Land prices in Bogota between 1955 and 1978 : a descriptive analysis

Urban andRegional Report No. 80-2 U LAND PRICES IN BOGOTA BETWEEN 1955 and 1978: A DESCRIPTIVE ANALYSIS * Rodrigo Villamizar A. Corporacion Centro Regional de Poblacion Bogota, Colombia April 1980 This report was prepared under the auspices of the City Study Research Project (RPO 671-47) as City Study Project Paper No. 10. The views reported here are those of the author, and they should not be interpreted as reflecting the views of the World Bank or its affiliated organizations. This report is being circulated to stimulate discussion and comment. Urban and Regional Economics Division Development Economics Department Development Policy Staff The World Bank Washington, D. C. ABSTRACT LAND PRICES IN BOGOTA: 1955-1978 Analyses of land prices in Bogota have been carried out over the period 1955 to 1978 using 6,100 observations of land price estimates for vacant parcels drawn from the archives of Wiesner y Cia., a Bogota real estate firm. These analyses show that overall average land prices have increased regularly over time at a low rate, while land py7ices in particular parts of the city have varied much more. For example real land prices in peripheral locations have increased at high rates while real land prices in central areas have declined in absolute terms. In addition, real land prices in radial sectors located in the south-west and north-west have increased more than in other radial sectors. Over the twenty-four year period land prices in the central business district (CBD) prices have been the highest levels observed in the city, although CBD prices have fallen over time in absolute terms. Overall, the price "surface" has kept its peak at the center but the pattern of relative price changes has tended to "flatten" this surface by lowering the peak and raising the edges. These trends are confirmed by the estimation of land-price gradients and a price index for the city and its sectors. The ratios of land prices in different rings to that in the CBD as one moves out from the CBD for the last three-year period were as follows: .50, .43, .37, .27. and .24 of that of the CBD price. Similar ratios estimated for the 1955-1959 period are as follows: .37, .22, .14, .04 and .03 and show the CBD price level to be greater than the prices of contiguous rings in the earlier period. Finally the estimated total "value" in real terms of the land of Bogota has been growing at an annual rate of about 4 percent to reach a value of Col$362,000 million pesos in 1978 (about US$9 million). PREFACE I would like to thank Rakesh Mohan for initiating and guiding the course of this research. I have benefited significantly from comments on earlier versions by members of the City Study Research Project of the World Bank and the Corporacion Centro Regional de Poblacion, especially from Gregory K. Ingram,and Alvaro Pachon. I am also indebted to contributions made by Bernardo Kugler, Ramiro Cardona and Alvaro Lopez. A very special effort was made by research assistants Horacio Osorio and Manual Salgado and secretarial assistance both at CCRP ind the World Bank's Urban Development Staffs. The study would have been impossible without the cooperation of Mr. Guillermo Wiesner of Wiesner and Cia., Bogota, who kindly provided the data set on which this study is based. We are very grateful for his considerable help. This paper is part of a program of research currently being conducted by the World Bank on Bogota and Cali, Colombia. The goal of the program, entitled The City Study, is to increase our understanding of the workings of five major urban sectors - housing, transport, employment. location, labor markets and the public sector - in order that the impact of policies and projects can be assessed more accurately. This paper is one of a series that focuses on the urban land market. A background paper on land markets in Bogota is also available: Guillermo Wiesner, "Cien Afos de Historia Sobre los Precios de la Tierra en Bogota: 1878-1978", CCRP City Study Working Paper No. 3. TABLE OF CONTENTS Page No. I. INTRODUCTION ............................................... 