exports relative to nominal
GDP (openness) has drawbacks for empirical cross-country productivity
analysis that are easily explained. Suppose
that trade increases productivity but that the implied productivity-gains
are greater in the tradable goods sector (e.g. manufacturing) than in the
non-tradable goods sector (e.g. services).
Will countries or regions that are more productive because of trade
have higher values of openness? Not
necessarily, because the relatively greater productivity-gains in the
tradable goods sector lead to a rise in the relative price of non-tradable
goods, which may decrease openness when the demand for non-tradable goods
is inelastic. We show this
formally in a model where productivity-gains from trade arise due to
increasing returns to specialization.
We therefore propose an alternative measures, real openness, that
does not suffer from this problem and show that this measure works much
better than openness in cross-country average labor productivity analysis.
(v)
Technology, specialization and productivity growth
Silvia
Fabiani, Fabiano Schivardi and Sandro Trento (all Sassari) in “ICT
adoption in Italian manufacturing’’ study the Information and
Communication Technologies (ICT) adoption choices of Italian manufacturing
firms. Previous empirical evidence, both at the macro and at the micro
level, has shown that investment in ICT is associated with substantial
increases in productivity. Much
less is known on the conditions more conducive to its diffusion.
They investigate the factors, both at the individual and at the
environmental level, that might speed up or slow down ICT adoption,
identifying potential areas of criticality.
They relate ICT adoption to two sets of variables.
First, they consider firm specific variables, such as size, the
composition of the labor force, indicators of employment flexibility and
financial indicators that, according to the literature on technology
adoption, might play a role in determining the optimal rate of investment
in ICT. Second, given the
network aspect of many ICT, particularly the Internet-based applications,
one may expect that the local industrial structure impacts adoption
patterns even after controlling for individual characteristics.
For example, large firms might act as catalyst for the adoption of
web technologies of smaller intermediate inputs suppliers. They use a
survey that covers approximately 1500 manufacturing firms, explicitly
conducted in 2001 to monitor ICT investments.
They obtain indicators of the local industrial structure from a
dataset of the National Institute for Social Security (INPS) that covers
the whole population of Italian firms.
The results can be summarized as follows: the number of pc per
workers depends mainly on the human capital of the labor force, while
other aspects, in particular size and sector, play a minor role. We
interpret this as evidence that pc constitute a mature, relatively
inexpensive technology. When
considering Internet-based applications, they find that size is positive
correlated with adoption and that local industrial structure matters: in
particular, the presence of large firms in the area has a positive impact
on adoption, an evidence that could indicate that large firms play the
role of catalyzers of the adoption choice and help overcoming the
suboptimal adoption rate induced by network externalities.
Lusito
Bertinelli (CORE) and Rosella Nicolini (CODE, Universitat Autonoma de
Barcelona) investigate the effect of spillovers on location decision of
firms. They develop an analysis merging the geographer toolbox with the
standard econometric techniques. For
a chosen sample of sectors, through the spatial data analysis, they test
the existence of positive spatial auto-correlation for R&D investments
that lead R&D expenditure to cluster.
They succeed in detecting in how far the local environment may
influence the firm decisions in R&D investments. Data confirm that the
proximity to other firms investing in R&D may produce positive
externalities. The diversity
vs. specialization debate is also tackled.
In
“Detecting the Informational Effects of Bank Mergers”, Fabio
Panetta, Fabiano Schivardi (Sassari) and Matthew Shum formalise one
mechanism through which the operating of labour pooling economies dampens
volatility in the local labour market, improves the incentives of
producers to perform R\&D and smoothes the life cycle fluctuations of
a city. A trade-off between diversification and specialisation is pointed
out in the case of concave R\&D technology. Assuming that innovation
becomes more difficult at each round, this paper provides a theoretical
back up for and reproduces a well-known piece of empirical evidence:
upstart cities tend to be more specialised, mature cities tend to be more
diversified.
B
. Determinants of agglomeration
(i)
Policy
Philippe
Martin (CERAS) finished a book that will be published by Princeton
University Press on “Economic Geography and Public Policies”. The book
is co-authored with with Richard
Baldwin (HEI, Geneva), Rikard
Forslid (Stockholm), Gianmarco Ottaviano (Milano) and Frédéric Robert-Nicoud
(Geneva). Several chapters analyse the relation between geography and
growth and the impact of public policies on both. The chapters are
available on the web site of Philippe Martin: http://www.enpc.fr/ceras/martin/.
Although this has been completed under the auspices of another network,
several aspects of the research have applications to the research
objectives of this network which will be exploited
Teixeira
(CORE) studies empirically the link between transport policy and the
distribution of economic activities in Portugal between 1985 and 1998.
He asks whether the reduction of transport costs is likely to
contribute to spatial polarization. In light of new data, it results that
the Portuguese transport policy is not working towards spatial equity.
