Cities 31 (2013) 308–316

Contents lists available at SciVerse ScienceDirect

Cities
journal homepage: www.elsevier.com/locate/cities

Self-financing urbanization: Insights from the use of Town Planning Schemes

in Ahmadabad, India
Shishir Mathur ⇑
San Jose State University, San Jose, CA, USA

a r t i c l e i n f o a b s t r a c t

Article history: Through an in-depth review of the financing structures of two projects developed using land readjust-
Received 3 May 2012 ment (LR) in the State of Gujarat, India, this article provides insights into LR’s ability to selffinance urban-
Received in revised form 7 July 2012 ization. This ability rests on three intertwined factors. First, a rapid increase in the value of developed
Accepted 17 September 2012
urban land enables local governments to generate substantial revenue from the sale of reserved land.
Available online 22 October 2012
Second, the local governments retain the reserved land for a significant amount of time before selling
it, thereby benefiting significantly from increases in land prices. Third, a revolving fund system, wherein
Keywords:
the revenues from older LR projects fund infrastructure in new projects, helps to finance the up-front
Land pooling
Municipal finance
infrastructure costs and eliminates the need to sell the land early or to seek loans. Furthermore, by using
Infrastructure finance sensitivity analyses, the article shows that the timely accrual of betterment charges would bolster the
Developing countries revenue stream. Finally, the analyses highlight the concern that more land may be reserved than is
India required to fund the project and argues for more accurate estimates of the need for land reservation.
Ó 2012 Elsevier Ltd. All rights reserved.

Introduction community centers, parks and open spaces. Some land may also
be reserved for future sale. The remaining land reverts back to
Led by increases in two demand side factors—population and the original landowners in proportion to their original levels of
income, the demand for serviced urban land is rising rapidly in land ownership (Archer, 1992).
many developing countries. However, significant infrastructure LR can be initiated by the landowners (usually through a
backlog, poor fiscal capacity at all levels of government—local, landowner association) or by the local government. In either case,
state and national (Bird, 1999; Shah, 2004; Vera & Kim, 2003), the initiating agency can impose a fee on the landowners and/or
and landowner opposition to compulsory land acquisition (Lall, sell land to fund the costs associated with LR. These include: the
2009) limit the supply of serviced urban land. This demand-supply cost of preparing and administering the project, the cost of seeking
mismatch is the primary cause of the rapid rise in the prices of government approval, and the cost of providing infrastructure and
serviced urban land. In such a scenario, an urban infrastructure services (Yau & Cheng, 2009).
finance mechanism that is acceptable to the landowners, unlocks
the value of the land and allows for the capture of a part of the in- LR overcomes limitations of compulsory and negotiated land
crease in land value to finance public infrastructure and services is acquisition methods
welcome (Home, 2007; Krabben & Needham, 2008). Land readjust-
ment (LR) is one such mechanism (Hayashi, 2002; Turk, 2008). In developing countries, governments usually employ two
Indeed, LR has been widely employed for rural-urban land methods to acquire land for planned urbanization—compulsory
conversion, for example in countries such as Japan, South Korea, and negotiated land acquisition. However, both of these methods
Australia, Indonesia, Nepal and India (Archer, 1988, 1994; Hayashi, suffer from significant limitations. First, they require fiscally con-
2002; Karki, 2004; Lee, 2002; Li & Li, 2007; Sorensen, 2002), and to strained governments to expend significant up-front funds to ac-
redevelop urban areas in countries such as Holland, Germany, quire land (Cete, 2010). Furthermore, landowners often complain
Turkey, Japan and China (see Hein, 2010; Li & Li, 2007; Needham, about unfair compensation for their land, and hence oppose land
2002; Turk & Altes, 2011). In LR, a group of land holdings are acquisition. Even a small minority of landowners’ being unwilling
pooled to plan and provide public infrastructure and services such to sell their land puts the entire urban development project on
as roads, electricity, water and sewer infrastructure, schools, hold, a problem commonly known as the hold-out problem (Turk
& Altes, 2010). Moreover, these methods create inequities. Land-
⇑ Tel.: +1 408 310 7856; fax: +1 408 924 5872. owners benefit significantly without contributing to the project if
E-mail address: shishir.mathur@sjsu.edu their lands are not acquired (Cete, 2010). For example, such

0264-2751/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cities.2012.09.004
 S. Mathur / Cities 31 (2013) 308–316 309