1 II. THE EVOLUTION OF LAND PRICES IN BOGOTA: 1955-1978 ......... 5 A. Evolution of Prices for the City as a Whole ........... 5 B. Evolution of Real Prices by Rings ..................... 5 C. Evolution of Real Prices by Radial Sectors ............ 9 D. A Land Prices Index ................................... 14 III. THE BEHAVIOR OF LAND PRICES IN BOGOTA: 1955-1978 .......... 27 A. The Estimation of Price Gradients for the City as a Whole ......................................... 27 B. The Estimation of Price Gradients for Individual Radial Sectors ................................... 34 C. Comparison of Land Price Gradients with Other Cities 36 IV. CORRIDOR ANALYSIS .......................................... 39 V. CONCLUSIONS ................................................ 45 FOOTNOTES ................................................. 48 APPENDICES ................................................. 50 Appendix 1 - Maps 1 - 4 and Tables 1-1 to 1-9 ........ 52 Appendix 2 - Averaging Land Prices .................... 66 REFERENCES ................................................. 69 LIST OF TABLES Table No. Page No. 1 Four-Year Real Average Land Price in Bogota, 1955-1978 ........................................ 6 2 Evolution of Average Real Land Prices in Bogota by Ring and Three-Year Periods ................... 7 3 Evolution of Average Real Land Prices in Bogota by Radial Sector and Three-Year Periods .......... 11 4 Estimated Land Price Index (LPI) and Index Ratio (IR) for Each Ring and Three-year Period. 1955-57 - 1976-78 ................................ 15 5 Estimated Land Price Index (LPI) and Index Ratio (IR) for Each Radial Sector and Three-Year Periods: 1955-57 - 1976-78 ................................ 17 6 Evolution of Land Values in Bogota (LPI), as Compared with the General Price Level 1955-1978... 21 7 The "Value" of Bogota's Land 1955 - 1978.......... 23 8 Comparison Between the Total Land Value of Bogota and Both Gross Domestic and Regional Product for Selected Years 1960-1978 ......................... 24 9 Comparison Between Estimated Total Land Value of Bogota, D.E. and Cadastral Assessments 1974-78 ... 26 10 Estimation of Price Gradients and Elasticities from Negative Exponential Price Functions of Bogota: 1955-1978 ................................ 30 11 Estimation of Price Gradients from Negative Exponential Functions for Different Radial Sectors of Bogota: 1955-1978 ............................. 35 12 Chicago Land Value Gradients: 1836-1928 .......... 37 13 Land Price Gradients of Selected Japanese Cities .. 38 14 Four-Year Real Average Land Prices Along the Main "Corridors" of Bogota, 1955-1978 ............ 42 LIST OF FIGURES Figure No. Page No. 1 Land Prices in Bogota ................... 8 2 Land Prices in Bogota ......................... 12 3 Land Price Indices in Bogota .................. 16 4 Land Price Indices in Bogota .................. 18 5 Land Prices in Bogota ......................... 28 6 Gradients and Elasticities .................... 32 7 P (0) VS B .................................... 33 8 Evolution of Land Prices Along Main "Radial" Corridors ..................................... 43 9 Evolution of Land Prices Along Main "Ring" Corridor Sections.............................. 44 LIST OF APPENDIX MAPS Number Title Page No. 1 Bogota, D.E.............................. 52 2 Bogota, D.E.,Rings in Bogota ............ 53 3 Bogota, D.E. Radial Sectors in Bogota.... 54 4 "Radial" and "Ring" Corridors, Bogota .... 55 LIST OF APPENDIX TABLES Table No. Page No. 1-1 Frequency Distribution of Land Price Data Set by Comuna in Bogota: 1955-1978 ................... 56 1-2 Definition of "Ring" and "Radial" Sectors ........... 57 1-3 Distribution of Number of Observations by Radial Sectors and by Three-Year Periods: 1955-1978 ........ 58 1-4 Distribution of Number of Observations by Ring and by Three-Year Periods: 1955-1978 .................. 59 1-5 Monthly Series of Consumer Price Index For Bogota ... 