An assessment of the implications of the envisaged transport
network in 2010 for regional development and spatial equity based on a
simulation of the model is also made. This simulation proves right another
major prediction of the new economic geography: if transport costs are
lowered sufficiently the spreading of industry occurs.
Barrios
(Young Researcher, CORE), Görg (University of Nottingham) and Strobl
(University College Dublin) study the role played by regional public
incentives as a determinant of multinationals’ location choice in Irish
regions over the period 1973-1998 using nested logit technique.
They find regional policy and public incentives to have a positive
influence on multinationals’ location choice, especially from the
mid-1980ies onward. They also
find that urbanisation economies and the level of skills, for which the
richest areas of Ireland had a distinctive advantage over the rest of the
country, played a major role in attracting foreign firms in hi-tech
industries. However, further
evidence shows that regional policy has also been effective for this kind
of industries and has more than compensated the relative disadvantage of
the poorest counties of Ireland by attracting a substantial number of
hi-tech multinationals in those regions.
Eric
Bartelsman, Stefano Scarpetta and Fabiano Schivardi (Sassari) in
“Comparative Analysis of Firm Demographics and Survival: Micro-level
Evidence for the OECD countries” present evidence on firm demographics
and firm survival for a group of ten OECD countries. For each country the
patterns of firm entry, exit, survival and employment growth are described
and analysed across countries, sectors, and over time (data set described
above). The paper provides a
discussion of how these data may be used to gain a better understanding of
the process through which economic policy and institutions may affect
aggregate patterns of employment, output, and productivity growth.
(ii)
Multinationals
Barrios
(Young Researcher, CORE), Bertinelli (CORE) and Strobl (University
College Dublin) consider
a special case where agglomeration and growth may go hand in hand by
studying the co-agglomeration of domestic plants and foreign
multinationals in Ireland over the period 1983-98.
To this end they make use of the index developed by Ellison and
Glaeser (1997) and find co-agglomeration to be important for a number of
sectors. They further test
for the impact of co-agglomeration on domestic firms’ employment growth
using panel level data. Foreign
presence as well as foreign employment density are found to be important
determinants of employment growth, especially for those sectors with a
high degree of co-agglomeration.
Barrios
(Young Researcher, CORE), Bertinelli (CORE) and Strobl (University College
Dublin) investigate whether multinational corporations (MNCs) can foster
local indigenous development using the case study of Irish manufacturing.
To do such they utilise a detailed plant level data set to
determine whether the presence of MNCs affected domestic start-ups
regionally within Ireland. Results show that apart from nation-wide
spillovers, MNCs also have a purely local impact. If inter-industry
spillovers are allowed for, such spatial spillovers can reach as far as
neighboring regions.
(iii) Capital, labour and industry mobility
Paba
and Murat (Modena) are also undertaking research on the role of labour and
capital mobility for the growth of local areas and industrial districts.
The focus is on region specific human capital and local codes of
behaviour, which may prevent the mobility of entrepreneurial firms and
workers and undermine the growth potential of industrial districts.
Economic
integration and the localization of economic activity
Agglomeration,
specialization and integration
Texeira
(CORE) analyses how economic integration in Europe has affected the
industrial agglomeration in Portugal between 1985 and 1998.
The identification of the most and least localized industries
reveals similar patterns between Portugal and other countries.
Also, old establishments as well as large establishments tend to be
more concentrated than the industry average.
In addition, both job creation and job destruction played an
important role on the increasing of industrial agglomeration during this
period of time.
Ghio
(Young Researcher, Modena), working with co-authors Catin and Van Huffel,
developed a model of economic
geography which analyses the influence of transport infrastructure on
economic efficiency and regional equity during an integration process.
Using numerical simulations, he showed that different kinds of
transport infrastructures policies such as those adopted by the European
union may lead to an inconsistent result both with economic efficiency and
regional equity. Part of this
research program has appeared in
the Revue d’Economie Politique.
Barrios
(Young
Researcher, CORE) and Strobl (University College Dublin) study the
concentration of industries in Europe.
Their analysis starts from Dumais et al. (2002) who have shown for
the US that industries are extremely mobile and that non-historical
factors attributable to randomness in industries’ location play a major
role in geographical concentration. They
apply this methodology to the case of EU countries and regions between
1972 and 1995 and find that changes in concentration levels were mainly
due to industry mobility rather than historical accidents and past levels
of concentration as often argued by the New Economic Geography literature.
