10
1970 3,811 93 634
1971 3,812 93 634
1972 3,813 93 634
1973 3,814 93 634
1974 3,868 92 629
1975 3,925 90 624
1976 3,970 89 619
1977 2,526 155 880
1978 2,549 153 873
1979 2,566 152 870
1980 2,576 153 860
1981 2,643 149 848
1982 2,674 148 843
1983 2,810 140 820
1984 2,851 138 815
1985 2,919 135 805
1986 2,953 134 779
1987 2,972 133 776
1988 3,023 131 769
In order to measure the degree of land fragmentation in this area, we apply the methods that are often used to study the relation of fragmentation of agricultural land to its productivity. The first method is Januszewski index. The theoretical JI value is between 0 and 1. The smaller the JI, the more fragmented the land parcels are. The figures in Table 3 clearly show that the distribution of land parcels in our study areas is fairly fragmented. There are clearly a large number of small land parcels in this area all the time. The same conclusion is drawn from another similar fragmentation index, the Simpson index (Table 4). The theoretical SI value is between 0 and 1. The larger the SI, the more fragmented the land parcels are. Both JI and SI suggest a high degree of land fragmentation in this area and no sign of improvement over 19 years.
Table 3 Measurement of Land Fragmentation (Januszewski index)
Year JI Year JI
1970 0.02138 1980 0.02670 1971 0.02138 1981 0.02644 1972 0.02138 1982 0.02630 1973 0.02137 1983 0.02591
11
1974 0.02123 1984 0.02580 1975 0.02104 1985 0.02562 1976 0.02092 1986 0.02556 1977 0.02691 1987 0.02556 1978 0.02674 1988 0.02547 1979 0.02669
Table 4 Measurement of Land Fragmentation (Simpson index)
Year SI Year SI
1970 0.98750 1980 0.98732 1971 0.98751 1981 0.98735 1972 0.98751 1982 0.98747 1973 0.98752 1983 0.98747 1974 0.98753 1984 0.98747 1975 0.98752 1985 0.98749 1976 0.98762 1986 0.98816 1977 0.98692 1987 0.98816 1978 0.98690 1988 0.98820 1979 0.98689
We also apply another measure that has been used to study the land distribution among people in ancient China. Gini coefficient measures the (in)equity of income or assets among its owners. The theoretical GI value is between 0 and 1: the larger the value, the more unequally land is distributed among owners. It is interesting to find that in Table 5 over time, the GI has slowly but persistently reduced. In the long term, the size distribution of land parcels has widened. In other words, the land parcels are not equally small; some continue to be fragmented but others (likely a very small number though) have become larger through land assembly.
Table 5 Measurement of Land Fragmentation (Gini index)
Year Gini Year Gini
1970 0.68382 1980 0.70226 1971 0.68372 1981 0.70559 1972 0.68379 1982 0.70683 1973 0.68371 1983 0.71594 1974 0.68345 1984 0.71859 1975 0.68155 1985 0.72273
12
1976 0.68177 1986 0.72722 1977 0.70034 1987 0.73009 1978 0.69846 1988 0.73409 1979 0.69975
The long-term adjustment of parcel size, or the ongoing activities of land segmentation and land assembly is evidenced in Figure 4 through the changes of Simpson and Gini indexes.
Figure 4 The Simpson and Gini Indexes over time
Tragedy of the Anticommons is Real
The statistical measurement of land fragmentation over time suggests on-going activities of land assembly. In order to provide further evidence, we compare the size distribution of land between those of newly developed and others. Figure 5 indicates that the size of newly developed projects tends to be larger than the rest. This again provides evidence of land assembly.
13
Figure 5 Distribution of Sizes for New Development Sites and Others
We carry out a t-test to see if the size of newly developed projects is larger than the rest. The statistical results support our expectation (Table 6).
Table 6 The Size Differences between New Development Sites and Others (t-test)
Year Development States No. of Observations
Average Size
Standard
Deviation t Value P value
1970 new development 72 336 570
3.628 .001
others 3674 90 641
1971 new development 78 329 550
3.790 .000
others 3647 90 644
1972 new development 85 322 532
3.947 .000
others 3620 90 646
1973 new development 99 348 606
4.183 .000
others 3569 90 647
1974 new development 107 326 583
4.139 .000
others 3604 89 644
1975 new development 129 609 3253
1.860 .065
others 3617 76 203
14
We also undertake a non-parametric Mann-Whitney test and the results remain robust.
The size of newly developed projects is larger than the rest (Table 7).
Table 7 The Size Differences between New Development Sites and Others (M-W test)
Year Mann-Whitney
U Statistic Z Statistic Significance Level
15
1970 41126.5 10.029 .000
1971 42498.5 10.613 .000
1972 46229.0 11.041 .000
1973 49924.0 12.195 .000
1974 54658.0 12.65 .000
1975 62950.0 14.114 .000
1976 66224.0 14.779 .000
1977 61769.0 13.541 .000
1978 62933.0 13.968 .000
1979 63963.5 14.216 .000
1980 65762.5 14.724 .000
1981 67671.5 15.422 .000
1982 70658.5 15.943 .000
1983 74540.0 16.855 .000
1984 75302.0 17.211 .000
1985 78751.0 17.776 .000
1986 81766.0 18.566 .000
1987 83014.5 19.212 .000
Finally, we examine the frequency of change of ownership, through land sales, over time.
