Introducing more transparent and efficient land management in post-socialist cities: lessons from Kyrgyzstan/Skaidresnes ir efektyvesnes zemetvarkos ivedimas posovietiniuose miestuose: kirgizijos pamokos.

ABSTRACT. The Urban Institute (UI) worked with five cities in post-Soviet Kyrgyzstan to apply better management practices through the development of Strategic Land Management Plans. Kyrgyzstan transferred property to local governments, but municipal land management had remained poor owing to a proliferation of responsible agencies, lack of rule of law, corruption, and passiveness on the part of local governments. UI worked with local governments to make an inventory of municipal land, publicize the results, and develop a strategy that articulated principles for land management and an implementation plan. This led to several improvements including proper registration of parcels and proactive policies to lease and sell land through open competition. It also established a model for determining public policy that countered corruption and public deliberation of costs and benefits in the use of local assets. Donor involvement to promote good land legislation, the property registration system, and decentralization was also critical to success.

KEYWORDS: Land management; Local government; Municipal property; Decentralization; Transparency

SANTRAUKA

Urbanistikos institutas bendradarbiavo su penkiais posovietines Kirgizijos miestais, kad, pletodamas strategines zemetvarkos planus, ivestu geresne vadybos praktika. Kirgizijoje nuosavybe perduota vietos valdziai, taciau zemetvarkos bukle savivaldybese isliko vargana del atsakingu tarnybu gausos, istatymu trukumo, korupcijos it vietos valdzios pasyvumo. Urbanistikos institutas bendradarbiavo su vietos valdzia, siekdamas inventorizuoti savivaldybiu zeme, paskelbti rezultatus ir sukurti strategija, pabreziancia zemetvarkos principus ir jgyvendinimo plana. Tai leido kai ka patobulinti, iskaitant derama sklypu registravima ir aktyvia zemes nuomos bei pardavimo per atvirus konkursus politika. Be to, sudarytas modelis, nustatantis viesaja politika, kovojancia su korupcija, ir viesus sanaudu ir naudos svarstymus naudojant vietini turta. Prie geru zemes istatymu, nuosavybes registravimo sistemos ir decentralizacijos sekmingo propagavimo daug prisidejo it remejai.

1. INTRODUCTION

Local governments are often considered to be poor managers of land because they lack the incentives of the private sector to maximize utility from land and because politicians and the citizens that hold politicians accountable tend to view land as simply one of the component means to deliver public services rather than a discrete asset to be managed. Yet public land often is a substantial part of local governments' asset bases, and improved management will necessarily have broad budgetary, public policy, and service delivery impacts. This is particularly evident in countries undergoing decentralization: newly empowered local governments can demonstrate greater responsiveness and effectiveness in their management of a key local resource. Improved municipal land management is thus at the intersection between efforts to (i) improve regulations and practice to foster effective property/land management and (ii) decentralize public management to obtain the interrelated benefits of increased participation of citizens in governance and more effective, responsive, and accountable local government.

The authors of this paper were the primary staff of the Urban Institute seeking to apply better land management and decentralization principles in local governments in the former Soviet republic of Kyrgzystan. This work was undertaken as part of a technical assistance project funded by the U.S. Agency for International Development to support more effective and responsive local government. We faced specific circumstances where public land management and particularly decentralized government in practice did not fully match the ideal conditions due to the political realities of how Kyrgyzstan had conducted reforms in these areas. These included in particular cynicism about formal legal process and lack of genuine participation among citizenry and corresponding motivation and responsiveness of local governments. The imperfect, incomplete nature of public land management is more likely the norm than the exception (Rajack, 2007; Garba and Al-Mubaiyedh, 1999; Buckley and Kalarickal, 2006). Therefore the realities of improving municipal land management in Kyrgyzstan are likely relevant to other transitional countries, particularly in the former Soviet Union. Yet despite the challenges, improvements could and did occur, albeit not as rapidly or as fully as might have been the case with more effective public land management systems and decentralization.

This paper summarizes the work performed and assesses the context, lessons learned, and factors of success in the development of Strategic Land Management Plans (SLMPs) in five Kyrgyz cities--Osh, Jalalabat, Karakol, Cholponata, and Karabalta. The SLMPs emerged thanks to cooperation between local officials in each of the 5 cities and the Urban Institute. The paper first reviews the country context for the SLMPs, including an overview of Kyrgyzstan's land management and decentralization reforms and a listing of key factors blocking effective public land management at the local level. Than, the paper outlines a detailed framework for strategic land management and discusses results and implications of applying this framework in five cities. It concludes with reflections on lessons that can be learned by donors and local actors.

2. COUNTRY CONTEXT

2.1. Review of reforms

Kyrgyzstan is a small country in Central Asia, with the population of about 5.2 million of which 1/3 is urban. It became an independent state in 1991 following the break-up of the Soviet Union. During the first 10 years of independence, the country was a leading reformer among post-Soviet countries, including ambitious efforts to denationalize real estate and enterprises and to decentralize government. In particular, there were several important changes in terms of property rights to land:

--The Constitution established that there are distinct state, municipal, private, and other types of property, and all forms are equally protected by law;

--The Constitution and other legislation established decentralized local 'self' government, i.e. locally constituted governments empowered with autonomy from the central state, particularly with regard to municipal property. This meant that local governments enjoyed, at least in principle, independent authority to make and implement land management decisions, which is unusual among post-Soviet countries where central governments retain a variety of controls over local administrations;

--The Land Code (adopted in 1999, revised in 2001) mandated that all land inside city, towns, and villages, which is not state and not private, is in municipal ownership. Moreover, this provision was interpreted by government agencies to be directly applicable, i.e. requiring no additional implementing regulations to transfer land from state to municipal ownership. This comprehensive granting of property rights to local governments made transfer of urban land into municipal ownership substantially easier than, for instance, in Russia where a similar transfer requires site-by-site approval by regional branches of the central government agency.

--The 1999 Land Code also stipulated that allocations of land to private individuals or legal entities must be in the form of either a transfer of full ownership rights or lease right, thus precluding the provision of "indefinite" or "permanent" use rights to private persons and entities. Moreover, legislation introducing the 1999 Land Code called for the automatic conversion of << permanent >> use rights (the main land tenure form in the Soviet Union) to full ownership. The Land Code identified a special category of land--"public use land" (streets, roads, public parks, etc.)--that may not be privatized and may only be leased for up to 5 years. The Land Code stipulated that allocation of vacant land to private parties should be conducted through auctions or tender procedures (except land plots allocated, once-in-a-lifetime, for individual housing);

--The Civil Code established that property rights are not legally valid until they are registered by a special state agency. The State Agency for Registration of Rights to Immoveable Property (Gosregister), was established for registering property rights and transactions involving such rights in 2000. Thanks to a World Bank loan and technical assistance Gosregister was able to establish 50 regional offices and conduct systematic registration of all real estate in cities and towns by the end of 2004.