Thoenig
(CERAS) built a theory where the decision of outsourcing some part of the
production processes enables the firms to balance between the gains of
specialization and the possibility of sharing (product market) risks with
their foreign suppliers. In a paper with David Thesmar (INSEE) a
macroeconomic model is developed where firms may endogenously outsource
part of their production process. Starting
from the premise that adaptation to uncertainty cannot be contracted upon
in the worker - employer relationship, outsourcing decisions then balance
flexibility gains against hold-up costs of opportunistic behavior by
outside contractors. In
equilibrium, the degree of outsourcing is shown to depend on the degree of
product market competition, contractor's bargaining power, and the
volatility of demand shocks. The
main result is that an increase in the degree of outsourcing amplifies the
volatility of firm sales and employment; it does not, however, amplify
aggregate uncertainty. This
theory is therefore a good candidate in explaining the rise in firm level
uncertainty witnessed in the European Markets over the past 30 years. It
also provides valuable insights on the relation between globalisation,
technical change, firm level uncertainty and job instability.
Finally, theory's implications are brought to the test on French
data. Evidence from firm level data is shown to be largely consistent with
the main implications of the theory.
Thoenig
(CERAS) has investigated on the links between the destination of export
flows and their impact to labour markets.
In a paper “Globalisation and the demand for skills: an export
based channel”, written with Eric Maurin (INSEE) and David Thesmar (INSEE),
international trade is shown to affect the demand for skill an export
based channel. The working
hypothesis is that the very act of exporting requires an effort of skill
upgrading, in particular among occupations related to marketing and
development. Using firm level data, they estimate a model that breaks down
production into two stages: product development and marketing, and actual
production. Once the biases
arising from the endogeneity of export decision are corrected, they find
strong support for our hypothesis. The
skill requirement in development/marketing occupations increases with the
share of exported output. Overall skill upgrading is as important among
firms exporting to OECD countries as among those exporting outside of the
OECD to the LDCs. This paper
has been submitted to the Journal of
International Economics where it is currently under revision.
Integration
and trade
Andrea
R. Lamorgese (Sassari) and Gianmarco I.P. Ottaviano in “Blind
Date”, build a multi-country Chamberlinian-Heckscher-Ohlin model
with iceberg trade costs to study the effects of preferential trade
agreements (in particular, enlargement).
In particular, they study the impact of enlargement on an existing
member depending on its relative factor endowments and distance with
respect to the average.
Integration,
risk-sharing and insurance: the role of policy
Helge
Sanner (Young Researcher, Modena) has studied the
impact of three different policy regimes for federal or regional as
opposed to central unemployment insurance.
A straightforward application of this analysis is the question
whether unemployment insurance within the European Union should be
organised on the European or on the national (regional) level.
He finds that workers are always in favour of central unemployment
insurance, while it depends on the type of regionalisation whether or not
firms fare better with federal or with central unemployment insurance.
Integration,
risk-sharing and insurance: the role of financial markets
Luigi
Guiso (Team Leader, Sassari) and Jorge Rodrigues (Young
Researcher, Sassari) in “Co-movement
and debt capacity ”, investigate the empirical relation between the
market liquidity of a firm's assets measured by its co-movement with other
firms and the firm debt capacity. The
findings show that firms with more liquid assets are able to raise more
funds externally. This
relation is stronger among firms with higher cash flow and a lower risk
that their managers transform the assets in place to their advantage.
However, for firms with low cash flows and high transformation risk the
relation between debt capacity and assets liquidity bends backwards at
high levels of liquidity. These
findings are consistent with recent theories of liquidity paradoxes which
emphasise the dark side of liquidity.
Consistent with this view they find that high co-movement firms
tend to invest more in less intrinsically liquid assets so as to harden
their commitment not to transform assets. They also analyse the effect of
assets liquidity on debt maturity.
In
a paper with Hélène Rey (CEPR), Philippe Martin (CERAS) has analysed the
impact of size on financial markets.
The paper is entitled “Financial Super-Markets: Size Matters for
Asset Trade”. Empirically,
demand and market size effects play an important role for international
trade in assets and in the determination of asset prices.
Asset prices are higher in larger financial areas, financial
integration decreases the cost of capital and market size determines
financial flows. They present
a two-country macroeconomic model where the interaction of a risk
diversification motive and market segmentation explains those facts.
The imperfectly competitive structure of financial markets in our
setting also provides a new explanation for the home bias in equity
holdings. Due to
co-ordination failures, the extent of financial market incompleteness,
which is endogenous in their model, is inefficiently high.
Antonello
d’Agostino, in “Testing financial market integration: a structural
approach” tests the degree of financial markets integration for 5
European countries using a sample of 15 years excess returns data.
The analysis is performed for three sub-samples: pre-convergence,
convergence and Euro period. Significant
evidence indicates an increasing integration in the convergence period and
a slight decrease in the Euro period. The growing importance of the global
influences seems to be the main driving force in explaining co-movements
in equity returns. A decrease of national influences in the three periods is
also observed.
Marina
Emiris carried out work on “Measuring Capital Market Integration”.
