Informality: Exit
and Exclusion
OECD Development Centre
Paris
July 2007
LAC not significantly more informal
than comparable countries…
Self Employment (% of Labor Force) 2/
Lack of Pensions (% Labor Force) 1/
2 Measures of Informality vs Income per Capita
100
90
80
70
60
50
40
30
20
10
0
-
10,000
20,000
30,000
40,000
60
50
40
30
20
10
0
-
10,000
Advanced Countries
30,000
40,000
2005 GDP per capita PPP adjusted
2005 GDP per capita PPP adjusted
Latin America
20,000
Rest of the world
Latin America
Advanced Countries
Rest of the world
…But, a disturbing trend in the nineties
% Informal Labor Force (Productive Definition)
% Informal Labor Force (Legalistic Definition)
11.9
Argentina 1992-2005
7.7
Venezuela 1995-2003
18.5
Venezuela 1989-2003
Uruguay 1992-2004
7.0
Honduras 1992-2002
6.8
5.4
Panama 1995-2003
Brazil (Met) 1992-2002
6.2
Nicaragua 1993-2001
5.9
Colombia 1996-2004
4.8
Peru 1997-2003
4.6
3.4
Jamaica 1996-2002
4.0
Mexico 1990-2004 *
1.1
Chile 1990-2003
-0.7
Ecuador 1994-1998
Brazil 1990-2003
-0.8
Paraguay 1997-2003
-0.9
-2.4
Colombia 1996-1999
3.1
Dominican Republic 1996-2004
Paraguay 1997-2003
2.6
Mexico 1996-2002
2.4
El Salvador 1991-2003
2.2
Argentina 1995-2005
1.9
Bolivia 1997-2002
1.5
Ecuador 1994-2003
1.1
Costa Rica 1992-2003
-0.4
Nicaragua 1993-2001
-0.9
-3.3
Brazil 1992-2003
El Salvador 1991-2003 -12.0
-5.2
Chile 1990-2003
-15.0
-10.0
-5.0
0.0
5.0
10.0
15.0
-10
-5
0
5
10
15
20
Why do we care?





Unprotected workers: issue of welfare, equity and
efficiency
Productivity: firms too small? Barriers to growth?
Indicator of regulatory distortions
Low and distorted tax collections and poor provision
of public services
Weak rule of law and Governance: A problem in our
social contract?
Exit and Exclusion

Traditional exclusion view of informality



Exit view: agents analyze costs vs. benefits of becoming
formal




Labor: inferior jobs in a segmented market
Firms: denied access to services by high entry costs (de Soto)
Labor: informality offers flexibility, avoidance of poorly designed
benefits programs, and provides “decent” work
Firms: don’t need/want State programs, don’t pay for them
Larger firms and individuals: Why pay taxes if can avoid?
Both exist to varying degrees across countries but have
different implications for policy
The Informal Worker
Comparative Advantage and Constraints
Two Distinct Types of Informal Worker
Distribution of informal workers in Latin America
% contributing to social security system
Peru 2002
Bolivia 2005
Ecuador 1998
Nicaragua 2001
Colombia 2006
Guatemala 2002
Dom. Republic 2006
Mexico 2002
Informal independent
Venezuela 2003
El Salvador 2003
Argentina 2005
Brazil 2003
Informal Salaried
Uruguay 2004
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
Chile 2003
% employed workers
% Not Contributing to Social Security System
Two distinct types of informal workers.
Role of Each Changes across Life-Cycle
Informal Salaried

Self employed
Cunningham (2007)
Formal
Informal
67
63
59
55
51
47
43
prevalent among prime or
older
have capital and skills to
open a business
39

35
Self Employment
31

27

50
45
40
35
30
25
20
15
10
5
0
23

port of entry for youth
accumulate experience for
Formal Salaried or
independent work
Most in micro firms
19

Brazil: Employment as share
of age cohort
15

Most informal workers found
in small firms
75% of Mexican and Argentine in firms of at most 5 workers
However, expansion in large firms in Arg, Bra over 1990s