--Privatization of urban land was conducted through simple channels: for individuals, land plots occupied under various "use rights" and used for single family homes or household kitchen gardens were automatically converted to private ownership; enterprises, which were privatized before the new Land Code was enacted in 2001, received their land sites in private ownership free of charge as well; owners of apartments in multi-apartment buildings are entitled to common shared ownership of land sites appurtenant to the buildings (in practice, the process of identifying the borders of these sites is not completed yet).

To date, more then 6% of the country's entire territory is privately owned, with a much higher share of private land in urban areas (the vast majority of territory are state-owned mountain ranges and pastures). For example, in five cities discussed in this paper private land makes up from 18 to 50 percent of city area, being close to 50 percent in three of these cities. Moreover, the real estate market in urban areas is quite active. For example, in 2006 5.3% percent of all registered residential properties and 3.3% of non-agricultural land sites in our five cities went through sale transactions. This is a very active market, by any standard (Kaganova, 1999).

One of Kyrgyzstan's fundamental steps in launching its program of government decentralization was to transfer substantial portfolios of built-up property (e.g., school and kindergarten buildings, town halls, clinics, and local cultural facilities) and municipal service enterprises (e.g., water/sewer utilities and companies providing solid waste management, street cleaning) into municipal ownership. This transfer happened, similarly to many other transitional countries, quickly (Kaganova, 2006), almost overnight. For example, as the result of this transfer, the cities of Osh and Jalalabat received about 140 properties each. The very fact that there was some process and procedure of the transfer, and the transferred properties were identified, established a clear initial point for local governments to start recognizing their new role of property owners (though this explicit transfer was not prepared well, especially for the enterprises, and has been associated with some long-lasting legal, accounting, and operational issues. However, given the strong inertia of centralized government from the Soviet era, the newness of the concept of municipal ownership, and initial weakness of local governments, property asset management in cities was quite embryonic until technical assistance on this subject was offered by USAID-sponsored projects. With technical assistance, some cities moved energetically with improving some elements of asset management, such as testing competitive procedures of engaging private investors (Kaganova et al., 2001). Also, many cities started registering their municipal ownership on the transferred properties. By late 2003, about 56 percent of the transferred properties were registered, on average, in all 25 cities of Kyrgyzstan.

Unlike built-up properties, taking control over municipal land turned to be a bigger challenge for local governments, and progress on this front has been substantially delayed compared to the built-up properties explicitly transferred to local governments. Such a lag for asserting control over vacant land parcels has been typical for cities in many, if not most, transitional countries. For example, many cities in Croatia, even those that progressed well with improving management of their portfolios of built-up properties, still don't have complete inventories of their land. In Kyrgyzstan and other Newly Independent States much of the confusion lies in the Soviet-era determination that land rights followed, rather than were primary to, the rights to buildings (FIRS, 2005; Strong, 2003). This led to a situation where there were transfers of buildings with land being an afterthought, providing grounds for competing assessments of the borders of land parcels appurtenant to buildings. Furthermore, it meant that land without buildings were generally not established as distinct parcels.

2.2. Challenges to effective land management

The overall situation with municipal land in Kyrgyzstan's cities prior to introduction of Strategic Land Management Plans can be summarized as follows:

--Competing agencies involved with land management. Management of urban land in Kyrgyzstan, as in most post Soviet Union countries, is one of the most non-transparent and confusing areas of public management. There are three key institutional players involved in a process of allocating vacant land and land rights for new construction: a local branch of the State Architecture & Urban Planning Agency ('Architecture Department'), local branch of Gosregister, and local government. Both the Architecture Department and local Gosregister are subordinate to a central government agency rather than to local government. Traditionally, the allocation of vacant land has been the stronghold of the State Architecture & Urban Planning Agency, which controlled the process and held information about vacant land. Gosregister attempted to carve a niche for itself in the process, well beyond its direct responsibility to register rights and transactions, emerging as a second strong player through its control over securing the legal rights to land. Gosregister was created out of older Soviet-era registration and mapping entities but granted additional powers to 'promote the land market' and was the implementing agency for successive World Bank-funded projects. It had resources, a large staff, a broad but vague mandate, as well as a dynamic Director to compete with the Architecture Departments for power over land management. Municipal Property Departments at local governments that, theoretically, should have assumed management of all municipal property including land as the representative of the 'owner', i.e. the municipality itself, have been the weakest institutional link. Moreover, in most cities, particularly larger ones with independent Architecture Departments, these departments have been explicitly removed from dealing with land management in any substantial way. At most the Municipal Property Departments oversaw short-term leases on small plots located on "public-use lands" to street retailers. The key process of allocating vacant land for construction has been controlled by the better staffed and almost always more experienced Architecture Departments and Gosregister. This is despite the fact that allocation decisions must be formally approved by the Mayor or his/ her deputy; these figures in effect wholly relied on agencies not under their direct authority rather than their own property departments. Inertia from the Soviet era where local mayors were directly subordinate to the State contributed to local governments' passiveness.

--Fragmented, unreliable information about land. Information on municipal land has been fragmented, unreliable, and on the whole not available to citizens and local governments themselves. Thus, local governments--which technically by law are owners of all land that is not private or state--have very incomplete information about which sites are municipally owned and where they are located. Some information about vacant land has been in possession of the local Architecture Departments and some information in Gosregister branches, but these agencies typically do not share or join information on a systematic basis. During systematic registration by Gosregister, most parcels, which should have been identified as municipal land, were given only "preliminary" registration in the absence of a specific document detailing the given parcel. Moreover, Gosregister officials routinely commented that local governments did not bother to utilize the systematic registration process to clarify their rights and their specific land parcels. We found that many cities indeed were passive in asserting their rights, despite repeated urgings to be more cognizant of the need, responsibility, and benefit to having a clearly defined portfolio of municipal land. This led to a situation where Gosregister offices arbitrarily determined municipal land in 2000-2002, leading to a very large margin of error both in favor and against cities. For instance, according to surveys of municipal lands undertaken with our project's help in five cities, the difference between actual holdings of municipal land and official Gosregister data is in the range of 7-24% of the city territory (for further discussion see Table 1 in section 3).

--Prevalence of transactions inconsistent with law. Local practices of land allocation to private parties very often occurred with significant violations of land legislation. Typical violations included:

* allocation of land to private parties without competitive procedures,

* allocation of the land rights without terms or without specific types of use,

* privatization of "public use land," especially park land, and

* creation of conflicting claims due to allocation of the same sites to various holders.