The convergence of European economies in view of the European
Monetary Union together with increasingly common dynamics in currency and
equity returns suggest that capital markets are at least partially
integrated. In her study, Emiris imposes a dynamic factor analytical model
for the returns on currency and stock portfolios, on eight European
markets, taking into account predictability by forward premia and dividend
yields. The resulting asset pricing model is characterized by
time-varying risk premia, and constant betas and return variances. She proposes a measure for the degree of integration and
examines its evolution from 1979 until 1997.
She found that the degree of integration for equity markets has
increased in the nineties but that this is mainly due to an increase in
the premium for extra-European currency risk.
She also found that the sources of comovement lie only in part in
the US equity markets. (published
in Conference Papers Series, Bank of International Settlements (June 2002)).
Marina
Emiris, in “Liquidity and asset price linkages studied financial market
linkages between Europe, US and Japan during periods of tight monetary
conditions. In particular, she investigated whether monetary conditions
generate asymmetries in financial market linkages. Liquidity states are
determined using a threshold model for the monetary policy instrument in
each country. For each state of liquidity in the foreign or local economy,
the common dynamics of monthly returns on bond yields of different
maturities and sectoral stock returns over the period 1980-2000 are
estimated using a dynamic factor model (Forni, Hallin, Lippi and Reichlin
(2000)). Then, the importance of location (geography), common currency and
real and monetary convergence in spillover asymmetries is investigated.
Furthermore, for each liquidity state, a measure for short run and long
run interdependence defined as the dynamic correlation (see Croux, Forni
and Reichlin (2000)) between markets and between assets is estimated. This
measure is used to investigate three issues: whether national linkages are
similar to cross-border stock-bond market linkages, whether short run
interdependence increases during low liquidity times and finally, the
presence of flight-to-quality phenomena. [paper presented at the Paris
conference].
Domenico
Giannone (Young Researcher, ECARES) and Michele Lenza (ECARES) in
“Explaining the Feldstein-Horioka facts” studied the effect of global
factor in accounting for cross-country saving behavior and explain the
Feldstein-Horioka puzzle. The Feldstein-Horioka finding is a very well
known and puzzling empirical regularity in International Macroeconomics:
domestic saving and domestic investment in OECD countries are highly
correlated contrary to what is predicted by partial equilibrium versions
of the intertemporal Theory of Current Account. It has been argued that
General Equilibrium effects could play an important role in generating
this empirical regularity. They took into account these effects by
controlling for co-movement generated by global factors in the
saving-investment regressions. When controlling for comovements they find:
(1) The Feldstein - Horioka puzzle is to be de-emphasized as the saving-
retention coefficient considerably decreases in a panel regression on OECD
countries; (2) The saving - retention coefficient seems to decline over
time becoming not significant in the 90.s. They interpreted this finding
as a consequence of increased capital mobility in OECD countries [paper to
be presented at the Paris network conference in December 2002].
In
another paper with Rey, “Financial Globalization and Emerging Markets:
With or Without Crash?” Philippe Martin (CERAS) analysed the impact of
financial globalization on asset prices, investment and the possibility of
crashes driven by self-fulfilling expectations in emerging markets. In a
two-country model with one emerging market (intermediate income level) and
one industrialized country (high income level), they show that
liberalization of capital flows increases asset prices, investment and
income in the emerging market. However, for intermediate levels of
international financial transaction costs, we find that pessimistic
expectations can be self-fulfilling, leading to a financial crash. The
crash is accompanied by capital flight, a drop in income and investment
below the financial autarchy level and more market incompleteness. They
show that emerging markets are more prone to financial crashes simply
because they have a lower income level and not because of the existence of
market failures (moral hazard or credit constraints), bad monetary
policies or exchange rate regimes.
Sorenson,
Yi-Tsung Wu and Yosha (Tel Aviv) in “Is less home bias associated with
more international risk sharing? An empirical analysis’’ obtained
evidence on the link between home bias and lack of international risk
sharing by exploiting cross-sectional variation (across countries) in
patterns of foreign security holdings and ask if countries with more
foreign asset holdings obtain better income and consumption insurance.
Localization
and macroeconomic fluctuations
Kalemli-Ozcan,
Sorensen and Yosha (Tel Aviv) published “Economic Integration,
industrial specialization and the asymmetry of macroeconomic
fluctuations” in the Journal of International Economics (55, 2001).
They showed empirically that regions with a more specialized
production structure exhibit output fluctuations that are less correlated
with those of other regions (less “symmetric” fluctuations). Combined
with the casual relation running from capital market integration to
regional specialization found in an earlier study, this finding supports
the idea that higher capital market integration leads to less symmetric
fluctuations. This mechanism
counter-balances the effect of lower trade-barriers on the symmetry of
fluctuations quantified by Frankel and Rose (1998).