landowners benefit from the future increase in the value of their ment both bear risks during the construction period (called con-
land due to the project (for example, a road construction project), struction risks). The government usually assumes a larger share of
while owners whose lands are acquired do not get the opportunity the risks because it often bears the entire up-front cost of providing
to benefit from such future increases in land values. Finally, nego- the infrastructure. The landowners’ construction risks and their
tiated land acquisition methods leave open the possibility that share of project costs increase if they pay a fee and dedicate land
some land owners may be able to negotiate smaller land deduction to fund the project.
than the others.1 Furthermore, LR may be vertically inequitable3 if the landown-
LR overcomes the above-described shortcomings of land acqui- ers are low-income (they are frequently small farmers) and have
sition methods. LR is equitable because landowners contribute very little ability to pay the fee up-front. Moreover, to the extent that
land and pay fees in proportion to the benefits received from the landowners have to sell land to provide their share of the LR project
project (Yau & Cheng, 2009). The equitable nature of LR also min- cost, such as in Finland (see Viitanen, 2002), they bear very high con-
imizes the hold-out problem (Turk & Altes, 2010). Finally, LR does struction risk, and receive smaller share of the future increases in
not require funds for land acquisition. land values. Therefore, it is customary to allow fee payment to be
made in installments (Gielen & Altes, 2007) and/or defer the fee until
Aim of the study project is complete or until the landowners apply for a change in
land use or a building permit.
The existing literature widely notes LR’s potential to self- The amount of the fee also impacts vertical equity. In a situation
finance urbanization (Cete, 2010; Karki, 2004; Turk, 2008; Yau & where a majority of the project costs is recovered from the fee paid
Cheng, 2009). However, only a few studies (see Archer, 1988; Karki, by the landowners, the amount of the fee can be very high and,
2004) provide high-level numbers to show how the projects are therefore, burdensome for low-income landowners. The literature
financed. None of the studies systematically document the project suggests several ways to keep the fee low. First, require the govern-
costs and revenues in enough detail to enable the reader to get a ment to bear the up-front project cost (Viitanen, 2002). Second,
complete picture of the project’s financial structure. Policy makers provide low-interest loans for the project (Turk, 2007; Yau &
and practitioners would not only like to know that LR can poten- Cheng, 2009), or third, provide no, little or low-quality infrastruc-
tially self-finance urbanization or that it has self-financed urbani- ture (Turk, 2008). However, governments in developing countries
zation in specific cases, they would also benefit from answers to are often fiscally constrained and credit markets are not well
questions such as: How large an area was developed using LR? developed. Therefore, opportunities for low-interest loans are lim-
What types of costs were incurred and revenues realized? When ited. Furthermore, the governments’ poor fiscal capacities limit
were these costs incurred and revenues realized? Which legal their abilities to bear the project costs up-front. Moreover, LR pro-
and administrative rules influenced the timing and magnitude of jects that provide no, little or low-quality infrastructure run the
the cost and revenue streams? What was the relative contribution risk of providing sub-standard or un-serviced urban land or delay-
of revenues from the sale of land? Could less land have been re- ing the provision of infrastructure (Cete, 2010).
served? Finally, how do factors such as construction delays and Finally, the project benefits are usually apportioned based
changes in the amount of land reserved for sale impact LR’s financ- either on the area or the value of the land owned prior to land pool-
ing structure? ing (Turk, 2007). There is an agreement in literature that the land
This article aims to answer these questions through an in-depth value based method is fairer, but it also contains warnings that
review of two projects developed using LR in the state of Gujarat, high-level real estate appraisal expertise is required to implement
India. Focusing on the cost and revenue streams, and the corre- this method fairly (Cete, 2010; Yau & Cheng, 2009).
sponding legal and administrative enabling environment, the arti-
cle identifies various factors that contribute to LR’s financial
success in Gujarat. Finally, sensitivity analyses are conducted to LR in India
demonstrate the financial impacts of time delays and changes to
the amount of land reserved for sale. In the absence of a national-level enabling legal environment,
the use of LR in India varies from state to state, is predominantly
used in the state of Gujarat, and to a smaller extent, Kerala,4 and
Literature review predates India’s independence from the British.5 First introduced in
the country through the Bombay Town Planning Act of 1915 in the
Turk (2008) provides a broad overview of the use of LR interna- erstwhile Bombay Presidency,6 LR was implemented through Town
tionally. Therefore, without duplicating Turk (2008), I focus on Planning (TP) Schemes. The TP Schemes were widely used in the first
reviewing the literature that examines how LR projects can be half of 20th century to urbanize Bombay (now called Mumbai), the
made more equitable for the two major stakeholders—the land- capital of the erstwhile Bombay Presidency and the present day state
owners and government, and hope that the literature review find- of Maharashtra. Large parts of the city’s suburban neighborhoods of
ings would provide insights into devising an equitable financing Mahim, Khar and Borivali were developed using TP Schemes. How-
structure for LR projects. ever, the use of TP Schemes declined when the Maharashtra Regional
and Town Planning Act of 1996 chose detailed Development Plans
Ways to enhance horizontal and vertical equity in LR projects (DPs) over TP Schemes. Furthermore, the TP Schemes suffered from
two major drawbacks. First, the long period between scheme initia-
Focusing on ways to enhance horizontal equity,2 the literature tion and final government approval (an average of 15 years) made
calls for balancing the benefits from a LR project with the costs
and risks associated with the project (Turk, 2008; Yau & Cheng, 3
Operationalized through the ability-to-pay (ATP) principle, vertical equity has its
2009). In a government-led LR project, the landowners and govern- roots in welfare economics. In public finance, the vertical-equity rule calls for the rich
to pay more than the poor for government-provided goods and services.
1 4
I would like to thank an anonymous referee for pointing out this drawback of Recent media reports from India suggest that following Gujarat’s success with LR,
negotiated land acquisition methods. several states are closely examining the feasibility of employing LR.
2 5
The beneficiary-to-pay (BTP) principle operationalizes the horizontal-equity rule India gained independence from the UK on August 15, 1947.
6
in public finance. The BTP calls for those benefiting from a public infrastructure or The Bombay Presidency included, among others, the post-independence states of
service to pay for it in proportion to the benefit derived. Gujarat (the entire state) and Maharashtra (the majority of the state).
310 S. Mathur / Cities 31 (2013) 308–316