Many of these violations were not intentional, especially in early 2000s. They occurred due to several factors, including simple ignorance of legislation at the local level among all of the actors--local governments, state agencies, entrepreneurs, and general public; insufficient professional skills of municipal staff, often as a result of high turnover; and an aversion to directly applying new, reformist legislation without a specific implementing regulation from the central government, a typical problem with the legal culture of many former states. The general ignorance of existing laws at all levels of government is well illustrated by the fact that the central government agencies and the government of Bishkek attempt, time to time, to promote their interests by introducing new laws and regulations that explicitly conflict with existing laws. For instance, legislation was adopted that effectively contravenes the principle of competitive allocation of property rights in order to facilitate a particular deal.

--Corruption, at least in perception. Land allocation and registration, at least in popular perception and in surveys on corruption, are perceived to have high incidence of corruption and abuse of power. In particular, there is the widespread perception that the following occurs: (i) allocation of land sites by corrupt officials for informal payments approximating market prices, while the formal prices are substantially below market or zero, and (ii) land grabs by those in power and their networks of land that is otherwise not available (for example, land plots for individual housing allocated to high-ranked officials on land cut from parks). This perception was confirmed in several cases against municipal officials for improper allocation of land.

--Lack of a proactive government strategy to manage land, particularly allocation. Not surprisingly, with the competing agencies and poor information, local governments are not prepared to engage in proactive management. It takes private sector to be motivated enough to navigate these problems and secure the allocation of public land. Thus, in general, the process of allocating land to new tenants--be this a small spot on a sidewalk for a short-term lease or a lucrative site for new construction--has occurred when potential tenants "find" land sites suitable for improvement and go to authorities with requests to allocate the sites to them. Local governments haven't been supplying sites for lease or sale by their own initiative and through open procedures, except some very limited "show cases" in large cities such as Bishkek and Osh.

3. STRATEGIC LAND MANAGEMENT PLANS

3.1. Objectives

Given this situation, we thought it would be useful to offer cities assistance on improving management of municipally owned land. This work was to be conducted in the context of our project to support effective decentralized governance; making land management a matter of local public policy was an important element in support of this project goal. The main objectives in proposing to develop land management plans were to:

* Help local governments assert management control over municipal land, which should include oversight by local councils;

* Introduce greater transparency in the land allocation process to build public trust as well as the opportunity for more popular control;

* Switch the land allocation process from re-active and non-transparent to more pro-active, with local governments offering land on market within explicit long-term land allocation plans;

* Facilitate transition to full observance of all legislation governing land allocation and management;

* Improve the investment climate for private investment in real estate;

* Establish land allocation as part of public policy, with requirements of government to balance interests of various groups in the community in order to meet development and social equity goals, and

* Increase budget revenues from land sales and leases and establish the practice of earmarking capital revenues from land sales for capital investment.

The fiscal objective of increasing local budget revenues through realizing full market value of municipal land was considered very important because cities in Kyrgyzstan are poor and the taxes assigned to local governments are mostly insignificant and subject to de facto control by the central government. There were various indications that revenue-generating potential of municipal land was underutilized; moreover the increased revenues by law should not be subject to reallocation as was the case with most key local taxes. It should be noted, however, that local officials often did not believe that their governments would benefit from increased revenues due to the range of central government controls over the local budget process prior to fiscal decentralization reforms in 2007. These controls included requirements for centralized confirmation of revenues and expenditures, which in practice included even individual line items, and a centralized treasury system that was selectively responsive in administering local budgets. Local officials expressed fear that increased revenues from land would de facto be offset by cuts in other revenues, as the State would seek to retain as many revenues as possible. Part of our team's task was to use our involvement in order to raise awareness about the detrimental effect of the threat of offsets and advocate for national legislative changes providing local budget autonomy to eliminate the controls that allowed for such offsets.

Land is the most valuable asset that local governments in Kyrgyzstan own--they are "income poor, land rich." Demand for land in many Kyrgyzstan's cities is manifested in relatively high and quickly growing land prices on the private land market. Also, there is high interest to land sites occasionally auctioned in big cities, at least in cases when sale conditions where reasonably formulated. Finally, international experiences pointed to the high revenue potential of municipal land in an increasing number of cities in various countries that were able to leverage their land to achieve remarkable development (Peterson, 2006, 2007).

The new approach to land management, which was proposed in Kyrgyzstan cities, involved a strategic rather than ad hoc approach. This means that municipal land was to be acquired, held, used, and disposed of based on relevant information and explicit policy, in a systematic (planned) and transparent way. An important milestone of implementing such an approach would be adoption by local government of "Strategic Land Management Plan" (SLMP). Furthermore, this adoption should ideally be codified as a local normative act that is therefore legally binding for all parties, above all the local government.

3.2. Framework

The obvious starting point was information. None of the above objectives could be achieved without local governments having explicit, reviewable, current, and reasonably complete information about municipal land. Therefore, creating an inventory of municipal land at Municipal Property Departments was a core task. Even in developed countries, such as the U.S., developing in maintaining a good parcel-based geographic information system (GIS) is still a challenge that requires highly qualified expertise and effort (Hall, 2001), and by late 1990s, the estimated share of jurisdictions in the country working on developing such systems was about 25% (Moundon and Hubner, 2000). At the same time, the encouraging experience from the pre-digital times is that in the U.K., yet back in 1980s, local governments were required to have registers of sites in public ownership that might be available for development (Bramley, 2001). Based on this, from the very beginning, we were oriented toward low-cost and simple methods of inventorying, which would be reproducible in all cities at costs affordable to local governments.

As a pilot project, we offered technical assistance on inventorying land and developing a SLMP to the City of Jalalabat, the 3d largest city in the country. The city was chosen for subjective reasons: the Municipal Property Department has been an active partner since 2000 in its attempts to improve management of municipal property assets; moreover, despite some turnover there was substantial continuity of practice. Soon, several additional cities asked for assistance, including the capital Bishkek. Bishkek, as the center of the country's economic activity, was particularly attractive. Five of the cities succeeded with developing SLMPs, though to varying degrees of detail and implementation. Bishkek, however, did not proceed with an inventory. The Mayor's office resisted conducting an inventory of plots that would be made public, despite strong support for the measure from the City Council, including several formal resolutions ordering the Mayor's office to conduct the inventory of land.

This section lists key activities undertaken, with UI's assistance, by participating cities for introducing strategic land management. It also includes some "know-how" comments based primarily on the complete experience with the pilot city of Jalalabat. Several steps are straightforward; however the key process of field work to conduct an inventory in the Kyrgyz context (step 3) is discussed at greater length.

1) Secure high-level commitment and support at the Mayor's office. Either Mayor himself should be interested in having the SLMP prepared or his First Deputy with sufficient "weight" and dedication.

2) Establish a special inter-agency working group. The group should be lead by either Mayor or his Deputy and include Heads (and staff) of the three key players: the Municipal Property Department, the Architecture Department, and Gosregister. The task of this group is developing the SLMP and addressing all technical and institutional issues that arise during the process, including cooperation among these three agencies.