Deriving a simple closed from expression for the gains from risk
sharing for CRRA utility is an independent contribution of the present
article.
Sorenson,
Yosha (Tel Aviv) and Wu published “Output fluctuations and fiscal
policy: US state and local governments 1978-1994’’ in the European
Economic Review (45-2001). They studied the cyclical properties of U.S.
state and local government fiscal policy.
The budget surpluses of both are pro-cyclical over short-and-medium
term horizons. Pro-cyclical surpluses are the result of strongly
pro-cyclical revenue and weakly pro-cyclical expenditure.
These results hold whether aggregate (U.S.-wide) fluctuations are
controlled for or not. Federal
grants to state and local governments are pro-cyclical, but this is due to
aggregate output fluctuations: With respect to sate-level output
fluctuations, federal grants are counter-cyclical.
The budget surpluses of trust funds and utilities are also
pro-cyclical, but there is no evidence that this effect is stronger in
states where balanced budget rules are tighter.
The cyclical patterns of state and local budget surpluses are
affected by various political institutions; for example, budget surpluses
are less pro-cyclical in conservative states. In an election year that
occurs in “good times”, state governments refrains from accumulating a
larger surplus. In “bad
times”, no such asymmetry between election and non-election years is
apparent.
Ostergaard,
Sorenson and Yosha (Tel Aviv) published “Consumption and aggregate
constraints: evidence from US states and Canadian provinces” in Journal
of Political Economy (110-2002). State-level consumption exhibits excess
sensitivity to legged income to the same extent as U.S. aggregate data,
but state-specific
(idiosyncratic) consumption exhibits substantially less sensitivity or
lagged sate-specific income – a result that also holds for Canadian
provinces. They propose the
following interpretation: borrowing and lending in response to changes in
consumer demand are easier for individual U.S. states than for the United
States as a whole, and therefore, the measured deviation from the
benchmark permanent income hypothesis model is smaller.
However, lagged state-specific variables help predict
state-specific consumption, suggesting that the PIH model still requires
qualification.
Sorenson
and Yosha (Tel-Aviv) published ‘Is state fiscal policy asymmetric over
the business cycle?’ in the Kansas
City Federal Reserve Economic Review (Third Quarter, 2001). A number
of stabilizers are thought to mute the business cycle.
One key stabilizer is federal fiscal policy.
The federal budget surplus tends to rise during economic booms and
fall in downturns, helping to stabilize consumers’ disposable income and
thereby mitigate economic fluctuations.
During booms, for example, the budget surplus typically rises
because tax-revenues rise more than expenditures.
Another stabilizer that has traditionally received less attention
is state fiscal policy. Like
the federal budget surplus, state government surpluses end to rise during
economic expansions and decline during downturns.
For instance, Nebraska’s budget surplus rose from $91 per capita
in the recession year 1990 to $326 (in, 1990 prices) per capita in 1998,
when the economy was booming. Moreover,
like the federal budget, state budgets represent large shares of the
economy. For example, in 1998
state government expenditure was 10 percent of gross state product in
Kansas, and 9 percent in Missouri. The
stabilizing influence of state fiscal policy, however, may differ across
business cycle expansions and downturns – making statistical policy asymmetric.
For example, state budgets could be more effective at mitigating
economic slumps than at muting booms if taxes fall more sharply during a
slump than they rise is an expansion of equal magnitude. Asymmetry in
fiscal policy could be caused by a number of factors such as balanced
budget rules, which are constitutionally imposed restrictions on a state
government’s ability to incur debt.
This article examines the business cycle behavior of state fiscal
policy to determine whether policy is asymmetric and, if so, to identify
the causes. The first section
of the article reviews the general business cycle behavior of state
budgets. The second section discusses some theoretical explanations of
asymmetry and then examines whether the fiscal policies of U.S. states are
asymmetric. The third section
explores whether stringent balanced budget rules are associated with
stronger asymmetry in state fiscal policy.
The article concluded that state revenue and expenditure display
significant asymmetry over the business cycle, with nearly offsetting
effects on the budget surplus. As
a result, state fiscal policy tends to mute economic booms to roughly the
same degree it mitigates slowdowns. The
asymmetries in revenue and expenditure appear to be associated with
balanced budget rules, although their fundamental causes cannot be clearly
identified.
In
“Monetary transmission in a small open economy” Ber, Blass and Yosha
(all Tel-Aviv) used the newly constructed data set described in Section
1.0 to study the effect of monetary policy on investment. They show that
in a small open economy like Israel, contrary to standard theory, monetary
policy affects investment. They also show that the impact is transmitted
trough the balance sheets of firms whose access to foreign funds is
relatively constrained.