60 1-6 Annual Consumer Price Index (White-Collar Workers) for Bogota ......................................... 61 1-7 Moving, Weighted Real Land Prices in Bogota by "Comuna" and Selected Periods ....................... 62 1-8 Average Real Land Prices of "Ring Corridors" in Bogota for Four-Year Periods Between 1955-1978 ..... 64 1-9 Average Real Land Prices of Radial Corridors in Bogota for Four Year Periods Between 1955-1978 ...... 65 I. INTRODUCTION The main objective of this document is to analyze the behavior of the land market in Bogota, Colombia, through the examination of land prices over time. A unique data set is employed to describe the structure .and evolution of Bogota's land market. Furthermore, we attempt to under- stand the dynamics of the changing "land price surface" of the city using estimates of the structural parameters of land-price gradient functions. Analysis of this type is very important, for it is known that the rapid pace of urbanization,. in the last quarter of the century has drastically changed the structure of cities in developing countries. Classic as well as contemporary analyses of land markets have tended to argue that land prices in cities of developing countries exhibit strong irregularities and disequilibrium behavior. The belief that stable equilibrium cannot be reached in urban land mar!-ets seems to be based on the assumption that the land market reflects, very directly, the strong effects of high population growth rates, rising incomes, and an inelastic supply of urban land. The findings of this paper suggest that land price patterns in Bogota exhibit regularities over space and time, and are not chaotic. The standard procedure to document shifts in the structure of cities has been through the use of population density gradients, which show an almost universal pattern of population decentralization. 1/ Although we have very little empirical knowledge about land-price gradients, theory predicts a similar pattern. As Mills (1972) points out "theory leads us to believe that land prices decline exponentially from the center of the city." This hypothesis has not been extensively tested, however, because of lack of readily available data. 2/ 2 We have obtained a rich source of information on land prices over a period of twenty-five years, that enables us to describe the changing patterns of the land market and its more important determinants. This land price data bank is comprised of transaction and appraisal values of vacant lots realized by the "Wiesner and Cia. Ltda.", a real estate firm that started operation in Bogota in the late 1920s. The data set analyzed here consists of 6,100 observations selected from Wiesner and Cia. files on the basis of lot size, infrastructure characteristics, access to public transit systems, past transactions and transactions of nearby lots. The period covered by these observations is January 1, 1955 to December 31, 1978. Following the same spatial framework used in other areas of the City Study, the City of Bogota was divided into thirty-eight 1/ "comunas" (see Map 1) which are compact "barrio" groupings developed by DANE (Departamento Administrativo Nacional de Estadistica). The "comunas" have been, in turn, aggregated in two ways. The first pattern of aggregation, named Riag sectors, groups the comunas into semi-circular rings around the city center (Map 2). The second pattern of aggregation, named Radial sectors, divides the city into radial corridors or "pie- slices" (see Map 3). These two aggregations of comunas are employed to describe the gross patterns of land prices in Bogota. Moreover, these comuna aggregates are identical to those used in parallel analyses being conducted in the City Study in the areas of employment location, transportation, and residential location. In addition to being shown graphically in Maps 2 and 3 in Appendix 1, these aggregates are defined in Table 1-1 of Appendix 1. 1/ All maps are placed at the end of the paper, Appendix 1. -3- We use Arabic numerals for rings and Roman numerals for radial sectors. The distribution of the observations by comuna, ring sector, and radial sector for three-year periods is shown in Tables 1-2, 1-3 and 1-4 of Appendix 1. 