3) Conduct initial inventory of land that is expected to be municipal and clarify city's ownership rights. Because of the incomplete information, the working group with our assistance had to engage in identifying parcels, in part mimicking the way that entrepreneurs had sought to 'find' land parcels for allocation. With respect to Kyrgyzstan's cities, the types of land to be identified fit into four main land portfolios:

Portfolio 1. Land sites associated with municipal buildings and facilities, including those under control of municipal enterprises and institutions;

Portfolio 2. Vacant, potentially buildable land, not encumbered with third-party rights; Portfolio 3. Public-use land (parks, streets, etc.);

Portfolio 4. Land leased out to third parties or encumbered by other partial rights of third parties.

Portfolio 4 of leased land may overlap with both Portfolio 2 and Portfolio 3, and sorting this out is one of subsequent tasks. However, the first step is to create initial lists of these four portfolios.

Identifying Portfolio 1 of land associated with municipal buildings and facilities is relatively easy because the address lists of these properties exist. Moreover, in Jalalabat, most of these properties already went through registration in Gosregister, which legally defined the size of land sites. However, in other cities, where properties from these portfolios were not registered yet, a problem encountered was that sites' size often varied according to varying sources of information. This lack of registration of the parcel was due primarily to local governments' earlier passiveness during systematic registration.

For Portfolio 2 of vacant land, we recommended establishing initially only basic information in order to reduce costs and delays:

* Site identification number;

* Address (location);

* Size (sq. m).

Later, during clarification of a legal status of the vacant sites and for planning their future, two additional characteristics were added for each site: its legal status and land-use zone according to the official city zoning.

Our initial discussions with the Architecture Department and Gosregister in Jalalabat quickly lead us to a conclusion--which later was confirmed in other four cities--that neither of these agencies had reliable information on vacant sites. Therefore, a lone feasible solution was to conduct a survey of the entire territory of the city, in order to identify such sites. The technique used was very simple and based on walk-through and visual identification of vacant sites on the entire city territory (except districts known as fully built-up). We hired a team of surveyors (generally from among professional surveyors unemployed at that moment and/or staff of the Architecture Department and Gosregister who did this work on weekends or took vacation from their regular work to participate in this survey), assigned to each surveyor a specific territory, gave him a map of this territory in the scale 1:500 provided by Architecture Department (topography maps shown existing buildings and borders of sites, and though the maps were not completely current, they provided some reasonable background) and asked to walk through his assign territory, block by block, and identify, measure, and mark on the map vacant sites and record them on note pads (the lists were then computerized). The minimum plot site subject to recording varied by city. For example, in Jalalabat, the minimum size was about 50 sq.m., while in Osh--about 70 sq.m. During these surveys, surveyors drew sketch plans of each site. When available in other cities that conducted an inventory later, copies of Gosregister's digitized maps in the scale 1:2000 were used instead of the Architecture Department paper maps, and vacant sites were consequently marked on digital maps. Results submitted by surveyors were subject to verification by the working group and random walk-through inspections by a UI project supervisor. The local government of one of the participating cities, Karakol, organized the land inventory independently in cooperation with a respected local NGO.

The next step after the vacant sites were listed and mapped was to clarify their legal status, i.e. whether they had been allocated to somebody and which rights were registered in Gosregister. For doing this, the lists were given to both the Architecture Department and Gosregister. The Architecture Department provided site-by-site information whether it had issued or was in the process of issuing land development permits, and Gostegister provided site-by-site information on which rights were registered, to whom, and based on what right-establishing documents. This exercise demonstrated that a very substantial part of vacant sites had encumbrances, of which some were legitimate and some were questionable. For example, in Jalalabat, among 129 sites found by the survey and believed to be free of encumbrances, 41 sites turned out to have some rights registered on them. After such sites were identified, the Municipal Property Department and Gosregister started a process of clarifying each site's legal situation in order either to exclude the site from the municipal inventory list (if it was legitimately privatized), register the parcel in municipal ownership (in addition to already registered partial private rights, such as lease), or challenge the legitimacy of private rights in court. This clarification revealed that some sites initially identified as vacant (neither state or private) and hence municipal were already privatized. The City decided not to dispute ownership of these sites because it was suitably convinced by documentation that privatization had occurred while its claim was based solely on the general deductive approach set for in the Land Code (what is not State or private is municipal). In the case of other sites, the process of clarifying legitimate rights would take a long time. It should be noted that in some cities the review by Gosregister and/ or the Architecture Department took a large amount of time. The staff of both entities, as they were not subordinate to the city, received no extra payment, and transactions were not imminent, often placed little priority on this work. It would take six or more months to complete the process.

Portfolio 3 consists mainly of streets that usually are not registered as separate "objects" of real estate, and specific "objects" such as parks, standard protected strips along rivers and canals, and parking lots. Identifying land in this portfolio is not difficult, in general.

Identification of Portfolio 4 of leased and otherwise encumbered municipal sites was based on combining two lists: the list of land lease contracts kept by the Municipal Property Department and the lists of all sites with some private rights (such as lease and "use" rights, but not private ownership) registered in Gos-register. However, finalizing this combined list presented some challenges. First, we were led to understand that not all leases, especially short-term leases, are formalized in written contracts. Second, similar to Portfolio 2, many sites turned out to be burdened by rights, but these rights were improperly or illegally granted, requiring legal clarification. For example, in Osh, a large amount of land (about 86 hectares, or about 10% of the estimated amount of municipal land in the city) was registered in "temporary use" (but not lease) of 178 private individuals and legal entities, which was in direct violation of the law. This revelation took Osh Mayor by surprise, and the Work Group was tasked to clarify the situation on a site-by-site basis, with the intent to convert these arrangements into leases.

The task of systematic clarification of land rights related to municipal land and legal "cleaning" of past mistakes was included in SLMP of Jalalabat as an on-going task. The Municipal Property Department works on several cases, but the magnitude of past violations is so high--everywhere, not in Jalalabat alone--that a more radical, nation-wide solution might be needed (see Section 4.3).

4) Consult with the public on identification of municipal land. In Jalalabat, the public was given an opportunity to comment on the initial identification of land owned by the municipality in two ways. First, the lists of vacant sites were displayed on the public information board right in front of the entrance to the Mayor's Office and sent to the Jalalabat Association of Condominium Associations. The purpose was to make sure that the public had an opportunity to inquire about the sites and assert their own claims to parcels. In addition, the novel exercise in transparency increased public trust in local government. Condominium associations were especially targeted because land parcels appurtenant to multi-unit buildings have also either not been defined or have been poorly defined and hence this was perhaps the single most likely source of counterclaims.