Barrios
(Young Researcher, CORE) and de Lucio (Universidad Alcalá de Henares,
Spain) provide evidence concerning the positive impact of economic
integration on EU regions’ business cycles convergence by focusing on
two neighbouring countries: Spain and Portugal. They show that while a
rise in cross-country business cycle correlation has also been experienced
by other European countries, it has been relatively more pronounced for
Iberian regions. Econometric evidence suggests that the existence of an
administrative border, the economic size of regions and their industrial
structures can explain a substantial proportion of regional cycles.
Luca
Sala (Young Researcher, Tel Aviv) in “Monetary Transmission in the euro
area: a factor model approach” studies the heterogenous effect of common
monetary policy across different countries within the Euro area. Since the
European Central Bank (ECB) targets euro-wide aggregates, it is important
to understand what is the degree of cross-country heterogeneity of common
monetary policy. This is an
important problem, but also a difficult one.
The first difficulty arises from the fact that there are not enough
data on the history of the EMU and a feasible empirical project can only
be defined for past data, when monetary policy was decided at the country
level. The second difficulty is due to the econometric problem of
modelling the joint dynamics of macro variables for many countries.
To tackle the first problem, Sala carries out two kinds of
exercises. He first defines a
common monetary policy shock (equivalent to the ECB shock) as the German
policy shock. He estimates the latter on the basis of data on Germany and
identifies it for the other countries as the shock with “minimum
distance” with respect to the German’s one. Second, he performs a
counterfactual exercise assuming that, as in the EMU regime, monetary
policy has the same effect on interest rates.
To tackle the second problem he adapts the model developed by Forni
and Reichlin (Review of Economic Studies, 1998) to the problem.
That model, unlike VAR cross-country regressions, allows for the
estimation of shocks and impulse response functions without imposing the
arbitrary restriction that the impulse response functions of the shocks
are the same across countries. Luca
Sala adapts it creatively to analyse the possible heterogeneous effect of
common EMU monetary policy across the countries of the Union. The paper is
now being revised. Giannone (Young Researcher, ECARES), Reichlin (ECARES)
and Sala (Young Researcher, Tel Aviv) are still developing the project on
the dynamic of common European policy shocks across single European
nations. This is an extension of Sala’s work and is based on new
methodological developments described in the last Section.
Jacopo
Cimadomo (ECARES) in “The Effects of Systematic Monetary Policy on
Sectors: a Factor Model Analysis” studies the monetary policy
transmission to sectors in the US economy. Recent literature has stressed
the importance of focusing on disaggregate variables in order to
investigate if monetary shocks have asymmetric effects across economic
entities, if they change relative process and relative outputs, if they
alter the production structure of the economy.
This work proposes an approach to analyse the sectorial
sensitivity of the systematic
component of monetary policy, within a factor
model econometric framework similar to that of the one formulated by
Forni, Hallin, Lippi and Reichlin (2000), and by Stock and Watson (1999).
A factor model analysis allows to exploit the covariance structure
among hundreds of time series and to overcome some drawbacks of VAR
literature, when the focus is on disaggregate variables. He finds that
systematic monetary policy affect heterogeneously the economy: industries
with higher capital intensity of production processes and relatively
sluggish production prices are the most “vulnerable” to endogenous
monetary actions, industries closer to the final demand the least.
Ghio
(Young Researcher, Modena) and Pistoresi (Modena) examined the relative
importance of local and global externalities for the U.S urban
development, by using the dynamic factor analysis
proposed by Forni and Reichlin (2001). These measures are used to derive
indirect evidence on the role of local and global externalities. Their
findings suggest that there is not a single type of externalities
predominates in the explanation of the urban industrial development.
Petrongolo
(CEPR) and Wasmer (ECARES) have completed the work on regional spillovers
in the matching process between jobs and unemployed workers.
The
econometrics of space and time
Mion
(CORE) estimates a model of economic geography using a space-time panel
data on Italian provinces. Particular
attention has been devoted to address endogeneity issues that naturally
arises when dealing with both simultaneity and spatial data. Results are
consistent with the hypothesis that product-market linkages, coming from
increasing returns and trade costs. Those
linkages influence the geographic concentration of economic activities and
their spread over space is, contrary to previous findings, not negligible.
Forni
(Modena) and Reichlin (ECARES-ULB) in collaboration with Hallin (ECARES-ULB)
and Lippi (CEPR/Rome) have continue to investigate econometric techniques
for large panels of time series, which are potentially useful to study the
propagation of macroeconomic shocks and spillovers.
They have published several papers in international journals such
as the Economic Journal and the Journal of Econometrics.
In this year report, we can see that papers already cited last year
are either published or under revision. Their work on dynamic factor
models has been developed in several directions, including empirical
applications. They explored
the role of financial variables in forecasting, in a paper using a large
data set, consisting of 447 monthly macroeconomic time series concerning
the main countries of the Euro area, to simulate out-of-sample predictions
of the Euro area industrial production and the harmonized inflation index.