3/ A more detailed description of this data set can be found in a paper by Mr. Guillermo Wiesner, along with an overall description of land price behavior during the last one hundred years.- The rest of the paper is divided into five sections. Section II describes the evolution of land prices averaged over "comunas" in both nominal and real terms. To illustrate the former'a land-price index (LPI) is estimated and compared to a general consumer price index (CPI). As for the latter, the evolution of real prices is described along with estimates of their annual rates of growth in different ring and radial sectors. The third section consists of a description of the behavior of land-price gradients. Such gradients are the result of the estimation of exponential land-price functions. The estimation of the functions is undertaken for the city as a whole on a yearly basis and for different radial sectors of the city as well. Section IV deals with what is called "corridor analysis" or a close-up look at the land-price surface behavior along main streets of the city. An important conclusion derived from section IV is that although its results seem to differ somewhat from those obtained in section II, they are in fact complementary rather than contradictory. As will be illustrated below, the finding in Section IV of non-central locations where the price-surface reaches its highest values is the result of different levels of aggregation. While in sections II-III we deal strictly with comuna averages, in sertion IV we deal with averages along specific corridors or street networks.- Price -4- "peaks" and "valleys" found along corridors are suppressed when comuna averages are considered. Section V consists of general concluding remarks and observations. Finally, the appendices contain maps, data description, and details of methodology followed in some intermediate procedures. 5 II. THE EVOLUTION OF LAND PRICES IN BOGOTA: 1955-1978 A. Evolution of Prices for the City as a Whole Contrary to much of what has been said and written about spectacular increases of land prices in Bogota, average land prices in Bogota have increased over time at a relatively modest rate. As one would expect, prices of land have increased dramatically in certain areas, while in others they have decreased in both real as well as absolute terms. The overall modest real growth is illustrated in Table 1 which shows relative changes in both real and nominal average unit prices of land in Bogota. Real prices have been increasing since 1955 at a rate o0r about 4 percent per year. In nominal terms, the price of a square meter of land has on average increased from Col$36.00 in 1955 to Col$784.00 in 1978. This estimate, as well as others which will follow in the sections below, has been calculated keeping the area of Bogota constant.6/ City-wide averages are, however, somewhat misleading. There are sectors of the city in which prices have increased at a relatively high rate of growth while others such as the Central Business District or CBD have even declined in absolute terms. Thus, the evolution of average prices for different sectors of the city is the subject of the next two sections. B. Evolution of Real Prices by Rings The evolution of moving-weighted-average real (see Appendix 2) prices by Ring during the twenty-four year period analyzed (1955-1978) is summarized in Table 2 and Figure 1. As one can see in Table 2 and - 6 - Table 1: FOUR-YEAR REAL* AVERAGE LAND PRICE IN BOGOTA, 1955-1978 Average Price* Number of Period (Col$/M2) Observations (1) (2) 1956-1958 36.0 36.0 924 1959-1962 43.4 67.2 1334 1963-1966 48.6 123.0 1057 1967-1970 51.9 193.0 910 1971-1974 63.5 335.3 909 1975-1977 70.4 783.6 970 Avg. Total 1955-1978 52.3 256.4 6100 Notes: * Average Prices are moving-weighted-average prices for "comunas" calculated according to the procedure explained in Appendix 2. (1) Real Prices (1955 = 100) (2) Nominal Prices l/ 2/ Table 2: EVOLUTION OF AVERAGE- REAL- LAND PRICES IN BOGOTA BY RING AND THREE-YEAR PERIODS (All Figures in Col$ - 1955) Annual Three-Year Periods Rate of Ring 1956-59 1960-62 1963-65 1966-68 1969-71 1972-74 1975-77 Change 3/ 1 324 276 255 223 245 232 185 -2.5 (1.00)* (1.00) (1.00) (1.00) (1.00) (1.00) (1.00) 2 120 105 111 120 108 100 93 -1.1 (.37) (.38) (.43) (.53) (.44) (.43) (.50) 3 73 82 80 80 78 81 79 0.3 (.22) (.30) (.31) (.36) (.32) (.35) (.43) 4. 