Land Rights Systems for Shared Prosperity: on jobs, rents and
power Stefan Dercon DFID and Oxford University
Slide 2
Property Rights are Good Acemoglu and Robinson: Property rights
key for growth, and a feature of inclusive institutions.
Acemoglu/Johnson (JPE, 2005). Property rights institutions that
protect citizens against expropriation and powerful elites first
order importance for long-run growth
Slide 3
Property Rights are Good High Level Panel to increase by x% the
share of women and men, communities, and businesses with secure
rights to land, property, and other assets DFID Result Framework
for current 5 year period Number of people supported through DFID
to improve their rights to land and property --- 6 million David
Cameron Golden Thread of Development the economic, political and
institutional building blocks for development Property rights one
of the essential features
Slide 4
Property Rights are Good What is it about them? Efficiency
(incentivises effort and investment) Equity reasons right/power,
including the poor and for women Leading to Interventions: To
ensure rights are given and protected, including for women or other
marginalised groups That are presumed to lead to growth and shared
prosperity
Slide 5
Are we asking the right questions? Land titling programmes have
an opportunity cost, so are we giving enough thought about: Can we
do it? Do we have the institutional conditions that will allow land
rights programmes to lead to equitable, efficient outcomes? How to
handle property rights so that are right for growth? Good for job
creation? Good for structural transformation? What land system is
required for job creation? Can we do within the state capacities we
have? Do we create new opportunities for rent-seeking
behaviour?
Slide 6
Feeding into the key questions: Are we building sustainable,
transparent land systems that support growth, jobs creation,
poverty reduction and equal opportunities? Are we
learning-while-doing whether they actually work deliver what they
claim to deliver? Or are we creating
Slide 7
Big questions, partial answers Can we design programmes that
lead to equitable outcomes? A story from urban Tanzania: a nudge to
gender equality in land titles; but is female empowerment a
fantasy? Are property rights going to lead to shared prosperity
through growth and jobs? A cautionary tale from Ethiopia (with
surprising facts) and nods to Singur (West-Bengal), historical
France and England Are land titling systems fit for purpose? The
ambitious protective land laws offering land rights including to
the poor in Dar es Salaam, Tanzania and the sobering tale of the 12
steps to land titles that cost of fortune to follow, if ever with
success and with few lasting effects....
Slide 8
Not just stories Results based on studies using RCT (Tanzania)
and discontinuity design (Ethiopia) Daniel Ayalew Ali, Matthew
Collin, Klaus Deininger, Stefan Dercon, Justin Sandefur and Andrew
Zeitlin (2014), The Price of Empowerment: Experimental Evidence on
Land Titling in Tanzania, mimeo Matthew Collin, Stefan Dercon,
Hunter Nielson, Justin Sandefur, Andrew Zeitlin, (2014), The
practical and institutional hurdles to obtaining land titles in
urban Tanzania, IGC issues paper. C. Anthony Harris, Forced to sell
the farm: Effects of expropriation and compensation programs on
small-scale farmers in Ethiopia URB-03: EXPROPRIATION March 24,
15:00, MC 2-800
Slide 9
1. The price of empowerment Tanzania 1999 Land Act: Direct
influence of Hernando de Soto Expanded scope for ownership in
urban, unplanned settlements Two types of titles: Residential
License (RL) and Certificate of Right of Occupancy (CRO) RL is
short term (5 year) lease on land, renewable, non-transferable, no
collateral value CRO is long term (99 year) leasehold,
transferable, can be used as collateral, perceived as more secure
But by now: less than 3% of households living in unplanned
settlements have obtained CROs
Slide 10
Female Inclusion in Land? In principle: Land act has several
provisions to protect women (default is female co-ownership, and
spouse need to have each others permission) In practice: only 45%
of men report that their wives will be consulted in selling even if
they have CROs, in general, less than quarter of RLs have female
name on it. Research programme: 1. What drives general low uptake
of CROs? 2. Can we induce greater legal protections for female
household members? 3. (in due course) What is the impact of CROs
and female inclusion in CROs?
Slide 11
Design: RCT Urban Dar es Salaam 2 neighboring sub-wards
(mitaa), roughly 1,000 parcels each Focus on one today: Mburahati
Barafu
Slide 12
The study area 2 neighboring sub-wards (mitaa), roughly 1,000
parcels each Focus on one today: Mburahati Barafu
Slide 13
Intervention (surveying plus repayment programme) randomised to
blocks 50 `blocks' of roughly 40 parcels each Randomly assigned to
treatment or control group Treatment: cadastral surveying and
repayment program
Slide 14
Intervention: price variation Parcel-level randomization of
prices: Baseline price of C.R.O. post-cadastral survey is TSh
100,000 (< $50)
Slide 15
Unconditional vouchers Some vouchers are `general' and act as
immediate discounts.