They find that financial variables do help forecasting inflation,
but do not help forecasting industrial production.
Together with Altissimo., Bassanetti,
Cristadoro and Veronese (Bank of Italy) they developed a real time
coincident indicator for the Euro Area business cycle, which uses the
information of a large data panel to obtain an indicator which, unlike
other methods used in the literature, takes into consideration the
cross-country as well as the within-country correlation structure and
exploits all information on dynamic cross-correlations.
Cristadoro, Forni, Reichlin and Veronese also proposed an index of
core inflation for the euro area which again exploits information from a
large panel of time series on disaggregated prices, industrial production,
labor market indicators, financial and monetary variables.
The indicator is shown to have a number of desirable
characteristics and to perform very well as a forecaster of the euro area
harmonized consumer price index at one and two years horizon, which is the
relevant horizon for the ECB monetary policy.
The published empirical work has been focused on short-term
macroeconomic issues, not central to the topic of this network.
However, the methodology has shown to be useful for studying many
issues central to this network. As
described above, the method has been adapted in work by Sala (ECARES),
Pistoresi and Strozzi (Modena), Ghio and Pistoresi (both Modena), Giannone
and Lenza, Emiris, d’Agostino, Cimadomo (all ECARES), Lamorgese and
Ottaviano (Sassari). In this respect, the new work on identification of shocks and
propagation in large panels (Forni, Lippi and Reichlin, “Opening the
Black Box: Structural Factor Models versus Structural VARs”) will have
interesting application for the study of the propagation of macro shocks
in the geographical space.
On
the last topic Reichlin (ECARES) has collaborated with students Giannone
and Sala from ECARES and Doz (Cergy-Pontoise) to explore different
methodological alternatives. Three
papers are of significance in that respect:
A.
“Tracking
Greenspan: systematic and unsystematic monetary policy reconsidered”
(with Giannone and Sala, CEPR W.P. 2002).
This paper is closely related to Forni, Lippi and Reichlin but
proposes a slightly different estimator beside showing how to perform
counterfactual exercises in dynamic factor models in large cross-sections.
This is the method used by Cimadomo (2002) cited above.
B.
“VARs,
common factors and the empirical validation of equilibrium business cycle
models reconsidered” (with Giannone and Sala, CEPR W.P. 2002).
The paper compares estimation performance for the impulse response
coefficients based on a VAR approximation to this class of models and an
estimation method that explicitely takes into account the restrictions
implied by the factor structure. Bias and mean squared error for both
factor based and VAR based estimates of impulse response functions are
quantified using, as data generating process, a calibrated standard
equilibrium business cycle model. They show that, at short horizons, VAR
estimates of impulse response functions are less accurate than factor
estimates while the two methods perform similarly at medium and long run
horizons.
C.
“A
GMM estimator for common factors in large panels of time series” (with
Doz and Giannone, in progress). This
paper proposes an alternative estimator to Forni et al. 2001 which is
shown to have efficiency properties.
Reichlin
(ECARES) was invited to the 2000 World Congress of the Econometric Society
in Seattle to present the work on dynamic factor models in large panels.
Her presentation is now published: “Extracting business cycle indexes
from large data sets: aggregation, estimation, identification”, in
Dewatripont, M., Hansen, P.L. and Turnowsky S. (eds) Advances in economics
and econometrics: theory and applications, 8th world congress of the
econometric society, Cambridge University Press, 2002.
Jorge
Rodrigues (Young Researcher, Sassari) in “Common
and idiosyncratic shocks to exchange rate volatility” extends the
Forni, Hallin, Lippi and Reichlin (2001) dynamic factor approach to the
analysis of GARCH time series and in particular to distinguish between
common and idiosyncratic shocks to exchange rate volatility. The
contribution to the financial literature is to allow the analysis of large
dynamic GARCH cross-sections free from prior beliefs on the conditional
mean of the data and to estimate a time-dependent common and idiosyncratic
components of exchange rate volatility. The good performance of the
approach is supported by a Monte Carlo simulation exercise. He considers
54 dollar-valued exchange rate returns observed monthly from January 1982
to June 1997. Results reveal that although the conditional mean co-move
strongly, the exchange rate conditional volatility is mostly
idiosyncratic. The method provides a useful framework for the analysis of
topic 2. (v) above, in particular in connection with the work by Emiris.
Giannone
(Young Researcher, ECARES) in “Testing for linear restrictions in large
panels of time series” develops a GMM type of test suitable for the test
of linear restrictions in large panels. This test is likely to be very
useful for verifying hypotheses in data sets where both the time and
geographical space matter, i.e. topics in Sections 2 and 3. This paper has
been revised with respect to the version cited in the previous report.