the process very slow. Second, ownership disputes over a single par- it is called a ‘‘Sanctioned Preliminary TP Scheme’’ (Ballaney & Patel,
cel of land could hold up an entire scheme. These two drawbacks, in 2009). The third round of hearings focuses on the financial issues
conjunction with the shift in focus to the DPs, contributed to the (such as the betterment charge). The scheme is again modified
abandonment of TP Schemes in Maharashtra (Interview with V. and submitted for government approval as the ‘‘Final TP Scheme.’’
Phatak). Upon government approval, it is termed the ‘‘Sanctioned Final TP
Gujarat continued using TP Schemes, first through the Bombay Scheme.’’ The TPO has 1–2 years to prepare the Preliminary and Fi-
Town Planning Act of 1915, which it inherited after seceding from nal TP Schemes. The state government is given 2 months to ap-
the Bombay Presidency after India’s independence, and later prove the Preliminary TP Scheme and 3 months to approve the
through the Gujarat Town Planning and Urban Development Act Final TP Scheme. However, substantial delays occur in the post-
(GTPUDA) of 1976. Sanctioned Draft TP Scheme stage due to delays by both the TPOs
and the state government (Interview with K. Gajjar; Ballaney,
TP Scheme preparation process in Gujarat 2008). Therefore, a process that should ideally take 3–4 years
may take more than a decade.
The Notification of Intention (NoI) is the first step in the TP
Scheme preparation process. Issued by the appropriate local public
Key features of the TP Scheme preparation process
agency (typically the municipal corporation or development
Two key features of the TP Scheme preparation process have
authority), the NoI identifies the location and scheme boundaries.
contributed significantly to the TP Schemes’ popularity in Gujarat.
Next, the public agency collects the land ownership records from
the state revenue department and other agencies, such as the vil-
(a) Early development of ‘‘priority’’ infrastructure: The develop-
lage local governments. These records also include land parcel de-
ment of roads is a major infrastructure project in any urban
tails, such as the ownership type, encumbrances, the area and
development scheme and contributes significantly to
value of the land and, wherever applicable, the value of the existing
increases in land values. Under the GTPUDA, infrastructure
structures on the land. The public agency must match and recon-
could not be developed until the final state government
cile the land ownership records through a site survey, and get
approval of the TP Scheme, leading to long delays in the pro-
the matched records certified by the state revenue department
vision of infrastructure. Therefore, Gujarat amended the
(Ballaney, 2008). The site survey also notes the location of other
GTPUDA in 1999 to allow local governments to appropriate
man-made and natural features; infrastructure and services, such
land for road construction and other ‘‘priority’’ infrastruc-
as regional water, power and sewer lines; and the location of major
ture within the road right-of-way (such as water and sewer
roads already constructed or to be constructed in accordance with
pipes, and electric poles and transmission wires) soon after
the city master plan. A site base map is prepared based on this
the state government approves a draft TP Scheme, rather
survey.
than after final TP Scheme approval. Approval of the draft
Next, a layout plan and an infrastructure plan are prepared. The
TP Scheme usually occurs within 15 months of scheme initi-
layout plan shows the street network; the new plots to be returned
ation (specifically, after the publication of a ‘‘Declaration of
back to the landowners after taking a portion of the land; the loca-
Intent’’ to prepare a TP Scheme). The approval of the final
tion of various services and amenities, such as parks, schools,
TP Scheme can still take more than a decade.
shops, and an electric substation; and the land parcels reserved
(b) Land ownership disputes do not hold up the process: The TP
for future sale by the public agency. The infrastructure plan shows
Scheme process does not settle land ownership disputes; it
the location of the various infrastructure systems, such as streets
transfers them to the newly reconstituted plot. Therefore,
and the water, sewer, and electricity supply systems. The plan is
such disputes do not hold up the TP Scheme preparation
accompanied by a detailed cost estimate.
process (Interview K. Gajjar).
The next major TP Scheme step entails the determination of: (a)
compensation to be paid to the landowners for the land taken from
Indeed, TP Schemes are the predominant urbanization tool for
them by the public agency; and (b) betterment charges (the better-
all major cities in Gujarat, including its capital Ahmadabad (Inter-
ment charges equal half of the difference between the pre- and
view K. Gajjar). In Ahmadabad, 18 TP Schemes (2300 ha) were pre-
post-TP Scheme land values) payable by the landowner to the pub-
pared in the 20 years before the 1999 amendment. In contrast, 50
lic agency. The scheme costs that are not recouped from the better-
TP Schemes (5028 ha) have been completed since then, and an
ment charges are covered through the sale of the reserved land
additional 47 (4,890 ha) are under preparation. These post-1999
parcels.
TP Schemes serve 1.5 million people, almost the entire increase
At this stage, the draft TP Scheme (comprising the layout plan,
in the city’s population (AUDA, 2012).
infrastructure plan, cost estimates, calculation of compensation
payable to the landowners, and betterment charge calculations)
is presented to the landowners for comments. Next, the scheme Case study LR projects
is modified to address the comments and forwarded to the state
government for approval. Upon receiving the approval of the state This section reviews two TP Schemes—Bodakdev-1B, and TPS-
government, the scheme is called the ‘‘Sanctioned Draft TP 50—located in western Ahmadabad, to illustrate the self-financing
Scheme.’’ Until this stage, the TP Scheme preparation process usu- nature of LR in Gujarat (Fig. 2 provides a map of India showing the
ally takes approximately 14–17 months, or one and one-half years location of Ahmadabad; and the map of Ahmadabad showing the
(see Fig. 1). location of the two TP Schemes). Developed by the Ahmadabad Ur-
Next, a Town Planning Officer (TPO) is appointed. A state town ban Development Authority (AUDA), both schemes urbanized peri-
planning department employee, the TPO, undertakes three rounds urban rural land. Bodakdev-1B is a pre-1999 scheme, where the
of public hearings, in which the individual landowners can contest infrastructure could only be provided after state government ap-
the Sanctioned Draft TP Scheme. The first two rounds of hearings proval of the final TP Scheme. TPS-50 is a post-1999 scheme, where
focus on the physical issues (such as the location, size and geome- the infrastructure development began immediately after state
try of a land parcel). After the scheme is modified by the TPO based government approval of the draft TP Scheme. All infrastructure
upon these two rounds of hearings it is submitted to the state gov- and services have been provided in Bodakdev-1B, while 80% of
ernment as a ‘‘Preliminary TP Scheme.’’ Upon government approval road work is complete in TPS-50. Table 1 provides the scheme
 S. Mathur / Cities 31 (2013) 308–316 311