% of Informal and Self Employed workers
Mexico
(Urban Areas)
50
45
40
35
30
25
20
15
10
5
0
1994
2004
1
2-5
6-10
11-15
16-50
Firm Size
51-100 101-250 250 +
% of Informal and Self Employed workers
The allocation of Informal Workers across firm size
Argentina
(Gran Buenos Aires)
50
45
1980
40
2003
35
30
25
20
15
10
5
0
1
2 to 5
6 to 25
26 to 100
Firm Size
101 to 500
501 +
Workers tell us of both exit and
exclusion
Self-rated Poverty Relative
to Formal Workers
Most (~2/3) of
independent workers are
voluntary



not queuing for formal
jobs
value flexibility, esp.
women
opting out of Social
Protection
odds ratio

1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Independent
Argentina
Informal
Salaried
Dominican Republic
Workers tell us of both exit and
exclusion
Self-rated Poverty Relative
to Formal Workers
Most Informal Salaried
are involuntary




would prefer to be FS or
SE
Brazil: 30% do not want
to leave
Exceptions Dominican
Republic, Mexico
youth: difficulty entering
workforce
odds ratio

1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Independent
Argentina
Informal
Salaried
Dominican Republic
And their transitions suggest both
voluntary entry in Mexico
.14
.12
.1
.04
.05
.06
Self-Employed to Formal
Formal Salaried to Self Employment
.16
.07
.18
Mexico
.08
.03
Self Employment to Formal Salaried
1987q1
1991q1
1995q1
Formal to Self-Employment
Bosch, Goñi, Maloney (2007)
1999q1
2003q1
Self-Employed to Formal
Cyclical Patterns of Labor Informality

.8
Mexico: Job Finding Rates:
Transitions from U to Work
What drives
countercyclically?

.6
IS
.4

FS
.2


0
SE
1987q1
1991q1
1995q1
q_unm_for
q_unm_se
1999q1
q_unm_inf
2003q1
However, new insights from
OECD literature on why
Job Finding in FS: volatile
and pro-cyclical
Job finding in the I Sector
Informal sectors constant.
Net: I-sector absorbs labor
in downturns
What Was Behind Trends in 90’s?



Boom in capital inflows and non tradable sectors in
the early nineties
Rise in female participation? (AR, perhaps, CO, BR no)
Trade Reform: Ambiguous effects




Some
in BR, AR; MX
Increased real minimum wages and/or more
restrictive labor regulations (CO, BR, PE)
Social security reform often lowered Benefit/ Cost
ratio (except in CH)
Temporary contracts and relaxed enforcement (AR)
Cyclical Patterns
Mexico
1.4
2.6
1.3
2.4
1.2

To Formal/Tradable
To Informal/Nontradable


1.2
0.5
Unconditioned Earnings F/SE
Real exchange rate
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1.0
1991
0.4
FS rigidities less binding in booms
Boom in nontradables:
REER appreciation due to boom in
capital inflows:
opening of Capital account, Exchange
Rate Based Stabilization, improved
expectations due to reforms
Formal/SE
Colombia
1.6
2.4
2.2
1.4
2.0
1.2
1.8
1.0
1.6
0.8
1.4
0.6
1.2
Countercyclical/Segmented
Unconditioned Earnings F/SE
Real exchange rate
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1.0
1985
0.4
Procyclical/Integrated
Formal / SE
1.4
0.6
Increase in Informality in early 1990s

1.6
0.7
Formal/SE
Formal / SE

0.8
1990

1.8
1989

0.9
1988

Show voluntary entry
Depend on positive shocks
2.0
1.0
1987

2.2
1.1
Relative earnings
Real exchange rate

Informality mostly countercyclical
But can be procyclical in some booms
Relative earnings
Real exchange rate

What Was Behind Trends in 90’s?



Boom in capital inflows and non tradable sectors in
the early nineties
Rise in female participation? (AR, perhaps, CO, BR no)
Trade Reform: Ambiguous effects




Some
in BR, AR; MX
Increased real minimum wages and/or more
restrictive labor regulations (CO, BR, PE)
Social security reform often lowered Benefit/ Cost
ratio (except in CH)
Temporary contracts and relaxed enforcement (AR)
The symmetric pattern breaks down in
Brazil after 1990
1
4
SE to FS
3
1.2
1.4
1.6
Self-Employment to Formal
1.8
5
Brazil
FS to SE
2
.8
Trade & Constitutional Reform
1983m1
1988m1
1993m1
time...
Formal to Self-Employment
Bosch, Goñi, Maloney (2007)
1998m1
2003m1
Self-Employment to Formal
Actual and predicted size of the formal
sector in Brazil
90
85
80
75
70
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
65
Actual
No trade
No Constitution
What Was Behind Trends in 90’s?