45 55 52 55 56 65 68 2.0 (.14) (.20) (.20) (.25) (.20) (.28) (.37) 5 15 27 34 33 43 48 51 6.6 (.04) (.10) (.13) (.15) (.17) (.21) (.27) 6 9 16 22 23 27 42 44 9.5 (.03) (.06) (.08) (.10) (.11) (.18) (.24) Notes: *Numbers in parenthesis represent the ratio of prices in each ring to prices at the CBD (ratio = Pring/PCBD). Ring 1 coincides with the CBD proper. 1/ Moving weighted average prices (see Appendix 2). 2/ Prices of 1955 (1955 = 100) / Annual rate of growth (%) estimated by fitting a log function through annual observations of each ring. FIGURE 1 o-R I NG "1'"., m-R I N "2" -RING "3" +-RING 524" -RI NG "6 z-RING "6" LRND PRICES IN B0GOTR 1956 1960 1964. 1966. 1972. 1976 3 2 b.-0 325.0 300 .0 300.0 75.0 275..0 c 2b0.0 250.0 LW 225.0 225.0 (9) 200.0 200.0 175-0 175.0 150 . 0 -150.0 150.0 25 .0 125.0 100.0100.0 .0 5.0 1956. 1960 1964 1966 1972 1976. YERRS 9 Figure 1, the average land prices of rings 1 (CBD) and 2 (located from 1 and 3 kilometers from the CBD) have been decreasing over time. Ring 3 (3-6 kilometers from CBD) shows hardly any change over the period; while Rings 4, 5 and 6 (located at an average of 6.9 to 12 kilometers from the center) show a systematic pattern of price growth as one moves toward the periphery of the city. Price increases have been more significant in ring 5 than in ring 4 and even more so in ring 6 than in any of the other rings. From its 1955-59 level, ring 6 went up about five times (at almost 10 percent per year) by thi end of the period while rings 4 and 5 only increased 1.5 and 3.5 times respectively (growing at 2 and 7 percent annually). In short,-although the overall average real land price has grown at a modest rate, relative decreases around the center along with increases at the periphery result in an overall "flattening" of the price vs. distance curve through time. This last point is illustrated in Table 2 in the figures in parentheses. The numbers in parenthes under- neath each three-year average value show the price of each ring as a pro- portion of that of the CBD. The value of 1.0 is assigned to Ring 1. These numbers illustrate that the ratio of Ring to CBD prices has in- creased overtime. The last column of the Table shows that as one moves out from the city center of CBD, the price level increases at an increas- ing rate. C. Evolution of Real Prices by Radial Sectors Eight sectors form the radial-shaped or "pie-slices" sectors of Bogota. Since each sector contains a portion of each ring, the results are not controlled for distance from the center; instead we have a sectorization that can explain price variations in terms of other factors like density, income, transportation facilities, and neighborhood quality. - 10- Both Tables 3 and Figure 2 show that the areas of the city which have faced the most significant price increases are those located in the south-west (Sector III), in the west (Sector VI) and the south Sector II). These sectors increased their 1955-59 level by 310, 270 and 230 percent by the end of the period, growing at annual rates of 6.8, 4.7 and 4.5 percent respectively. The areas covered by these three sectors, although somewhat heterogeneous, are generally characterized by low-socio-economic strata. Over 80 percent of the barrios classified as low and low-low income stratum are located within these three sectors. This evidence contradicts the general impression that it has been the north (Sector VIII) where "the action is".7/ This sector show only a slightly better "performance" than those sectors located in the industrial corridor (Sectors IV and V), whose prices have remained very stable through the years. This land-price