Fig. 1. TP Scheme preparation process in Gujarat, India. Note: TPS is acronym for Town Planning Scheme. Source: Adapted from IDFC (2009). In turn based upon GTPUDA
(1976) and Ballaney and Patel (2009). Using the Development Plan-Town Planning Scheme Mechanism to Appropriate Land and Building Urban Infrastructure. In India
Infrastructure Report, ed. N. Mohanty, R. Sarkar, and A. Pandey. N. Delhi, India: Oxford University Press.

preparation and infrastructure development timelines. Further- ment charges and the sale of reserved land. As noted earlier, the
more, both schemes serve approximately the same area—209 ha betterment charges equal half of the difference between the pre-
for Bodakdev-1B and 181 ha for TPS-50. Table 2 provides the land and post-TP Scheme land values. Therefore, pre and post land val-
use breakdowns for the two schemes. ues are calculated for each land parcel individually (see Eq. (2)).

B ¼ 0:5  ½ðApre  DÞ  ðLpost  Lpre Þ ð2Þ

TP Scheme costs and revenues
where B equals betterment charges payable by the landowner, Apre
In Gujarat, TP Scheme costs and revenues are estimated at the is the land area before the deduction, D equals land area deducted,
time of the preparation of the draft and final TP Schemes (Inter- Lpre is the land value at the time of the TP Scheme initiation and Lpost
view with K. Gajjar). As Bodakdev-1B is complete, the costs and is the estimated land value once the infrastructure and services are
revenues estimated during the preparation of the final stage of provided as per the TP Scheme.
the scheme are available. These estimates were prepared in As per Eq. (2), a landowner with an initial land holding of 1000
1992. The TPS-50 is still a draft TP Scheme; hence the estimates square meter valued at $5 per square meter and who dedicated
were prepared in 2004, when the draft scheme was sent for state 200 square meter of land, valued at $12 per square meter after
government approval.7 the provision of infrastructure and services, has to pay $2800 in
The scheme costs include the cost to prepare the scheme betterment charges: 0.5  [(1000  200) square meter  ($12/
(including staff salaries, legal expenses and publication costs), square meter  $5/square meter)], or 0.5  800 square meter  $7
the cost of compensating landowners for the value of their contrib- per square meter = $2800].
uted land, and the cost of providing the road, water, sewer and The landowner pays the difference between the betterment
electricity infrastructure. The landowners are compensated for charges and the land deduction compensation. This difference is
land deduction at the pre-TPS Scheme land value rate (see Eq. (1)). called net demand (see Eq. (3)).
C ded ¼ D  Lpre ð1Þ
N ¼ Eq ð2Þ  Eq ð1Þ ð3Þ
where Cded is the compensation to the landowner for land deduc-
where N is the net demand.
tion, D equals land area deducted and Lpre is the land value at the
In the above scenario, the net demand is $1800: $2800  $1000.
time of TP Scheme initiation.
Payable at the time the landowner applies for a change in land use
As per Eq. (1), a landowner with an initial land holding of 1000
or a building permit, the net demand is realized over an extended
square meter valued at $5 per square meter would be compensated
period of time.
$1000 for a 200 square meter land deduction: 200 square
meter  $5 per square meter = $1000.
The scheme revenues are realized from two sources: better- Estimate of revenue from sale of land
All of the reserved land for Bodakdev-1B was sold during the
1988–2006 period. The sale dates and price data provided by AUDA
7
Construction cost for TPS-50 is about 18 times higher than the cost for Bodakdev- have been used in this study. Land sales for TPS-50 have not begun.
1B likely for the following two reasons: a) Quality of infrastructure provided in the TP
Therefore, Ahmadabad Municipal Corporation (AMC) staff was con-
Schemes likely went up significantly between 1992 (when the cost estimates for
Bodakdev-1B were prepared) and 2004 (when the cost estimates for TPS-50 were tacted for their advice on the 2009 land values in the TPS-50. The
prepared). For example, the specifications for sewer pipes and electric poles likely staff advised that prime residential and commercial land can fetch
changed during this period, driving up the scheme costs. b) The construction cost 30,000 Rupees per square meter (Interview with V. Patel). A conser-
increased by more than the increase in the consumer price index for industrial vative land value estimate of 20,000 Rupees per square meter is
workers (CPI-IW)—the index used to calculate the NPV. Although recent data suggests
that construction cost increases mirror CPI-IW, to the extent this was not the case in
used in this article because some of reserved land in TPS-50 may
the decades of 1990s and early 2000s, the NPV analysis might underestimate not be in prime location, and may only be used for other uses that
Bodakdev-1B’s construction costs. have much lower land values, such as civic centers and libraries.
312 S. Mathur / Cities 31 (2013) 308–316

India

Ahmadabad

Ahmadabad

Fig. 2. Location of the two case study TP Schemes. Source: Google Maps.