The clear finding is that several times of land sales are needed prior to a development project. For example, a project of between 500 and 600 m2 on average requires 7 to 8 times of land assembly. And it takes an average of 6 years between each land sales. Given this development pace, this study area will take a total of 85 years to complete development.
16
Figure 6 The Frequency of Ownership Changes in Sites of New Development
Attenuation of Property Rights when Land is at Stake
Our study evidently shows that a significant development project at a larger scale will need to solicit a great number of owners for their agreement. A unanimous consent from numerous owners is fairly unlikely if not impossible at all, but it is needed for virtually every land redevelopment project. Likely because of this kind of dilemma, several pieces of legislations employ the majority rule in development of land in joint-ownership. Below are some of them.
Land Act: Article 34- 1
For the disposal of ownership, or changes of, or setting encumbrance of superficies, Agricultural Right , Servitude of real property, or dien over co-owned land or constructional improvements, the consent of more than half of the Co-owners whose holding of ownership is more than half of the total share shall be required. But if the holding of ownership is more than two thirds, the numbers of consenting co-owners need not be taken into account.
Urban Renewal Act: Article 10
The owners of the lands and legal buildings of an area that has been designated for renewal implementation may designate the renewal units by themselves as per renewal units defined by the authority, or based on the criteria for designating a renewal unit, conduct a public hearing. They may then present a business summary together with the public hearing records to the municipal, county (city) authority to apply for approval.
17
Finally, they can organize a renewing group to implement the urban renewal business of that area or entrust it to an urban renewal business institution for implementation.
The application mentioned in the above paragraph should be agreed by more than 10% of the owners of the private lands and legal private buildings within the renewed area, and the total land areas and the total floor areas of the legal buildings owned should also exceed 10%. If the proportion of agreement has reached the value in article 22, presenting a business summary can be omitted and replaced by presenting an urban renewal business plan directly.
Urban Renewal Act: Article 22
When the implementers is drafting or revising the urban renewal business plans to submit for approval, the applying for approval of urban renewal business in accordance with the regulations in Article 10 should obtain enough agreement as follow. On one hand, in the urban renewal area designated in accordance with article 7, it should be agreed by more than 50% of the owners of private lands and private legal buildings owners within a renewal unit. Furthermore, the sum of their land area and floor area of the legal buildings should be more than 50% of all. On the other hand, it should be agreed by more than 60%
of the owners of the owners of private lands and private legal buildings within a renewal unit. Moreover, the sum of their land area and floor area of the legal buildings should be more than two thirds of all. In addition, the applying for approval of urban renewal business in accordance with the regulations in Article 10 should obtain more than two thirds of the owners of private lands and private legal buildings owners within a renewal unit. Furthermore, the sum of their land area and floor area of the legal buildings should be more than 75% of all. However, if the sum of consenters' private land area and floor area of the legal buildings is more than 80% of all, the computing of owners' agreement can be neglected.
The computation of the proportion between the number of persons and the ownership's of the lands and buildings mentioned in the preceding paragraph can be done to the regulations in Article 12. Competent authorities should verify the proportion of agreement by the end of the exhibition, except the affairs ruled in Article 88, Article 89 and Article 92 in civil law or both of the two ends agree the revoke. If the owners disagree with the urban renewal plan exhibited publicly, they can revoke their agreements by the end of the exhibition except the rights and obligation have not been changed since they assigned the agreement.
Under most situations, the right of ownership to land is protected by property right rule,
18
the owner of a parcel of land has the absolute and upmost right to accept or refuse an offer, or asking for any price he/she demands. However, as demonstrated in the Land Act and Urban Renewal Act, owners’ property right to land is attenuated when development under consideration involve numerous owners. Under this circumstance, the decision of an owner will affect other owners participating in the same development project. And now the rights to land ownership are protected instead by the liability rule that only warrant a fair compensation to the owners who does not agree to a joint project that is believed to benefit the general public.
Conclusions
Land assembly is a common phenomenon in redevelopment of old downtown of a city. The anticommons theory convincingly suggests an inevitable tragedy, but so far has little supporting empirical evidence. This paper sets out to empirically test this theory. Our study suggests that the tragedy of the anticommons has an element of truth, at least in this old part of Taipei City. The challenge faced by the Taipei City is to find a balance between the tragedy of anticommons in land redevelopment and the protection of private property rights.
References
Colwell, Peter F. and Munneke, Henry J. (1997) The Structure of Urban Land Prices, Journal of Urban Economics, 41(3): 321-336.
Fennell, Lee Anne (2004) Common Interest Tragedies. Northwestern University Law School Review. 98 (3): 907-990.
Glaeser, Edward. and Ward, Bryce (2009) The Causes and Consequences of Land Use Regulation: Evidence from Greater Boston. Journal of Urban Economics, 65: 265-278.
Heller, Michael (1998) The Tragedy of the Anticommons: Property in the Transition from Marx to Markets. Harvard Law Review. 111(3): 621-688.
Shoup, Donald C. (2008) Graduated Density Zoning, Journal of Planning Education and Research. 28(2): 161-179