5) Plan how to use the stock of vacant land. We urged local governments to be proactive and strategic in making decisions about land management after their portfolio was identified. The idea is that the local government should decide two things regarding each site included in Portfolio 2. first, whether the site can be used for capital construction or for temporary light structures only, and, second, what to do with it: sell, lease, use for public functions, preserve for the future, etc. After discussions with the working group in Jalalabat, we suggested the following groups for classifying vacant land:

* "Golden reserve"--several sites with high market values (current or expected), preserved for sales or borrowing in the future when City would need to generate funding for large infrastructure investment;

* Category 1, land for capital construction. A sub-group Category 1/2006 included sites for auction sales in 2006;

* Category 2, small plots suitable for capital construction only after consolidation with some adjacent sites. Can be offered for sale to owners of adjacent sites;

* Category 3, sites not suitable for capital construction. Can be offered for short-and mid-term lease;

* Category 4--land for future public uses.

Here is an example (see Table 1) of how Jalalabat classified its vacant land in the SLMP.

6) Identify main issues to address in strategic land management and reflect them in SLMP.

In Jalalabat, the SLMP had the following sections:

* Why is the SLMP needed?

* Overview of municipal land;

* Principles defining municipal land policy;

* Management of the vacant land portfolio;

* Management of the public-use land;

* Management of land leases and other partial rights;

* Organizational aspects;

* Financial aspects;

* Information for land management, monitoring and evaluation;

* Reporting.

The text was in large part drafted by the Urban Institute experts who were involved in this pilot project, because staff at the local government did not have either skills for such writing or time. Should be noted, though, that in the USA similar advanced strategic documents are produced by consultants as well. The working group studied the draft, made modifications, and finalized it for presentation.

There were several key issues to be addressed in the SLMP. First of all, as a matter of public policy the local government should articulate principles upon which the plan is based. The Jalalabat SLMP formulated 5 major principles: requirements to allocate most land through competitive procedures with exceptions only as allowed by law; provide full ownership rather than long-term leases for sites sold for capital construction; re-invest sale proceeds in capital projects; maintain transparency; sell sites according to annual sale plans approved by council.

Second, there were various organizational and financial issues and specifics related to particular land portfolios, such as who is in charge for SLMP implementation, how should auction preparations be funded, and what are procedures and responsibilities in converting "public use land" into land for private construction.

The third major area involved the organization of administration of key aspects to ensure the ongoing implementation and amendment as necessary of the SLMP. This included issues of collecting and maintaining information needed for land management, as well as monitoring, evaluation, and reporting.

While most cities followed the Jalalabat example closely, Karakol took the Jalalabat SLMP and revised it independently according to its needs.

7) Plan how expected sale proceeds will be used. Current budgeting practices at local governments do not systematically earmark capital revenues for capital expenditures, though there have been a few examples where cities invested revenues from disposing of some properties into capital construction or repair. In order to guard against using land sales to cover operating deficits, we advised Jalalabat and all other cities to explicitly plan how they were going to spend expected capital revenues from land sales. Thus the SLMP itself defined what would be done with proceeds, specifying which properties would be acquired/ improved, the nature of improvements, and the estimated amount needed. In Jalalabat, this list included 10 items of the total cost of about $68,800 (assuming some cost-sharing with donors on two projects).

8) Discuss the SLMP publicly and approve it at City Council. Upon completion of the draft, it was discussed at a public hearing. After input from the public, the working group then submits it to the City Council for approval. City officials must decide whether the SLMP is an internal document or a local normative act subject to registration at the Ministry of Justice. In the latter case, SLMP becomes binding for the local government. Jalalabat choose the second option.

9) Ensure monitoring and evaluation. Finally, an annual report on SLMP implementation is supposed to be open to the public.

3.3. Current status in five participating cities

Jalalabat formally approved its SLMP in May 2006, though it started its implementation prior to that, from January 2006. Karakol completed and approved SLMP in December 2006. The rest three cities went through inventorying municipal land by mid-late 2006, but have been stalled at two stages: clarifying the legal status of some vacant sites and classifying sites for future use. All of them planned to complete these stages and draft SLMP in April-May 2007.

4. IMPLICATIONS FOR CITIES AND LOCAL GOVERNMENTS

Perhaps the main implication is that the process of creating an inventory municipal land and developing SLMPs brought to light many challenging questions of land policy and land management practices. This first generation of SLMPs won't answer all of these questions, but at least formulating them sets a stage for explicit discussion.

4.1. Quantitative insights

Identifying municipal land parcels

Inventorying municipal land produced several important quantitative results. First, it made clear that the formal state statistics about municipal land ownership--which is based on data officially presented by the Gos-register annually (Form 22)--is not reliable and deviates from the reality substantially. Gosregister tends to provide an overall figure for municipal and state owned land based on historic breakdowns of the total area of cities rather than recalibrating the totals as the sum of individually registered parcels. Thus, Table 2 illustrates that when municipal land holdings were carefully reviewed and measured (see 3.2.3 above), the difference between the estimated area of municipal land and the formal data was in the range from 7.3% of the city area (Kara-Balta) to 23.8% (Cholponata). Deviations can be in either direction: for Osh, Jalalabat, Karakol, and Karabalta the state statistics reported more municipal land than the cities had in the reality; for Cholponata, the state statistics substantially underreported municipal land. And though the inventory estimates are not final since some land might be added or removed from this count after clarification of municipal land rights on some sites, these corrections would not significantly affect the gap. It's important to emphasize that:

* Gosregister officials at local branches in these cities admit that the data on municipal land is obsolete and has not been precise from the very beginning (because the initial systematic registration in Kyrgyzstan never aimed for precise measuring of land sites), and

* Until this process of developing SLMPs started, the owners of municipal land, cities, did not make systematic efforts to properly identify, document, and register municipal land.

Second, the inventory demonstrated that the situation with municipal vacant land varies dramatically from city to city. Thus, Karakol has 20% of the entire city territory in its ownership and vacant! At the same time, Osh owns only 69 ha of vacant land, which constitutes 0.4% of its territory. Obviously local land policy should vary with such differences among cities regarding land holdings.

Finally, there was a further substantial revision to estimates of municipal land based on a review of registered in Gosregister temporary use rights granted by local government resolutions or other decisions. This review was conducted in Jalalabat after the initial field survey and involved mining Gosregister for several months to identify these additional cases. The results were remarkable in terms of how local officials had been providing large amounts of land on an ad hoc basis and the amount of forgone revenue from not signing leases. By checking Gosregister's data with the Municipal Property Department's list of leases, the team in Jalalabat additionally identified a total of 380 additional parcels comprising 197 hectares. Of these parcels, only 34 comprising 8.2 hectares had existing leases. All of the other municipally owned parcels were being used without lease payments, resulting in striking amount of forgone revenues as estimated in the section below.