Table 1
Town planning scheme timeline. Source: Town planning scheme reports from Ahmadabad Urban Development Authority.

Scheme Stages in the TP Scheme preparation process Infrastructure milestones

Scheme Sanctioned Final Roads Other
Notification Draft TP Scheme TP Scheme
Bokakdev-1 B May 1978 March 1982 January 1993 1993–2005
TPS-50 December 2001 June 2004 – 80% Completea –
a
As of December 2010.

Net present value calculation of costs and revenues data were calculated for the end of December 2009 using the All
The construction costs, net demand, and land sales data are India Consumer Price Index for Industrial Workers (CPI-IW). CPI-
spread over a long time period. Therefore, to enable a comparison IW data were consistently available for the entire study period
of scheme costs and revenues, the net present values (NPV) of all (1992–2010), while the more region-specific or finer-grained
 S. Mathur / Cities 31 (2013) 308–316 313

Table 2
Land use breakdown for the case study Town Planning Schemes. Source: TP Scheme reports from Ahmadabad Urban Development Authority.

Land use Bodakdev-1 B TPS-50

Area (sq. m) Area (%) Area (sq. m) Area (%)
Final land parcels to property owners 1,556,866 74 1,157,591 64
Infrastructure and public facilities
Roads 275,963 13 311,685 17
Low-income housing 38,242 2 49,860 3
Othersa 144,053 7 125,740 9
Land reserved for sale 76,456 4 166,559 9
Total TP Scheme area (in ha) 209 181
a
Others include gardens, open space, social infrastructure, civic centers and playgrounds.

Table 3 is primarily due to two reasons: first, rapid increases in land

Price indices.
values in the scheme areas and, second, the large amount of
Index Time period Average annual increase (%) land reserved for sale.
CPI-IW 1992–2009 7.2a
CPI-IW 1988–2009 6.9b An analysis of the sale prices of the residential land sold by
CPI-IW 2004–2009 8.8c AUDA in Bodakdev-1B reveals an annual land price increase of
CPI-IW 2006–2011 7.6
15.2% (see Fig. 3); this is much larger than the 5.9% increase in
Ahmadabad-specific CPI-IW 2006–2011 7.4
CPI-IW 1997–2005 4.4
the CPI-IW during the corresponding period.10
WPI 1997–2005 4.4 Furthermore, AUDA reserves large amount of land for sale. For
example, in Bodakev-1B, the revenues from the sale of land were
CPI-IW stands for Consumer Price Index-Industrial Workers.
WPI stands for whole-sale price index.
expected to equal more than 60 times the minimum required to
a
Discount rate used to calculate NPV of construction cost, compensation for land fully fund the scheme (see Table 4, line 12, column 3). Indeed, I
deduction and betterment charges for Bodakdev-1B. estimate that the actual revenues were approximately 14 times
b
Discount rate used to calculate NPV of revenue from sale of land in Bodakdev- the required amount (see line 12, column 4). Similarly, in TPS-50,
1B.
c revenues from the sale of land are expected to equal more than 8
Discount rate used to calculate NPV of construction cost, compensation for land
deduction and betterment charges for TPS-50. times the amount required to fully fund the scheme (see line 12,
column 7). AUDA staff note that the surplus revenues fund regional
indices, such as the Ahmadabad-specific CPI-IW, the Construction infrastructure, such as a water treatment plant serving several TP
Industry Development Council’s Construction Cost Indices, and Schemes, and infrastructure within new TP Schemes.
the whole sale price index (WPI) were only available for part of
the study period. Moreover, these indices closely tracked the CPI- (b) Net demand funds a smaller proportion of the schemes costs
IW (see Table 3).8 than was originally planned. For example, in Bodakdev-1B,
As per the AUDA staff’s 2010 estimates, almost all of the net de- net demand was projected to fund 95% of the cost of the
mand from Bodakdev-1B has been realized, and the remainder is scheme. The charges ultimately only funded 23% of the cost.
likely to be realized over the next few years. Therefore, for NPV anal- Similarly, it is estimated that net demand will only fund
yses purposes, it is assumed that the net demand is realized over the about one-third (37%) of the cost of TPS-50, compared to
1993–2012 period, a 20-year period starting with the government the planned three-fourths (79%) (see Table 4, line 13). The
approval of the final TP Scheme in the year 1993. For TPS-50 it is as- reduced contribution of net demand is due to the fact that
sumed that the net demand will be realized over the 2004–2023 these charges are only paid by the landowners when they
period, a 20-year period starting with the government approval of seek a change in land use or apply for a building permit.
the draft TP Scheme in the year 2004.9 Furthermore, because 80% Therefore, the charges accrue to AUDA over a long period
of the road infrastructure in TPS-50 was developed by 2009, it is as- (two decades in the case of Boakdev-1B). Furthermore, the
sumed that the entire infrastructure will be developed by 2013. charges are not adjusted for inflation. Hence, in present
Finally, the data from a TPS Scheme prepared by AUDA in 2009, value terms, the amounts of the charges decline over time.
Bopal-3, are used to determine the NPV of staff salaries, publication Indeed, these charges are estimated to be only 2% of the total
costs and legal expenses for Bodakevdev-1B and TPS-50. Table 4 pro- revenues for Bodakdev-1B, and 7% for TPS-50 (see Table 4,
vides results of the NPV analyses for the two case study TP Schemes. line 14, columns 4 and 6).
(c) The cost of providing infrastructure increases over time.
NPV analyses findings AUDA does not budget for such cost increases and relies
heavily on revenues from the sale of land to bear the con-
The analyses of the TP Schemes’ costs and revenues provide struction cost increases.
several insights into the schemes’ financing structures. The find-
ings are as follows:
Sensitivity analyses
(a) Revenues from sale of land are a significant source of reve-
nue for AUDA, far exceeding the agency’s expectations. The Infrastructure development within Bodakdev-1B took 12 years
enhanced contribution of the revenues from the sale of land and is estimated to take 9 years for TPS-50. The amount of time ta-
ken to develop infrastructure impacts the financing of TP Schemes
8
in two ways.
Extensive efforts to gather actual cost and revenue data proved unsuccessful.
Hence, estimates are used.
First, the delay leads to increased construction costs. However,
9
The amendments to the GPTUDA allow infrastructure development and land use
10
changes to begin immediately after the draft TP Scheme approval. The land was sold during the 1995–2006 period.
314 S. Mathur / Cities 31 (2013) 308–316