Boom in capital inflows and non tradable sectors in
the early nineties
Rise in female participation? (AR, perhaps, CO, BR no)
Trade Reform: Ambiguous effects




Some
in BR, AR; MX
Increased real minimum wages and/or more
restrictive labor regulations (CO, BR, PE)
Social security reform often lowered Benefit/ Cost
ratio (except in CH)
Temporary contracts and relaxed enforcement (AR)
The Informal Firm
Microfirm Dynamics:
Very similar in LAC and the US
Example: as in the US, self-employment increases with age
United States
60
50
50
Percent
Percent
Mexico
60
40
40
30
30
20
20
10
10
Age
0 18-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65
0 18-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65
Entry Rate
Exit Rate
Self-employment Rate
Age
Entry Rate
Exit Rate
Self-employment Rate
Other similarities:
 Entry increases with:



Education (firms with some employees)
Conditional wages of salaried workers (overachievers)
Firm exit and growth:


Higher for smaller microfirms (as predicted by Jovanovic)
Firm survival and growth increase with owner’s education, firm revenues
Why do the Majority of Microfirms
Formalize as they Grow?
0%
30%
60%
90%
Formality by firm size (Brazil)
0
1
2
3
4
5
Number of Paid Employees
Not Registered

No workers registered
Informal services & institutions may become insufficient:





No tax payment
Contract enforcement
Formal Credit
Sales to formal firms (e.g. if VA system)
Access to Government programs
…or just increased probability of detection
Detection does seem to play an important role
in driving firms to formalize (IFC surveys)
Main advantage of Formality (%)

Main motivation seems to be
risk of being caught:


Comply with the law
“comply with the law” (47%)
avoid bribes, fines (30%)
47
Avoid paying fines
20
Expand and Gain new clients
14
Avoid paying bribes

Access to markets/services less
important (<25%)
9
Improve access to credit
8
Improve contract enforcement
2
0
10
20
30
40
50
Source: IFC (Bolivia, Peru, Brazil, Honduras and Nicaragua)
And Small Firms often say they do not
formalize because “they don’t need to”

The DR and Mexico over 60%
and 80% resp. say “they’re too
small” or “they don’t need to
formalize”
Why do microfirms not register?
38
Argentina
63
84
Mexico
 Probably low detection
probability, lack of demand for
formal services

Argentina: somewhat different
picture, for 63% main
motivation are formality costs
14
63
Dominican
Republic
27
0
20
40
60
80
No need to register/business is too small
Formality costs
100
But note that informality is not just a
small firm phenomenon


Medium and Large Firms
evade considerable Taxes and
Social Security
Firm level correlates of
evasion:






firm size (-)
Productivity (-)
corruption (+)
rigidity of labor regulations (+)
linkages to large companies (-)
use of third party contract
enforcement (-)
Tax and Social Security Evasion (%)
20 and under
21 to 99
100 and more
Chile
20 and under
21 to 99
100 and more
Colombia
20 and under
21 to 99
100 and more
Mexico
20 and under
21 to 99
100 and more
Brazil
0
5
10
Tax
15
20
25
30
Social Security
35
40
45
Recent Evidence on Policies to reduce
Red Tape and Taxes

Mexico’s SARE program:


Business registration in less than 3 days
Positive Impact on formality but effect is of small
magnitude



Bruhn (2007): 5.6% increase in fraction of registered businesses
but mostly increased entry salaried workers
Kaplan et al. (2007): increase is of only 0.7%
Brazil’s SIMPLES program:

Tax burden reduction: 8% of revenues


Monteiro & Assunção (2005): 46% increase in registration
rate for retail firms…at least in 1st year after SIMPLES
Evidence of improved firm performance as a result of
formality
Impact of informality on Firm
Productivity?

Lower economies of scale as
firms remain small:



“Unfair” competition by low
productivity informal firms?



To avoid detection
Because of low access to
markets and services
Firm Productivity in High Evasion Areas
Labor
Productivity
Total Factor
Productivity
-12
-10
-8
-6
Unregistered Workers
-4
-2
0
Unreported Sales
Slows Schumpeterian creative destruction
Reduces market/investments of formal firms?
Efficiency gains thanks to avoiding regulations?