Slide 16
Some vouchers are conditional, and only apply if HH includes a
woman as an owner on CRO application Conditional discounts gender
conditions
Slide 17
Variation in conditional and unconditional vouchers Some get
general, some conditional and some both vouchers of various
values
Slide 18
Leading to price variation
Slide 19
Subsequently correlated with uptake We find high sensitivity to
price But do gender conditions matter?
Slide 20
Resulting into many more women registered on titles through
conditions Figure shows estimates of co-titling (probability of
woman to be included on actual title when application is submitted)
We find that co-titling rate shoots up from about half to close to
100% when conditional voucher is received instead of general
voucher.
Slide 21
But gender conditions dont come at a cost in terms of uptake.
Estimated takeup impacts of 0.379% and 0.377% per TShs 1,000 of
conditional and unconditional discount. They are statistically
indistinguishable.
Slide 22
Conclusion 1: land rights for the poor? Can we design
programmes that lead to uptake of titles for the poor? YES WE CAN
up to a point poor people are willing to pay for it, but make sure
it is not too much
Slide 23
Conclusion 1b: land rights for women? YES WE CAN - WE ARE IN
SILVER BULLET LAND: Gender conditions have no cost in terms of
uptake for a given price so simple nudges have huge impact on
female empowerment in land titling programmes. But lets be cautious
does it mean anything for women to have their name on the title? Do
we have the institutional conditions that will allow land rights
programmes to lead to equitable, efficient outcomes? (to be
continued)
Slide 24
2. Are land rights good for jobs? Jobs will come from
industrialisation and urbanisation Are we setting up land systems
supporting structural transformation?
Slide 25
Some lessons from history The economists view Industrialisation
could take off thanks to property rights (Acemoglu and Robinson)
and happened in UK thanks to Glorious Revolution of 1688. But
property rights by then at least as secure in France or even in
China by then (Allen, 2009). In France probably too secure: e.g no
profitable irrigation or canals in Provence as no way of overriding
private property rights; In England, parliament could override
private rights. A key part of development is ability to ensure land
can be used for higher productivity use and for public
infrastructure. A transparent system of protecting and at times
overriding private rights for the public good. (Eminent domain,
compulsory purchase, expropriation systems).
Slide 26
What (not) to learn from India Singur (West-Bengal) 400 ha
agricultural land taken in 2006 to be offered to Tata for the
production of the Nano car Forcible acquisition under an old
(British colonial) law compensation paid Protest and violence, and
international news coverage, and huge political fall-out
Interpretation of law rejected by courts Tata Motors left in 2008
New law very bureaucratic and slow? Not solving bad land records
Discussion March 26: RES-21 8:30 Papers by Dilip Mookherjee and
co-authors -Inadequate or inappropriate compensation/poor valuation
Comments on new law Naresh Saxena (new law anti-farmer and anti-
growth)
Slide 27
Solving this, Ethiopia style Bidding to attract industries Land
systems more like China than Africa All land belongs to the state
Constitutional user rights (with commensurate land registration
programmes) But government can take away with clear compensation
rules also for economic purposes
Slide 28
Ethiopias non-Singur 300 ha agricultural land, near Kombolcha
taken away to give way to Steel Plant, 2013 Land taken away
according to law, with compensation 10 years land productivity (not
quite headline news) C. Anthony Harris, Forced to sell the farm:
Effects of expropriation and compensation programs on small-scale
farmers in Ethiopia URB-03: EXPROPRIATION March 24, 15:00, MC 2-
800
Slide 29
Study Area: Kombalcha Survey well before land was taken away,
and afterwards. Red=treatment; green=control
Slide 30
Impact of expropriation On average: households affected lost
50% of their land received 100,000 Birr (the equivalent of 5,200
USD). Compensation received equivalent of USD 8,000 for unirrigated
land USD 16,000 for irrigated land (compared to USD
25,000-36,000/ha in stated policy Singur) Our data suggest in line
with regulations and principles (10 years productivity of
land)
Slide 31
Impact of expropriation Compared to those who did not in
diff-in-diff setting (note: average compensation=100,000 birr)
...increased savings by 58,000 Birr ...increased value of home by
9,000 Birr ...increased consumption by 6,200 Birr (possibly spent
more than that to compensate for income loss) ...increased
non-agricultural assets by 6,200 Birr ...do not change the value of
livestock holdings ...increased non-farm activity
Slide 32
Are they worse off? Those who lost land have consumption higher
than control group (6,200 birr per annum more despite income loss)
have less trust in local officials are indifferent in Cantrils
ladder of life (overall, how satisfied are you with your life,
using 10-steps) And expect to be better off than control group in
five years from now (subjective poverty scale) But not clear where
earnings will come from in data (high liquid savings, consumption
but limited off-farm or other investment)
Slide 33
Conclusion 2: land rights for jobs? Countries need to have a
transparent, fair system to transfer land from one use to another,
including where the market may suffer from hold-up problems For
public purposes For industrialisation and job creation? Handing out
individual land rights can protect the poor, but it should not be
fundamental hindrance of essential structural transformation,
required for poverty reduction Systems will have to be nested in
local institutional and legal contexts. Important efficiency and
equity trade-offs that must be confronted, and researched.
