Temporary contracts and
innovative investments: are they
substitute or complements at the
firm level?
Marco Malgarini (ISTAT)
MASSIMO MANCINI (ISTAT)
Lia Pacelli (UNIVERSITY OF TURIN AND LRR)
ISTAT, Roma, 20-21NOvember, 2011
OUTLINE
•
•
•
•
•
Aim
Background
Data
Empirical Strategy
Results
Istat, Rome, 21-22 November, 2011
AIM
Offer a contribution to the debate on labour
market flexibility (i.e. temporary contracts) and
its relationship with innovation (at a firm level)
• providing new insights on complementarity or
substitutability
• different kinds of labour contracts (more or
less flexible).
• between different kinds of investment (more
or less innovative)
 Are innovative investments inside the firm
accompanied by a specific kind of workforce?
 Which are the factors underlying investment and
hiring decisions?
•
Istat, Rome, 21-22 November, 2011
Background
 Flexibility debate:

Positive effect of flexible labour on firm
performance due to a decrease in labour cost
and an increased ability to innovate and
compete on the global markets.

Negative effect of deregulation of the labour
market due to a decrease in investment in
human capital and hence performance and
competitiveness in the longer run.
Istat, Rome, 21-22 November, 2011
Background
 Large part of the economic literature in the
‘90s and early 2000’s states that LM
flexibility (i.e. temporary contracts), allows
firms to adapt more rapidly to fluctuations of
demand, increasing marginal efficiency and
determining productivity gains via a
reduction of labour hoarding.
 The human resource managerial literature
challenged this vision (Hailey, 2001 for a
survey), underlying that a high turnover of
employees hinders the development of new
ideas and hence of innovation.
Istat, Rome, 21-22 November, 2011
Background
 From an empirical point of view, evidence on the
relationship between flexible contracts and (labour
or total) productivity is rather mixed.
 On the other hand, the relationship between
innovation and productivity is a well established,
positive one.
Istat, Rome, 21-22 November, 2011
Background
 If flexible contracts are found to be
positively related with innovative
investment, they may also be considered
to be productivity-enhancing. And
viceversa.
 Is this a stable relationship? Does it
change with the business cycle (or the
crisis)?
Istat, Rome, 21-22 November, 2011
Our contribution
 To investigate the Innovative Investment-
type of contract and effect of firm
characteristics on investment and hiring
decision…
 … We estimate a recursive mixed-process
model postulating a chronological ordering
in the decisions to invest and to hire at the
firm level.
Istat, Rome, 21-22 November, 2011
Data
 Business tendency surveys generally collect
entrepreneurs’ and managers’ opinions on
current trends and expectations for the near
future, regarding both their own business and
the general situation of the economy.
 Firms are asked to report about:


the current situation of a given variable
their short-term forecasts on possible outcomes in the
near future (usually three months ahead).
 Questions are qualitative, in the sense that firms
have to report their opinions, usually choosing
among three different alternatives arranged on a
Linkert scale
Istat, Rome, 21-22 November, 2011
Data
 In the paper, we integrate firm-level data from the
Manufacturing Confidence Survey (ISA-001) with those
from the Investment Survey (ISA-002) and from the
Export of the Manufacturing Sector Survey (ISA-008)
 The three surveys are realized on the same panel of
roughly 4,000 manufacturing firms
 Starting from January 2011, the surveys are carried
forward by ISTAT, in the framework of the EUHarmonized Program of Business and Consumers
Surveys
 We also provide a longitudinal integration among ISA001 survey data available on a monthly, quarterly and
annual basis, the latter referred to, respectively,
capacity utilization and the labor market
Istat, Rome, 21-22 November, 2011
Data
 The data used in the paper cover the period 2006-
2010
 Considering the Manufacturing Confidence Survey
(ISA-001), in the paper we use:

Data available on a monthly basis:


Assessments and expectations on the level of demand, production,
cash availability, growth of unit labor costs (in percentage points),
inventories
Data available on a quarterly basis:

Plant utilization (in percentage points); obstacles to production;
working hours
Istat, Rome, 21-22 November, 2011
Data
 The manufacturing survey also contains an annual section
concentrating on the labour market
 The Labor Market Section (LMS) focus on the hiring behaviour of
firms
 It provides qualitative (and also quantitative) information about:






Stock of workers at the end of the year
Hiring
Distribution of hiring and stock by type of contract (i.e. fixed term
contracts, temporary agency work, trainees …)
Skills of the new entrants
Recruitment channels preferred by firms
Previous employment status of hired workers
Istat, Rome, 21-22 November, 2011
Data
 The Investment survey (Isa-002) provides twice a
year (April and October) both quantitative and
qualitative information upon:



Actual amount of investment in 3 consecutive years
Its shares for substitution/rationalization/Increase
production (With new or existing products)
Factors influencing the revision of investment plans
during the year
 The quarterly Export activity survey ((ISA-008)
contains information referred to:

Export share on total turnover (in percentage points);
competitive position on national and international markets
Istat, Rome, 21-22 November, 2011
Data description: Hiring
Type of contract
2006 2006 2007 2007 2008 2008 2009 2009 2010 2010
% all % all H % all % all H % all % all H % all % all H % all % all H
no hiring
only open ended
only fixed term
only training
only agency
Open ended and fixed term
Open ended and training
Fixed term and training
Other combinations
68.1
12.0
10.5
3.4
n.a.
3.3
0.8
0.8
1.1
Istat, Rome, 21-22 November, 2011
37.7
33.0
10.6
n.a.
10.5
2.7
2.4
3.1
71.4
13.5
7.8
2.4
1.1
1.4
0.5
0.4
1.5
47.3
27.3
8.3
3.9
5.0
1.8
1.4
5.0
70.8
12.9
8.1
2.0
1.2
1.9
0.8
0.6
1.7
44.1
27.7
6.9
4.1
6.5
2.7
2.0
6.0
79.7
8.2
8.1
1.1
0.8
0.7
0.1
0.3
1.0
40.4
39.9
5.5
4.1
3.4
0.5
1.6
4.6
77.0
7.3
9.4
2.0
1.0
1.0
0.4
0.1
1.8
31.7
40.9
8.7
4.5
4.2
1.8
0.4
7.8
Data description: Investments
investment in t-1
2006
2010
Investment in t No
Yes No Yes
No
73.5 19.7 78.4 40.8
Yes
26.5 80.3 21.6 59.2
Substitution
Innovative
Both
Istat, Rome, 21-22 November, 2011
54.2
26.0
19.8
48.3 76.9 68.2
16.4 14.5 17.6
35.4 8.6 14.3
Merged Dataset
 We start matching the 2896 firms in the
November 2006 investment survey and the
3816 firms in the December 2006 labour
market survey obtaining 2700 matches.
 We then match the 2700 firms to their
monthly questionnaires one year before, to
obtain the exogenous determinants of the
above decisions (exogenous variables).
Istat, Rome, 21-22 November, 2011
Merged Dataset
 After the match with the monthly
questionnaires collected in December 2005
and in January 2006 (3726 observations) we
are left with 2272 firms (78.5% of the
potential sample of 2896 firms). We then
repeat the same for the following years.
 Conditional probability of being out of the
final sample is marginally higher among
large firms (above 100 employees) regardless
of their hiring decisions.
Istat, Rome, 21-22 November, 2011
Hiring and Investment
2006
Hiring
Investment
No investments Substitution Innovative
Both
Total
6.8
1.7
1.6
0.6
10.7
9.5
3.2
2.9
1.9
17.5
67.5
14.8
12.9
4.8
100.0
Investment
No investments Substitution Innovative
Both
Total
2.9
1.2
0.7
0.5
5.2
77.0
13.3
7.3
2.4
100.0
No Hiring
Permanent
Temporary
Both
Total
2010
Hiring
32.3
5.1
4.9
1.4
43.7
No Hiring
Permanent
Temporary
Both
Total
Istat, Rome, 21-22 November, 2011
47.5
6.6
3.6
0.8
58.5
18.9
4.7
3.5
1.0
28.1
21.5
4.5
2.3
1.0
29.3
5.1
1.0
0.7
0.1
7.0
Empirical strategy
 Estimation
of a recursive mixed-process model
postulating a chronological ordering in the decisions to
invest and to hire.
 Interactions among different decisions are captured by
including sequentially the decisions on more “rigid”
factors into equation of the more flexible ones;
 The size of adjustment costs generate the length of time
that defines short/long run: longest for innovative
capital, shortest for flexible labour.
 The size of adjustment costs generates the sequence of
decisions embedded in our recursive model.
Istat, Rome, 21-22 November, 2011
Empirical strategy
 innovative capital: highest non-convex
adjustment costs (research, patents, search of the
right asset ...)
 permanent labour: high non-convex adjustment
costs (search and training, investment in human
capital)
 ordinary capital: low non-convex adjustment
costs (replace existing capital, production
disruption when installed ...)
 temporary labour: lowest non-convex
adjustment costs (search, no investment in HC)
Istat, Rome, 21-22 November, 2011
Empirical strategy
 Decisions on innovative investment will not
depend on adjustment of other production
factors.
 Permanent hiring will depend on innovative
investments
 Substitution investment will also depend on
permanent hiring
 Temporary hiring (last decision) will depend
upon the decisions that the firm has made
Istat, Rome, 21-22 November, 2011
Model

We estimate the following interdependent equations:
I *in n o v  X 11   1
H *p erm  X 2  2  1 I *in n o v   2
I *rep l  X 3  3   2 I *in n o v   3 H *perm   3
*
*
H temp
 X 4  4   4 I *in n o v   5 H *perm   6 I repl
 4

Alpha1 measures the likelihood of hiring perms given that the firms has invested in
innovative K, wrt firms that have not. Etc …

X are predetermined observable firm characteristics capturing the Q*-Q gap; X
includes measures of demand and production process characteristics registered at the
end of the year preceding the investment/hiring decisions.

Errors are assumed to be identically and independently jointly normally distributed
with an unrestricted covariance matrix S, to be estimated.
Istat, Rome, 21-22 November, 2011
Main results
 The model is estimated for all the years
2006-2010.
 Up to 2008: Positive relationship among
permanent hires and innovative investments
 Since 2008: positive increase in temporary
hires with innovative investments.

Possible interpretation: cyclical pattern.
 No correlation between hires of permanent
and temporary workers.
 Decision to invest in substitution investment
and innovative capital is constantly
correlated.
Istat, Rome, 21-22 November, 2011
Main results
 Probability of making innovative investments
 is lower for non exporting firms
 Is higher for firms not facing liquidity contraints
 Permanent Hires
 depend on Demand (not increase in labour cost).
 Presence of liquidity constraints
 decreases permanent hires
 decreases temporary hires too (surprising result)
Istat, Rome, 21-22 November, 2011
Main results
INSERIRE TABELLA
Istat, Rome, 21-22 November, 2011
Caveats
 The model is not dynamic, due to attrition in the
panel dimension (questa non so se dirla)
 Endogeneity and unobservable factors.
Istat, Rome, 21-22 November, 2011