Giannone
with d”Agostino (ECARES) and Reichlin (ECARES) have worked on the
development of time varying parameter methods which would be used to carry
on the project on the evolution of real and financial market integration
in Europe. This project, as mentioned in the last report, was initially
designed to integrate work by Forni (Modena) and Reichlin (ECARES) and
Lamorgese (Sassari), Sorensen, Yosha (Tel Aviv).
Perleman,
Rodrigues (Young Researcher, Sassari) and Wasmer (ECARES) have now
published the work in which they develop a method for the identification
of clusters in the labour market.
Miscellaneous
Topics
Luigi
Guiso (Sassari), Paola Sapienza and Luigi Zingales (CEPR), (CEPR WP, 2002,
forthcoming Journal of Monetary Economics) in ”People’s
opium. Religion and economic attitude” study the effect of religious
beliefs and economic attitudes on a number of issues on economic
performance. Since Weber there has been an active debate on the impact of
religion on people’s economic attitudes. Much of the existing evidence,
however, is based on cross-country studies, where this impact is
confounded by differences in other institutional factors. We use the World
Values Surveys to identify the relation between intensity of religion
beliefs and economic attitudes controlling for country fixed effects.
We study several economic attitudes: attitude toward co-operation,
the Government, working women, legal rules, thriftiness, and the market
economy. We also distinguish across different religion denominations,
differentiating on whether a religious denomination is dominant in a
country. We find that on average religion beliefs are associated with
“good” economic attitudes, where “good” is defined as conducive to
higher per capita income and growth. Yet religious people tend to be more
racist and less favourable with respect to working women. These effects
differ across religion denominations, but not in the way emphasised by the
existing literature.
Chiara
Strozzi (Modena) and Julian Messina (European Central Bank) were involved
in a joint project which investigated the determinants of short- and
medium-run fluctuations of the labor share in European countries. They are
currently working on the formulation of a theoretical wage bargaining
model with the aim to analyze the adjustment of labour market variables to
business cycle fluctuations in the presence of trade unions and labour
adjustment costs. The implications of this model for the behaviour of the
labor share in Europe will be also tested empirically.
Chiara
Strozzi (Modena) has been working on a dynamic wage bargaining model aimed
at analysing the impact of an increase in product market integration on
the sustainability of tacit collusive agreements between trade unions
coming from different countries.
Bertocchi
(Modena) with Canova (UPF) completed a paper on growth, geography and
institutions focusing on the case of Africa as affected by colonization.
This work has now appeared in the European Economic Review. Bertocchi also
conducted research on institutions and growth in a comparative
perspective. In particular, she has investigated the role of labour
markets and unionization in an integrated world, and, in collaboration
with Spagat (CEPR) she has studied the interaction between growth and
education and welfare systems. A
recent project focuses on the interaction between the intergenerational
transmission of property rights and the sectoral evolution of the economy.
During the year under consideration some of these papers have appeared in
international journals such as the Journal of Comparative Economics and
the Review of Economic Dynamics.
Sanner
(Young
Researcher, Modena) has studied spill-over effects between imperfect
product and labor markets, developing a very general model to be applied
to many issues. Under a dual
labour market, he found that cross-country differences of the union wage
premium are mainly caused by two parameters: The degree of centralisation
of wage determination, and the size of the unionised sector of the
economy. Together with the technology employed, these two parameters also
seem to impact strongly on real income. In another project, he is
investigating the cross-market effects of (deregulation in goods and
labour markets. In particular, he considers regulatory measures
determining product market transparency, the bargaining power of workers,
and the bargaining structure. Preliminary results confirm and highlight
the requirement to take cross-market effects into account for an
assessment of deregulatory measures. Finally, in a joint project with
Felbermayr (European University Institute) he intends to examine
spill-over effects of labour market institutions on other countries. Part
of Sanner’s work on labour markets has appeared in The Geneva Papers on Risk and Insurance Theory.
Davide Ticchi (Young Researcher, UPF) has analysed the interplay between risk aversion and inter-temporal substitutability in a simple dynamic investment model with aggregate uncertainty (in his paper with Saltari, “Risk aversion, intertemporal substitution, and the aggregate investment-uncertainty relationship”). His main finding is that risk aversion cannot by itself explain a negative relationship between aggregate investment and aggregate uncertainty, as the effect of increased uncertainty on investment also depends on the inter-temporal elasticity of substitution. In particular, the relationship between aggregate investment and aggregate uncertainty is positive even if agents are very risk averse, as long as the elasticity of inter-temporal substitution is low. A negative relationship requires that the relative risk aversion and the elasticity of inter-temporal substitution are both relatively high or both relatively low. Ticchi also shows that the implications of our model are consistent with the available empirical evidence.
Return to Introduction
Research