Slide 34
3. Are land right systems fit for purpose? Providing and
enforcing land rights potentially bureaucratically intensive
process E.g new land law in India: may now take years to get rights
as well as to conclude expropriate from farmers, leaving all in
limbo In countries where much work is needed on land titling, are
we setting up functional real institutions or just systems not fit
for purpose? Matthew Collin, Stefan Dercon, Hunter Nielson, Justin
Sandefur, Andrew Zeitlin, (2014), The practical and institutional
hurdles to obtaining land titles in urban Tanzania, IGC issues
paper.
Slide 35
Experience with CROs in Tanzania After 15 years, only 3% of
households in urban unplanned settlements with titles At least 12
distinct steps each with risks of hold-up, cost escalation,
corruption, and confusion and with considerable tax uncertainty
Even in our RCT, started in 2010 we are still trying to ensure
those who paid are getting the CRO
Slide 36
The sobering 12 steps to a title Step 1: Hire a Surveyor Step
2: Surveyor Prepares a Base Map Step 3: Submitting the Base Map to
Municipality Step 4: Submitting the Base Map to Ministry Step 5:
Completing the Cadastral Survey & Placing Beacons Step 6:
Approval of Survey by Municipality/Ministry Step 7 Receipt of
Unique Plot Number Step 8: Completion of CRO Forms and Creation of
Deed Plan (five forms and a proof of payment) Step 9: Deed
Plans/CRO forms Submitted to Municipality Step 10: Deed Plans/CRO
Forms to Ministry Step 11: CRO Approved for Collection Step 12:
collect the CRO
Slide 37
Cost? Cost prohibitive, not just in time Official route
(without extras) $23 for plot of 20m by 20m (400 sq metres) In
practice, private route required costing up to $712, before back
payments of tax and extras. A system fit for bureaucrats and
rent-seekers, and not for poor people? People are willing to pay
for title but at a reasonable rate for the poor
Slide 38
Conclusions Our examples showed: That we can easily nudge men
to ensure women get onto titles but can we be confident that this
means real empowerment? That protecting land rights is surely good
but do we get systems that are fit for a dynamic economy and that
will promote jobs and structural transformation? That it is easy to
get top-heavy bureaucratic systems with much rents but no impact
for the poor They have in common: a deep understanding of the local
institutions norms, rules, legal system, bureaucratic culture will
be required to be successful
Slide 39
Conclusions (2) We tend to build land rights systems as they
work in other countries or in imagined histories. But do we think
enough about whether they will work in a particular setting,
contributing to solving the problems of growth, poverty reduction,
structural transformation and exclusion? Isomorphic mimicry Just
like tropical frogs that just look like poisonous frogs, we build
institutions (including land rights systems) that just look like
effective institutions, but are they?
Slide 40
Conclusions (3) Land rights systems need to work Contribute to
growth and poverty reduction Contribute to efficient and equitable
outcomes; protect and promote development They function within
institutional settings Norms and practices on exclusion, e.g. by
gender Legal and other institutional settings Bureaucratic cultures
and capacities At times, our programmes risk trying to build
institutions, without impact, wrong for particular settings We need
good land governance systems with strong institutional assessments
and transparent monitoring of rights and, when it happens, of
expropriation.
Slide 41
How to guard ourselves? Diagnose: Are land rights and the way
we offer them a real solution for the problem one tries to solve?
Also in the long run? Monitor: Does the programme work (i.e.
achieve outcomes and change) as one thought it would in this
context? Build up a real evidence base Evaluate: Learn from success
and failure, openly and transparently with strong evidence. And
dare to ask, all the time: Is this just a fad? A white elephant? A
great new rent-seeking opportunity for some? Is this just
superficial change? Top heavy, bureaucratic? Is this really
contributing to structural transformation? To Jobs? To poverty
reduction? To inclusion?