within the TP Schemes but regional infrastructure as well. It is note-

worthy that the state-level legal and administrative framework
does not identify the funding of regional infrastructure as a permis-
sible use of TP Scheme revenues. However, in the absence of clear
fund accounting systems to track the usage of TP Scheme costs
and revenues, this practice continues, albeit informally.
Given the importance of regional infrastructure to urbanization,
we suggest that empirical studies should be conducted to estimate
the monetary impact of a TP Scheme on regional infrastructure.
Furthermore, the cost of mitigating such regional impacts should
be included as a permissible TP Scheme cost. Accurate and trans-
parent estimation of such costs would, in turn, enable better esti-
mates of the amount of land that must be reserved to fully
recoup TP Scheme costs. Because landowners contribute land, bet-
ter estimates of the need for land reservation would make TP
Schemes more equitable. Finally, as seen in the case of Boadak-
dev-1B, AUDA held on to the reserved land parcels for several years
Fig. 3. Residential land prices in Bodakdev-1B. before selling them. While on the one hand, this practice restricted
the supply of developed land (often parcels reserved for prime
commercial and residential uses), on the other hand, it allowed
to the extent the rate of construction cost increase equals the dis-
AUDA to generate significant revenue. Moreover, the landowners
count rate used for the NPV calculation, construction delays should
who contributed the land did not directly benefit from these sales
not increase the NPV of construction cost. But, the NPV of construc-
(except to the extent that the revenue from the sale of land helped
tion cost would increase with construction delays when the con-
lower their betterment charges). A transparent method of sharing
struction cost rises more rapidly than the NPV discount rate.
the profits from land sales could make the TP Scheme financing
Second, the development of infrastructure increases land val-
process more equitable for the landowners.
ues, leading landowners to either develop their land for urban uses
or to sell it to a new owner who then applies for a land use change
Widely-applicable insights
and/or a building permit. Both cases trigger a net demand pay-
ment. Therefore, delays in providing infrastructure extend the
This article’s analyses of Gujarat’s use of LR to fund urban infra-
net demand accrual period. Furthermore, because these charges
structure and services provide practical insights to other develop-
are not inflation-adjusted, their amounts decline in present value
ing countries searching for urban infrastructure financing
terms. As noted previously, the net demand would accrue over
structures. Three such insights are noteworthy.
20 years in Bodakev-1B.
First, the local governments in Gujarat reap substantial financial
Furthermore, the review of the case study TP Schemes shows
benefits from the sale of reserved land. They retain the reserved
that excessive land is reserved for later sale. Therefore, sensitivity
land for a considerable period of time, allowing them to benefit sig-
analyses were conducted for TPS-50 to estimate the impact of a
nificantly from the increases in land prices. The ability to retain
shorter infrastructure development and net demand accrual peri-
land is primarily a result of local governments not requiring land
od, and to identify the minimum amount of land needed to recoup
sale revenues to bear the up-front scheme costs. Such costs are
project costs. Specifically, it was assumed that the infrastructure
funded through an informal revolving fund system, where the rev-
would be developed in 5 years (2004–2008 period), and the net de-
enues from previously developed TP Schemes help to finance new
mand would be realized over 10 years (2004–2013 period). Table 5
TP Schemes. Such a revolving fund system eliminates the need to
provides the sensitivity analyses results.
seek external financing, such as a loan from a financial institution
As expected, Table 5 shows that shortened construction period
or from a higher level of government (the state or federal govern-
does not reduce the NPV of construction costs. However, the short-
ment) to fund TP Scheme costs, and creates a self-sustaining urban
ened net demand period increases the NPV of the net demand. In-
development financing mechanism.
deed, for TPS-50, the net demand increases by almost 40%.
Second, Gujarat’s LR process is largely equitable for landowners.
Examining the amount of land that must be sold to finance the
The financially self-sustaining LR process obviates the need for
scheme, the analysis shows that AUDA could have reserved just
seeking loans, and in turn, incurring interest costs. Therefore, the
1.1% of the total scheme area for sale, instead of the actual 9.2%.
net demand payable by the landowners is reduced. Furthermore,
With shorter construction and net demand accrual periods, this
the net demand does not need to be paid until the landowner is
percentage could reduce even further to 0.9% of the scheme area
ready to realize the increased value of her land by applying for a
(a 23% decrease).
land use change or a building permit. In fact, it could be argued
In summary, for the two case study TP Schemes, AUDA reserved
that such latitude allows landowners to retain their land for spec-
more land than was required to fund infrastructure within the
ulative purposes. I believe that rather than adopting regulatory ap-
schemes. Furthermore, AUDA can strengthen the self-financing nat-
proaches to address the hold-out problem, such as requiring the
ure of TP Schemes by developing the infrastructure in a timely man-
landowners to sell their land, or to apply for a land use change or
ner (for example in 5 years, instead of the current 10–12 years).
a building permit within a specified period of time, market-based
approaches should be used.11 Such approaches could include the ra-
Study recommendations and concluding remarks pid development of high-quality infrastructure, which would then
incentivize urban uses of land, in turn incentivizing a landowner to
Gujarat-specific recommendations apply for a land use change or a building permit and pay the net de-