Competitive allocation and maximization of revenues to city

The Urban Institute's analysis of land allocation practices in Jalalabat and Osh provided quite important insights on land management issues. In connection with this it should be noted that Jalalabat not only was the first city in Kyrgyzstan to go ahead with registration of its built-up properties and development of SLMP, but it also pioneered processing of all (or at least, most) land allocations through competitive bidding as required by law. Jalalabat officials started regularly auctioning land in 2004, and by mid 2006, some data on land prices on municipal auctions has been accumulated.

Figure 1 shows two important trends on the market of municipal land in Jalalabat. First, it indicates that when land is sold competitively, on auctions, the land prices gradient is similar to the gradient in true land markets, be it New York over its history (Atack and Margo, 1998) or post-socialist cities (Bertaud and Renaud, 1997): prices differ by location, falling when sites are offered at less preferred and more remote from the center locations. Second, it shows a dramatic increase in prices from year 2005 to 2006 in similar locations--by 150% in the central part of the city and by about 100% in other locations. This uneven increase implies that the land price gradient became steeper, which indicated that the market matured and quickly, according to Bertaud and Renaud (1997). This increase in auction land prices reflects the overall trend of rapid and continuing crease in real estate market prices that have occurred over the past few years and is likely to continue. Increase in real estate prices has been fueled by a number of factors, chief of which are the fact that real estate was a favored investment for proceeds from remittances and the gray economy and increased demand due to cities' increasing populations.

Further, Figure 2 provides an important insight on a much-debated issue of whether auctioning land helps increase revenues from sales, comparing to sales by "estimated market value". It shows that even if starting prices try to mimic market prices, which is the case in Jalalabat where a special commission establishes the starting prices based on their knowledge of the real estate market without any discount, auctioning increased sale revenues very substantially: the increases between the actual sales price and the starting prices were on average 24% in 2005 and 46% in 2006. Moreover, the very provision of lots to the local market on a competitive basis allows the local government to help contribute to the market's growth.

The a gap between "estimated market prices" and auction prices in Osh, during the period from 2004 through the first six months 2006 was even bigger than in Jalalabat. Thus, Table 3 demonstrates that prices of so called "direct land sales" (i.e. sales to tenants of structures with incomplete rights to the appurtenant land parcel) conducted by "estimated market prices" determined by professional appraisers, were on average 2-3 times lower than the prices obtained on auctions, despite the fact that the sites for these "direct sales" were in better areas than the auctioned sites (as the last column of Table 3 indicates).

[FIGURE 2 OMITTED]

Another important glimpse into land management practices in Osh is provided by data on "use" vs. "lease" allocations (Table 4). This table indicates that, while "small" lease tenants occupying the total of 11.5 hectares have to pay for land use, a much larger area--86 ha--was given to various "big" tenants free of charge, without land leases signed with them. It should be noted that most of these "free" allocations happened during 2002-2006, after enacting the Land Code that outlawed such "use" rights for private tenants.

Not surprisingly, such practices of direct sales and free of charge land allocations lead to forgone income for the city budget. We estimated Osh losses for year 2005 to be at least as follows:

--6% of the total city budget for 2005, or 16% of total own source revenues, including:

--3.6%--<> instead of auctions;

--2.2%--<

--0.2%--recurrent annual loss due to land lease rates defined by "formula" instead of auctions.

Even more striking were the findings in Jalalabat. If one were to apply the lowest existing lease rates in the city (7.8 som or about $0.20 per square meter per month) to the 197 hectares given for free-of-charge temporary uses, then the annual forgone revenues from these properties is about $4.73 mln, which is 150% of the city's annual budget; at the average rate of 12.9 som per square meter per month, the forgone amount is 250% of the budget! By comparison, actual revenues from leases of land made up only 1% of the budget.

The striking difference in forgone rental incomes in Osh and Jalalabat sheds some light on another important issue: land lease rates are set up in Kyrgyzstan cities quite arbitrary, according to some local administrative norms and intra-city zones, and have no connection to market land values. In particular, in Osh, municipal land lease rates are substantially lower that in Jalalabat, despite the fact that the market land prices in Osh are much higher than in Jalalabat.

Nevertheless, being able to present the foregone income is a crucial argument for local governments to engage in proactive land management.

4.2. Institutional and policy achievements

The development of SLMPs and related improvements in land practices in five cities brought up several changes significant for urban land management in these and other cities and towns.

First and foremost, the pilot project in Jalalabat demonstrated that it's feasible for local governments to take municipal land under their control. A study tour to Jalalabat for officials from other cities sparked enthusiasm. Jalalabat's working group demonstrated the entire process--from taking visitors to the "golden reserve" sites and sites for sale to providing samples of documents prepared by Municipal Property Department for registering municipal ownership to hosting them at the public hearing on the SLMP. The City of Karakol made the next important step and demonstrated that strategic land management can be introduced with very limited help from outside: after Karakol's delegation participated in the study tour to Jalalabat, its local government energetically proceeded with development, approval, and implementation of its SLMP, with very limited technical assistance by Urban Institute.

Second, the realization by local governments that municipal land represents public value that should be protected is growing. In particular, the number of local governments asking the Urban Institute for help with land management increases, and they often lament why they weren't educated on this issue earlier.

Third, local branches of Gosregisters started to perform the "complete" registration of vacant municipal land based on the Land Code as "directly applicable law," which is an outstanding achievement for a post-Soviet country. Credit for establishing a precedent in early 2006 goes to the Director of the Jalalabat branch of Gosregister.

Fourth, preparation of SLMPs and related public hearings opened for public debate the issues of managing municipal land. Intuitively, local officials and ordinary citizens understand that municipal land cannot be given to everybody, as land quantity is very limited. In such a case, what could be a fair and pro-poor solution? Is selling public land, especially on auctions, when land goes to those who can pay more then others, fair? In the past, local governments avoided discussing this issue, but silent uncertainty and doubt did exist. We at the Urban Institute suggested, and our advice was well received by local governments and citizens at public hearings--that social justice and concern for the poor would come through governments being effective managers of municipal land maximizing its economic usage, while using the proceeds to provide facilities and services for all citizens. Therefore the governments should unlock the full potential market value of its most valuable land, sell it to those who can purchase it at the highest prices (through transparent competitive procedures), and use the revenues for the benefits of the entire city population and / or the poor. In the Jalalabat SLMP, this concept translated into two policy principles: land is sold and leased through auctions, and revenues from sales and long-term leases are invested in public infrastructure according to the list of projects approved by the City Council.

Finally, the work was particularly effective in countering corruption and perceptions thereof in an area that was notoriously corrupt. Information about what the city owns has been made publicly available in most of the target cities and the others are awaiting full review from Gosregister or the Architecture Department. Jalalabat has an established routine of processing all new land allocations through auctions. Another very encouraging example comes from Osh: during the process of inventorying portfolios of municipal land, the Municipal Property Department substantially improved its leasing practices. First, information management was improved (all leases were sorted out, documents assembled in lease-by-lease folders and also computerized); second, collecting cash was fully excluded from the practice of the Department and, instead, all lease payments became collected through bank payments by lessees. The most encouraging appears to be a concept underpinning these changes and promoted by the Head of the Department. His position boiled down to the idea that transparency was beneficial for the Department and that it was in Department's best interests to establish transparent practices that would protect it from potential suspicions or accusations in corrupt practices.