variation across radial sectors is not an isolated phonemenon; rather it is a result of socio-economic characteristics that are spatially concentrated more in terms of radial sectors than in terms of rings. In general, population, population density and income differentials are closely associated with these land prices variations. For instance, densely populated areas such as the Kennedy area ("Ciudad Kennedy", "Kennedy Oriental", "Kennedy Sur" and "Kennedy Central") and the Provivienda area (Provivienda Occidental", Provivienda Sur", and "Provivienda Oriental"), in sector III, have some of the fastest growing land prices of the city. In a similar way, sector II land prices are considerably affected by increasing population densities, although income levels, for the sector as a whole, are the lowest in the city. Sector II 1/ 2/ Table 3- EVOLUTION OF AVERAGE:- REAL- LAND PRICES IN BOGOTA BY RADIAL SECTOR AND THREE-YEAR PERIODS (All Figures in Col$ of 1955) Annual Radial Three-Year Periods Rate of Sector 1955-59 1960-62 1963-65 1966-68 1969-71 1972-74 1975-77 Change 3V I 324 276 255 223 245 232 185 -2.5 (1.00)* (1.00) (1.00) (1.00) (1.00) (1.00) (1.00) II 21 33 35 37 45 48 49 4.5 (.07) (.12) (14) (.17) (.18) (.21) (.27) III 14 22 30 41 40 45. 44 6.8 (.04) (.08) (.12) (.19) (.16) (.19) (.24) IV 51 61 69 72 63 67 61 0.8 (.16) (.22) (.27) (.32) (.26) (.29) (.33) V 48 47 44 50 51 53 63 1.1 (.15) (.17) (.17) (.22) (.21) (.23) (.34) VI 19 36 48 30 46 53 52 4.7 (.06) (.13) (.19) (.14) (.19) (.23) (.28) VII 30 41 46 46 47 66 63 4.0 (.09) (.15) (.18) (.21) (.19) (.28) .(.38) VIII 46 54 49 50 61 67 72 2.5 (.14) (.19) (.19) (.23) (.25) (.29) (.43) Notes: *Numbers in parenthesis represent the ratio of prices in each radial sector to prices at the CBD (ratio = Pring/PCBDJ I/ Moving weighted average prices (see Appendix 1). 2/ Prices of 1955 (1955 = 100) 3/ Annual rate of growth (%) estimated by fitting a log function through annual observations of each radial sector. - 12 - FIG.2 LRND PRICES IN -9,OTR 1956 1960 1964 1963 1976 325.0 -- 325 .0 300.0 -~ 300.0 275.0 275.0 250.0 250.0 225.0 - -225.0 200.0 200.0 x-RRD - SECTOR " 2 150.0 150.0 - 4-RRO.- SECTOR "2 150.0 -+- RRU. SECTOR "13" 2. 125.0RD SECTO "" 125.0 100.0 - 75.0 75 25.0 25.0 1956 1960 1964 1962 972 1976 YERR\S 195l 1960 1962 1964 1966 1968 197: 172 1974 1976 325.0 325.0 300.0 300.0 275.0 275.0 + -RRDO. SECTOR "5" 250.0 - SECTOR "6- 250.0 w ~'-RRD. SECTOR "'6" - 225.0 -- 225.0 z-RRD. SECTOR "7"K 200.0 Y-RRD. SECTOR "8 200.0 0 175.0 175.0 150.0 -E - 150.0 c 125.0 125.0 Lij 100,0 100.0 75.0 75.0 25.0 25.0 1958 1960 1962 1964 1966 1963 1 1972 1974 1976 YERRS - 13 - includes 26 out of 37 barrios ranked by DANE, (Departameito Administrative Nacional de Estadistica) in 1976, as slum areas.- The other sector with rapidly rising land prices, Sector VI, is very heterogeneous. It com- bines areas of high population growth with others of low population growth, and areas of medium and medium-high income strata together with areas characterized by low-medium and low-low socio-economic stratum. Other sectors with relatively low population density and/or high household in- come levels show relatively lower price increases than those mentioned above. Among such sectors are Sector VIII, the northern part of the city between "carrera" 7 and "Av. Caracas" and Sectors IV and V (the city's industrial area). It appears that other factors than distance from the center are important determinants of land price differentials and behavior. Among those factors are population density, income, neighbor- hood quality, infrastructure conditions, and amenities. An analysis of these determinants is, in fact, the subject matter of the next paper in this study.