The review of the two TP Schemes shows that LR, through TP

Schemes, is effectively used in Gujarat to fund urban infrastructure 11
Market-based approaches are actions or policies that work with, rather than
and services. In fact, the TP Schemes not only finance infrastructure against, private-sector market forces.
 S. Mathur / Cities 31 (2013) 308–316 315

Table 4
Net present value analyses of TP Scheme financing.

Line Bodakdev-IB TPS-50

number
Year 1992 values Year 2009 values Year 2004 values Year 2009 values
(‘000 US Dollars) (‘000 US Dollars) (‘000 US Dollars) (‘000 US Dollars)
Costs
1 Construction 211.1 687.1 8361 12,746
2 Publication cost 0.6 32.3 21.5 32
3 Legal expenses 0.3 32.3 32.3 32
4 Salaries 44.2 1076.9 53.8 1077
5 Total cost 256.1 1828.6 8468.7 13,888
Revenues
6 Net demand 243.4 413.5 6689.6 5137
7 Sale of reserved land 781.3 19573.5 No dataa 71,746
8 Total revenue 1024.7 19,987 6689.6 76,883
9 Surplus (line 8 minus line 5) 768.5 18158.5 1779 62,996
10 Revenue to cost ratio (line 8 divided by line 5) 4.0 10.9 0.8 5.5
11 Revenues from the sale of reserved land required to break even 12.8 1415 1779 8750
(line 5 minus line 6)
12 Revenue from sale of land as proportion of minimum revenue 61.2 13.8 – 8.2
required to break even (line 7 divided by line 11)
13 % of scheme cost funded through net demand (line 7 divided by 95% 23% 79% 37%
line 6)
14 Net demand as % of total revenues (line 7 divided by line 9) 24% 2% – 7%

Note: All figures, unless otherwise stated, are in thousand US Dollars.

It simply assumes that all the scheme costs would be recovered through sale of land.
The US Dollar to Indian Rupee conversion is based upon the December 31, 2009 currecny exchange rate of One US Dollar = 46.43 Indian Rupees.
a
The cost and revenue proforma for TPS-50 does not allocate a value for revenues from the sale of reserved land.

Table 5
Sensitivity analyses for TPS-50.

Line Year 2009 values Year 2009 values (with shortened infrastrcuture % Change from
number (Table 4 values) development and net demand accrual period) Column A to B
Column A Column B Column C
Costs
1 Construction 12,746 12,746 0%
2 Publication cost 32 32
3 Legal expenses 32 32
4 Salaries 1077 11,077
5 Total cost 13,888 13,888 0%
Revenues
6 Net demand 5137 7185 39.9%
7 Sale of reserved land 71,746 71,746
8 Total revenue 76,883 78,931 2.7%
9 Surplus (line 9 minus line 5) 62,996 65,044 3.3%
10 Revenue to cost ratio (line 8 divided by line 5) 5.5 5.7 2.7%
11 Revenues from the sale of reserved land required to break 8750.5 6702.6 23.4%
even (line 5 minus line 6)
12 Revenue from sale of land as proportion of minimum revenue 8.20 10.70 30.6%
required to break even (line 7 divided by line 11)
13 % of Scheme cost funded through net demand (line 6 divided 37.0% 51.7% 39.9%
by line 5)
14 Net demand as % of total revenues (line 6 divided by line 8) 7% 9.1% 36.2%
15 % of Scheme area reserved for sale 9%
16 Minimum amount of land to be sold to fund the entire 1.1% 0.9% 23.4%
scheme (as % of the entire scheme area)

Note: All figures, unless otherwise stated, are in thousand US Dollars.