4.3. Obstacles for further progress

There have been and remain many problems in introducing strategic land management in Kyrgyzstan's cities. Above all, ingrained practices in which some entities and individuals have vested interests are must be broken with. The key specific obstacles have been:

* Lack of follow through. Our close cooperation with local officials tended to generate additional commitment and interest in developing the SLMPs. Once developed, and our presence was not as constant, there was less will to implement the recommendations. This is particularly the case in unraveling prior, probably illegal, allocations of land in order to put them on a proper lease footing. This was true even if the SLMP was a local legal act, which reflects the generally weak rule of law. For instance, in Jalalabat in 2006, the spending of proceeds from land sales deviated from the approved list of capital projects very substantially. A positive side was, however, that the local government at least had a record of how the sale revenues where spent and had the courage to note these deviations from the plan in the annual report on SLMP execution.

* Resistance to transparency. In some cities, even participating in this project, there is resistance to compiling a comprehensive inventory of municipal land that would then be made available to the public. Without a strong political will by Mayors to overcome resistance, inventories will not be done. Further, there are opportunities for agencies involved to make the inventories purposefully incomplete. This option is important to acknowledge and neutralize. A way for protecting municipal land sites from being "hidden" from formal inventorying would be to have independent examiners authorized to access all agencies' information and qualified for conducting checks and verification. At the same time, frequently local officials were reluctant to publicize lists in order to avoid arguments with other claimants or more generally drawing attention to the holdings owned by the city (and what the city already lacks), or perhaps prior actions undertaken that were questionable from a legal standpoint.

* High cost of registration deters local governments from asserting rights. The high costs of registering municipal property or preparing information for registration is typical for many transition countries (e.g., Armenia, Georgia, Serbia). In Jalalabat, long negotiations were held between the local government, Gosregister, and the Architecture Department, to agree what kind of information is required for "complete" registration of municipal land, who will prepare it, and for which price. In general, in most transition countries, the problem is that even if the initial registration of public property is free of charge, the cost of preparing information required for registration, such as site and building plans, is high. This is often aggravated by registration agencies' requirements for excessive information, not needed for confirming property rights.

* Turnover. Kyrgyz local governments are heavily prone to turnover. A change in mayor frequently results in the turnover of almost all key staff. In the 25 cities in which we worked from 2000 to 2006, only 4 retained the same key personnel (and in most cases there were multiple cases of turnover). For municipal asset management (land management is part of it) this is especially devastating because this is a technically complex issue and learning this craft takes time.

* Legal conundrums. A large share of privately owned or leased land sites was allocated with violations of the Land Code and other laws. Clarifying property rights of these landholders is an important task, which often goes beyond local government authority and involves interests of many players. However, in many cases the current local governments are interested in resolving site-related conflicts. Some of our pilot cities are trying to sue land tenants, some land tenants would perhaps sue cities, but this is an inefficient and expensive process. The national government should consider means to clarify these problems on a systematic basis. Neighboring Kazakhstan is currently conducting a 'capital amnesty' whereby individuals can clean up title to property, which is added to tax rolls. There are social equity issues with this approach, but it does allow for achieving a clean slate on property issues rather than having land tied up in legal limbo.

5. CONCLUSIONS

Comparing the results of this pilot project in five Kyrgyzstan cities with existing practices in many transitional and developing economies, we believe this is a success and it can be reproduced elsewhere. However, for transferring this experience to other soils, one needs to look at factors of this success. Our interpretation of key factors is following:

Synergy among several successful technical assistance projects sponsored by international donors. The main three directions of reforms assisted by donors were:

* Enacting good legislation on land and municipal property, in particularly determining that all non-private and non-state land in cities, towns, and villages being municipal under 'directly applicable law," without site-by-site approval by the central government;

* Establishment of a working property registration system; critically important for our story is that the majority of private land holders--even if their rights were only term rights as leases or "use" rights--tended to register their rights in Gosregister; for local governments this implied that Gosregister became a depository of information that the Municipal Property Departments initially might not have, but were able to obtain from Gosregister during the inventorying (in particular, about the portfolio of sites with encumbrances). Although Gosregister has been the target for criticism, it has come to fulfill the role as the unquestioned depositor of property rights, allowing for building an information base to make property decisions in the public and private sector;

* Local government reform, which included substantial emphasis on endowing municipalities with autonomous property rights.

Linkage between municipal land management with overall attempts to improve local governance instead of focusing only on "land reform" technical assistance. We consider this a critical orientation: urban land management is a fundamental matter of local public policy, and good land management should be fostered and supported as a core skill set for local government with large implications for all of local government operations. Even in a small country like Kyrgyzstan the local land situation and practices varies dramatically, and locally specific (and relevant) solutions are needed. There are close synergies between land and property management and other local government functions such as service delivery, the budget process, and the promotion of economic development. Furthermore, the use of land assets should be subject to public scrutiny. This view does not imply, of course, that other components of technical assistance on land reform should not exist: as noted above, this success would not take place if at earlier stages of technical assistance donors would not help the central government to create a framework for land management, which included good land legislation, straightforward land devolution to local governments, and establishment of the land registration system.

A legal environment that rewards local governments for better governance is important. The benefits of better management must be tangible for local governments to break with past practice. Kyrgyzstan had relatively good legislation regarding autonomous municipal property rights. Moreover, changes in the budgeting system are creating the incentive for local governments to increase their local revenues to the benefit of their citizens. Local government skills and capacity have increased over a series of years not only in asset management, but also in citizen information and participation and strengthening local councils. The confluence of all these factors makes local governments more willing to implement reform and to do it well. This also reinforces the point above.

ACKNOWLEDGEMENTS

We are endlessly thankful to two groups of people. First, to our many counterparts in five pilot cities--and not only at local governments but also at local branches of the Gosregister and Architecture Departments--whose interest, dedication to the good of their cities, and hard work contributed in improvements of land management. Second, to our clients at the USAID who supported our work on land management, and in particular, to Ivan Apanasevich and Ted Priftis. We also are thankful to colleagues from the World Bank with whom we have discussed the issues of municipal property registration. We thank Victor Vorobyev for his help with presenting illustrative data. Finally, we are grateful to the anonymous reviewer for very useful comments and suggestions.

Received 28 March 2008; accepted 2 June 2008

REFERENCES

Atack, J. and Margo, R.A. (1998) "Location, location, location!" The price gradient for vacant urban land: New York, 1835 to 1900, Journal of Real Estate Finance and Economics, 16(2), pp. 151-172.