Example: evidence of higher firm productivity associated with lower
enforcement of labor regulations (Almeida and Carneiro, 2006)
Impact of informality on growth?

Difficult to identify formality
effects:

Regulation, education, etc.
affect both informality and
growth, but...
Impact of Informality on Growth
Cross country regressions
Controlling for:
Informality - % of GDP
Self Employment
None
Initial Education

Informality itself has negative
but insignificant effects on
growth
Regulation
Financial Depth
Investment
Corruption

Future research: how does the
impact of informality on growth
vary according to the specific
factors that drive informality
All
-.1
-.05
Estimated Impact
0
.05
-.1
-.05
95% Confidence Interval
0
.05
Social Protection
Protecting all LAC’s citizens against risk
Andrew D. Mason
Senior Economist
Informality and Social Protection –
Why do we care?



Lack of access to social protection has welfare, equity and
efficiency implications
While individuals/families employ private risk management
strategies – informally (prevention, self-insurance) and
somewhat through the market (market insurance) …
… Lack of access to formal social protection raises important
concerns on several fronts

From the perspective of household welfare


From the societal perspective


Inefficient coping, impoverishing shocks (particular concerns in health
and old age security)
Too much uninsured risk can have external costs, productivity effects
On the social protection system

Small, inefficient risk pools, foregone savings
Social Security Coverage is Low and
Stagnant – or Declining – in Many Countries
Coverage Rates for the Economically Active Population
(% of economically active)
80
70
60
50
40
30
20
10
0
BO
PY
PE
NI
GU
CO
EC
MX
1990s
Source: Rofman and Luccetti (2006)
SA
2000s
VE
AR
BR
UY
CL
CR
LAC’s “Truncated Welfare State” –
The Poor Have Largely Lacked Access
Pension Coverage by Income Quintile (% of economically active)
Percentage
100
90
80
70
60
50
40
30
20
10
0
Peru
Poorest Quintile
Colombia
Second Quintile
Mexico
Argentina
Middle Quintile
Brazil
Fourth Quintile
Chile
Wealthiest Quintile
Persistent “Truncation” Helped Spawn the
Rise of Social Assistance Programs
Over the last decade,
LAC countries have
launched a variety of
programs to better reach
the poor, e.g.:





CCTs
Social assistance
pensions
Subsidized health
insurance
Subsidized housing
etc.
Absolute Incidence of CCTs in Brazil and LAC
Absolute Incidence - CCTs in Brazil and LAC
(QUINTILES)
% of total benefits
received by each
QUINTILE

BR-Bolsa Familia
PNAD 2004
CL-SUF/Solidario 2003
80%
70%
60%
50%
40%
30%
20%
10%
0%
MX-Oportunidades 2002
AR-Jefes 2003
Q1
Q2
Q3
Q4
Poorest to richest (QUINTILES)
Q5
Government Failures are Pervasive and
Create Incentives for Informality


Government failures exist on at least two levels …
Weak design of Social Security programs





High costs via payroll contributions; rigid one-size-fits-all approach
(young vs. old; paying twice for health insurance)
Excessive “bundling” of benefits (health, pensions, disability, … but also
housing, childcare, sports & recreation) taxes contributors
Weak accounting for labor mobility (too-high vesting periods for
pensions, intermittent coverage in health)
Low quality of services/benefits
Lack of Incentive Compatibility within the “System”

Noncontributory assistance for informal workers increasingly
“compete” with Social Security reinforcing incentives for informality
The Design of Social Security is Inconsistent
with Movement In-and-Out of the Formal Sector
In Mexico



Low wage (affiliated)
workers spend less than half
their time in the IMSS
system, on average
Would take over 50 years of
qualify for a minimum
pension
In Uruguay

Less than 30 percent of
workers are estimated to
attain vesting requirements
by age 65
Workers’ Years Insured by IMSS,
Mexico, 1997-2005
Share of Workers (in percent)

45
40
35
30
25
20
15
10
5
0
1
2
3
4
5
6
7
Years
Workers with < 3 MW Workers with > 3 MW
8
9
Re-engineering Social Protection
to Protect all Citizens


Recommendations grounded in the economics of insurance
(Comprehensive Insurance Framework) and in public economics (the
appropriate role for public intervention)
In the long-term, ensure