mand; or inflation-adjust the net demand in order to realize its value agency usually sells reserved land in open market to realize its full
in present value terms. As noted earlier, the net demand is not infla- value; and the landowners are free to sell their land in open
tion-adjusted. Therefore, in present value terms, the cost to the land- market.
owner to retain their land for speculative purposes reduces over Finally, among other reasons, LR’s popularity in Gujarat can be
time. Furthermore, in a developing country such as India where attributed to two key factors: first, the ability of the local govern-
the local governments’ abilities to enforce regulations are weak, ments to develop infrastructure on the road right-of-way well be-
market-based approaches are likely to yield better outcomes. fore the final approval of the scheme. Second, land ownership
LR in Gujarat is aligned closely with market-based approaches. disputes do not hold up the scheme. Such disputes are simply
First, the betterment charges are based upon the market value of transferred to the newly reconstituted land parcels. The legal and
land. Second, the TP Scheme cost includes the full cost of provid- administrative feasibility of adopting these two innovations could
ing infrastructure and services, including the administrative cost be explored by other countries as well.
of preparing and implementing the scheme. Third, the public
316 S. Mathur / Cities 31 (2013) 308–316

References Karki, T. K. (2004). Implementation experiences of land pooling projects in

Kathmandu Valley. Habitat International, 28, 67–88.
Krabben, E., & Needham, B. (2008). Land readjustment for value capturing: A new
Archer, R. W. (1988). Land pooling for resubdivision and new subdivision in
planning tool for urban redevelopment. Town Planning Review, 79(6), 651–672.
Western Australia. American Journal of Economics and Sociology, 47(2), 207–221.
Lall, R. (2009). Foreword. In N. Mohanty, R. Sarkar, & A. Pandey (Eds.), India
Archer, R. W. (1992). Introducing the urban land pooling/readjustment technique
infrastructure report. New Delhi, India: Oxford University.
into Thailand to improve urban development and land supply. Public
Lee, T. (2002). Land readjustment in Korea. Cambridge, MA: Lincoln Institute of Land
Administration and Development, 12, 155–174.
Policy.
Archer, R. W. (1994). Urban land consolidation for metropolitan Jakarta expansion,
Li, L., & Li, X. (2007). Land readjustment: An innovative urban experiment in China.
1990–2010. Habitat International, 18(4), 37–52.
Urban Studies, 44(1), 81–98.
Ahmadabad Urban Development Authority (AUDA) (2012). History of planning.
Needham, B. (2002). Land readjustment in the Netherlands. Cambridge, MA: Lincoln
<http://www.auda.org.in/tp_scheme.html> (Accessed 14.02.12).
Institute of Land Policy.
Ballaney, S. (2008). The town planning mechanism in Gujarat, India. Washington, DC:
Shah, A. (2004). Fiscal decentralization in developing and transition economies:
The World Bank.
Progress, problems, and the promise. World Bank policy research working paper
Ballaney, S., & Patel, B. (2009). Using the development plan-town planning scheme
3282. Washington, DC: The World Bank.
mechanism to appropriate land and building urban infrastructure. In N.
Sorensen, A. (2002). Consensus, persuasion, and opposition: Land readjustment
Mohanty, R. Sarkar, & A. Pandey (Eds.), India infrastructure report. New Delhi,
organizing in Japan. Cambridge, MA: Lincoln Institute of Land Policy.
India: Oxford University Press.
Turk, S. (2007). An analysis on the efficient applicability of the land readjustment
Bird, R. M. (1999). Subnational revenues: Realities and prospects. In S. J. Burki et al.
(LR) method in Turkey. Habitat International, 31, 53–64.
(Eds.), Proceedings from the annual World Bank conference on development in Latin
Turk, S. (2008). An examination for efficient applicability of the land readjustment
America and the Caribbean. Washington, DC: The World Bank.
method at the international context. Journal of Planning Literature, 22(3),
Cete, M. (2010). Turkish land readjustment: Good practice in urban development.
229–242.
Journal of Urban Planning and Development, 136(4), 373–380.
Turk, S., & Altes, W. (2010). How suitable is LR for renewal of inner city areas? An
Gielen, D. M., & Altes, W. K. (2007). Lessons from Valencia: Separating infrastructure
analysis for Turkey. Cities, 27, 326–336.
provision from land ownership. Town Planning Review, 78(1), 61–79.
Turk, S., & Altes, W. (2011). Potential application of land readjustment method in
Gujarat Town Planning and Urban Development (GTPUDA) Act, 1976. An actof
urban renewal: Analysis for Turkey. Journal of Urban Planning and Development,
Government of Gujarat, India.
137(1), 7–19.
Hayashi, K. (2002). Land readjustment as a crucial tool for urban development.
Vera, R. D., & Kim, Y. (2003). Local government finance, private resources, and local
Cambridge, MA: Lincoln Institute of Land Policy.
credit markets in Asia. Economics and research department working paper series
Hein, C. (2010). Shaping Tokyo: Land development and planning practice in the
no. 46. Manila, Philippines: Asian Development Bank.
early modern Japanese metropolis. Journal of Urban History, 36(4), 447–484.
Viitanen, K. (2002). The Finnish urban land readjustment procedure in an international
Home, R. (2007). Land readjustment as a method of development land assembly: A
context: What can be learned commonly. Cambridge, MA: Lincoln Institute of
comparative overview. The Town Planning Review, 78(4), 459–481.
Land Policy.
Infrastructure Development Finance Company (IDFC) (2009). Land pooling and
Yau, Y., & Cheng, C. Y. (2009). Applicability of land readjustment in urban
reconstitution: A self-financing mechanism for urban development. Policy Group
regeneration in Hong Kong. Journal of Urban Regeneration and Renewal, 4(1),
Quarterly (Vol. 7, pp. 1–3). Mumbai, India: Infrastructure Development Finance
19–32.
Company.