Bertaud, A. and Renaud, B. (1997) Socialist cities without land markets, Journal of Urban Economics, 41(1), pp. 137-151.

Bramley, G. (2001) Monitoring and Managing Urban Growth in the United Kingdom. In: Gerrit J. Knaap (ed.) Land Market Monitoring for Smart Urban Growth, Cambridge, Massachusetts: Lincoln Institute of Land Policy, pp. 335-366.

Buckley, R.M. and Kalarickal, J. (eds.) (2006) Thirty years of World Bank shelter lending: What have we learned? Washington, DC: The World Bank.

FIAS (2005) Land Reform Privatization Procedures and Monitoring System. Final Report on Business Access to Land, Prepared for the Minis try of Economic Development and Trade of the Russian Federation.

Garba, S.B. and Al-Mubaiyedh, S. (1999) An assessment framework for public urban land management intervention, Land Use Policy, 16(4), pp. 269-279.

Hall, C. (2001) Identifying Vacant and Buildable Land. In: Gerrit J. Knaap (ed.) Land Market Monitoring for Smart Urban Growth. Cam bridge, Massachusetts: Lincoln Institute of Land Policy, pp. 53-68.

Kaganova, O. (2006) A Need for Guidance in Countries with Emerging Markets. In: Kaganova, O. and McKellar, J. (eds.), Managing Government Property Assets, Washington, DC: The Urban Institute Press, pp. 255-295.

Kaganova, O., Tian, V. and Undeland, C. (2001) Learning how to be efficient property owners and accountable governments: The case of Kyrgyzstan's cities, Public Administration and Development, 21(4), pp. 333-341.

Kaganova, O. (1999) Russian home-building in transition, Journal of Real Estate Literature, 7(1), pp. 65-76.

Moundon, A.V. and Hubner, M. (2000) Current Land Monitoring Practices and Use of GIS: Challenges and Opportunities. In: Moundon, A.V. and Hubner, M. (eds.) Monitoring Land Supply with Geographic Information Systems, New York, Chichester, Weinheim, Brisbane, Singapore, Toronto: John Wiley & Sons, Inc., pp. 17-40.

Peterson, G. (2007) Land Leasing and Land Sales as an Infrastructure Financing Option. In: Peterson, G. and Annez, P.C. (eds.) Financing Cities, The World Bank, pp. 284-306.

Peterson, G. (2006) Municipal Asset Management: 0nce Sheet Perspective. In: Kaganova, O. and McKellar, J. (eds.) Managing Government Property Assets, Washington, DC: The Urban Institute Press, pp. 145-169.

Rajack, R. (2007) Does the ownership and management of public land matter to land market outcomes? Commissioned Research Paper, 4th In ternational Urban Research Symposium, World Bank, May, 2007.

Strong, A.L. (2003) Tenure Choice for Urban Land Privatization in Ukraine. In: Bourassa, S. and Hong, Yu-H. (eds.) Leasing Public Land: Policy Debates and International Experiences, Cambridge, Massachusetts: Lincoln Institute of Land Policy, pp. 179-204.

Olga KAGANOVA (1) ([mail]), Abdirasul AKMATOV (2) and Charles UNDELAND (3)

(1) Senior Associate, Urban Institute, 2100 M St., NW, Washington, DC 20037, USA E-mail: okaganova@urban.org

(2) Land Management Expert, Urban Institute IBishkek, 195 Tynystanova St, Bishkek, Kyrgyzstan E-mail: aakmatov@ui.kg

(3) Independent consultant, 45 Orozbekov St., Bishkek, Kyrgyzstan E-mail: cundeland@gmail.com Table 1. Initial classification of vacant land in Jalalabat Category Number Area,

of sites sq.m. Golden reserve 15 192,356 Category 1/2006, 11 14,496 Category 1/2007 & 61 240,097 after and Category 4, together Category 2 3 329 Category 3 39 52,224 Category 4 See above See above Total 168 499,511 Table 2. Population and land of five cities

Osh Jalalabat Karakol Population, thousand 241 (1) 83.4 (2) 62.2 (3) Total city area, hectares (6) 16 769 2 499 4 402 By type of ownership, according to formal Gosregister data (6): 5 701 1 182 2 130 Privately owned, hectares & (34%) (47%) (48%) (% of total city area) State owned, hectares 8 661 295 419 Municipal, hectares 2 407 1022 1 854 1Estimated municipal land, 890 732 920 according to the direct survey of land sites, (7) hectares Estimated vacant municipal land, 69 (0.4%) 50 (2%) 880 (20%) according to this direct survey, hectares & (% of total city area) Surveyed territory a, % of total 90% 70% 95% city area

Cholponata Karabalta Population, thousand 8.9 (4) 43.5 (5) Total city area, hectares (6) 3 886 3 201 By type of ownership, according to formal Gosregister data (6): 683 1 611 Privately owned, hectares & (18%) (50%) (% of total city area) State owned, hectares 2 912 781 Municipal, hectares 291 809 Estimated municipal land, 1215 574 according to the direct survey of land sites, (7) hectares Estimated vacant municipal land, 233 (6%) 80 (2.5%) according to this direct survey, hectares & (% of total city area) Surveyed territory a, % of total 80% 70% city area Sources: (1) www.oshcity.kg estimates for 2006; (2) www.citykr.kg estimates for 2004; (3) www.citykr.kg estimates for 2003; (4) http://life.undp.kg estimates for 1999; (5) www.citykr.kg estimates for 2001; (6) Gisregister, Annual land statistics 2006, Form 22; (7) The sum of all identified and surveyed sites, including public use land and some sites with disputed municipal rights. Notes: (a) Specific territories excluded from the surveying were those where no municipal vacant land could be found as the matter of certain knowledge (i.e. either fully built-up or privately owned). Table 3. Results of "direct sales" and auctions, Osh, 2004--six months of 2006

Share of

sites inside

the 1-km zone

Average around city's

Number sale "central point"

of Total price (b), % of

sites area, (a), transactions

sold he $ / sq.m in this group 2004 Direct sales 93 5.00 2.15 32%

Auctions 20 0.77 4.61 10% 2005 Direct sales 77 5.74 1.15 28%

Auctions 23 0.33 3.83 4% 2006 Direct sales 29 0.86 3.0 13% (first Auctions 18 0.48 5.54 11% 6 months) Source: The Municipal Property Department, Osh. Notes: (a) In actual prices, with the exchange rates as follows: 2004: 1 USD = 48.4 som; 2005: 1 USD = 48.0 som; 2006: 1USD = 44.0 som; (b) The old bus terminal as city's "central point". Table 4. Free temporary use and lease of land in Osh

Number of Total Average size Type of rights tenants area, he of site, sq.m. Free-of-charge temporary use 178 86 4800 Lease 561 11.5 224 Source: The Municipal Property Department and Gosregister, Osh.