Universal “essential cover” in health, de-linked from the labor contract, and
financed by general taxation



By facilitating widest possible risk pool, also enhances efficiency of the insurance
model
Brazil-Fernandes, Gremaud and Narita (2006): replacing labor tax on first
minimum wage with neutrality reduces informality (1.5%) if capital investments
are exempt
Poverty prevention pensions, targeted to the poor, and incentive compatible
with multi-pillar systems


Both pooling and savings components are important (complementary)
Targeted social assistance pensions for “large gap” countries; incentive compatible,
integrated systems (a la Chile) for “smaller gap countries”
The Role of Institutions
and Social Norms
Towards an Inclusive Social Contract
Informality Partially Reflects
How Citizens Relate to the State
Self Employment and Quality of Institutions
(correlations controlling for GDP)
Government Effectiveness
.2
BOL
COL
PER
HON
GUA
JAM
.1
CHI
-1
-.5
0
.5
Goverment Effectiveness
coef = -.07366668, se = .01583045, t = -4.65


CHI
COS
PAN
SAL
-.2
-.2
-1.5

ARG
COL
PER BOL
HON
GUA
JAM
ECU
BRA
MEX
-.1
BRA COS
MEX
PAN
SAL
-.1
0
ECU
ARG
VEN
0
VEN
Share of Self-employed
.1
.2
Rule of Law Index
1
-1
-.5
0
.5
1
Rule of Law Index
coef = -.07296094, se = .01690364, t = -4.32
Poor, untrustworthy institutions increase the costs, decrease benefits of
formality
High informality limits the action of the state
Another vicious circle?
Institutional Performance also Shapes Citizens’
Perceptions about Effectiveness and Fairness of the State
Share of firms which are confident
that the judiciary will enforce
contractual and property rights
(controlling for GDP)
.2
0.9
Correlation of informality and the
Perception of impartiality of courts
0.8
0.4
0.3
0.2
PER BOL
HON
GUA
0
ECU
ARG
COL
BRA
MEX
PAN
SAL
JAM
CHI
COS
-.1
0.5
VEN
0.1
0
Middle East &
North Africa
OECD
East Asia &
Pacific
Sub-Saharan
Africa
Europe & Central Latin America &
Asia
Caribbean
South Asia
-.2
Percentage
0.6
.1
Share of Self-employed
0.7
-4
-2
0
Impartiality of Courts
Coef.=-0.0262,
coef = -.02625821, t=-4.3
se = .00611322, t = -4.3
2
4


Lead to exit and a “culture
of informality”
Symptom of a
dysfunctional” social
contract?
PAR
.1
BRA
NIC
HON
ARG
COL
0
Ex: tax morale is negatively
correlated to perceptions of
State capture
SAL GUA
CHI
MEX
URU
-.1

COS
VEN
PAN
ECU
BOL
PER
-.2

Collective perceptions of
fairness/efficacy of state
Strong reciprocity: I’ll
comply if others comply
Tax Morale

.2
“Social Norms” of Compliance:
Perceptions of State, of Each Other
-.2
-.1
0
Run By a Few Interest
coef = -.80478831, se = .24120409, t = -3.34
.1
.2
Is Inequality Related to a High Informality /
Weak Institutions Equilibrium?
Informality vs. Inequality
20
40

ECU
NIC
VEN
0
URU
PER PAR
ARG BOL
MEX
GUA
DOM
HON
SAL
BRA
COL
CHI
-20
JAM
COS
High inequality correlated
with:


PAN
-60
-40

-20
-10
0
Gini Coefficient
coef = .71532765, se = .20185487, t = 3.54
10
20

Unequal access and
exclusion
Differences in power, voice
and influence
State capture by the rich and
organized middle classes
and low state legitimacy
But, Chile: strong
institutions, low informality
Policies to Reduce Informality

Reduce opportunity cost of informality: raise productivity in
the formal sector




Remove segmenting distortions in labor markets where
applicable
Tilt the benefit/cost ratio of firms and workers to opt for
formality:




improved Investment Climate
higher human capital accumulation
removing distortionary incentives in Social Protection Systems,
reducing cost of doing business for all firms
improving services associated with formality especially for SME’s,
Improve quality and fairness of institutions and policies and
even handed enforcement: move social norms towards a
culture of compliance. A potential virtuous circle