- This paper focuses on the patterns of prices which are descriqd in the following section in terms of both time and distance variables which are implicit in the combination of sectors and periods selected. Based on the ring and radial sector averages, we can highlight two quite distinctive patterns of price behavior: 1. The CBD has registered the highest land price level 10/ since 19551- although its average price has consistently decreased in absolute terms (the CBD is defined as Ring 1 and Radial Sector I or simply "comuna" 31).A1/ -14- 2. The periphery of the city (Rings 5 and 6) has registered the largest price increases of land during the period 1955-1978. More specifically, the areas of these two outer rings which coincide with Radial Sectors II, III, VI and VII have captured the largest relative price in- creases in all the city. D. A Land Price Index Our conclusions about the parts of the city which experienced the largest price increases are somewhat different from popular belief and from the opinions of some real estate people interviewed. Most people not only believe that there have been "dramatic" price increases in the last 4 to 6 years but many are convinced that the city's north (roughly our Radial Sector VIII) has been the focus of most price increases. This has led us to hypothesize that perhaps people's perceptions about land price increases are geared to nominal rather than real prices. This hypothesis is related to the question debated among economists over whether wage earners negotiate in terms of nominal or real wages. In terms of land rents the question may be similar only that it focuses on the more recent developments of "rational expectations" theories: Do people suffer from "money illusion" in land-related markets? Are nominal rather than real terms the basis for land speculation? These and other related questions which in principle may be answered with more rigorous empirical work, were the reason for con- structing a preliminary price index for land on the basis of "comuna" price averages. The land price index was constructed using the land area of each ring and radial sector as weights. Table 4: ESTIMATED LAND PRICE INDEX (LPI) AND INDEX RATIO (IR) FOR EACH RING AND THREE-YEAR PERIOD: 1955-57 - 1976-78 RING 1955-57 1958-60 1961-63 1964-66 1967-69 1970-72 1973-75 1976-78 LPI 1/ 1 100.0 101.1 135.2 187.7 283.1 238.7 363.6 259.3 2 100.0 146.4 158.3 .267.1 379.3 372.6 567.9 1225.5 3 100.0 137.8 200.3 247.7 377.8 466.1 634.6 1323.2 4 100.0 164.3 233.0 335.7 486.7 583.1 1230.0 2218.8 5 100.0 199.4 478.3 546.4 1176.8 1544.4 2950.0 6125.8 6 100.0 272.2 430.0 785.3 1041.8 1793.7 3910.0 6448.3 IR 2/ 1 1.00 0.75 0.78 0.73 0.82 0.55 0.51 0.18 2 1.00 1.08 .92 1.04 1.10 0.85 0.80 0.84 3 1.00 1.02 1.16 0.97 1.10 1.07 0.89 0.91 m 4 1.00 1.22 1.35 1.31 1.42 1.34 1.72 1.53 5 1.00 1.48 2.72 2.13 3.43 3.50 4.13 4.22 6 1.00 1.69 2.49 3.06 3.03 4.10 5.48 4.45 1. (P.Q.)1957 Notes: 1/ Estimate from: LPI = iit . 100 where t=0 represents 1955-57 r i it=o where P, Q represent the nominal Price and Area of each comuna summed over each ring; r Ring, 2/ Index Ratio = Land Price Index (LPI) Consumer Price Index (CPI) FIG.3 LRND PRICE INDICES : BOGOTR -16- 1958 1950 1962 1964 19,6 1968 197 -.572 1974 1976 500.0 i56500.0 6000 .0- >6000>-0 5500.0 5500.0 5000.0 -5000.0 LLJ 4500.0 4500.0 4000.0 4000.0 - RING "4" 3S00 -0 3500 .0 RIG~5 30000 - RING "5"0 3000-0 - G3000 -0 LL ,- RING "6" 2500.0 2500.0 2000.0 2000.0 Q. 1500.0 1500.0 1000.0 ~ 00. 500.0- 500 0 .0 1958 1960 1962 1964 1966 1968 1970 972 1974 1976 YERRS 1956 1960 19.54 19568 72 1976 6500.0 650.0 5000.0 E ~ 6000.0 5500.0 5500.0 5000.0 -- . I 5000.0 L 4500.0 0- Ring "i" 4500.0 4000.0 - Ring "2" 4000-0 i300 .0 Ring "3" 350r0.0 30000 E-3000.0 L2500,0 -2500.0 2000.-0 2000,0 0 150.0.0 - - 1500 .0 1000 . EO 1000 .0 .0 -0 1955 190 1954 1968 1372 1976 YERRS. Table 5: ESTIMATED LAND PRICE INDEX (LPI) AND INDEX RATIO (IR) FOR EACH RADIAL SECTOR AND THREE-YEAR PERIODS: (1955-57 - 1976-78) RADIAL SECTOR 1955-57 1958-60 1961-63 1964-66 1967-69 1970-72 1973-75 1976-78 LPI 1/ I 100.0 101.1 '135.2 187.7 283.1 238.7 363.6 259.3 II 100.0 156.6 297.5 389.2 813.0 772.5 1812.2 3726.5 III 100.0 187.0 258.0 490.5 1048.0 1022.2 2134.7 3495.9 IV 100.0 140.4 251.7 348.3 532.2 503.6 942.8 1682.9 V 100.0 149.3 165.2 243.3 375.3 448.3 872.0 1971.5 VI 100.0 135.4 401.2 430.3 635.1 1167.6 2040.5 3600.8 VII 100.0 207.1 312.8 527.5 670.9 1030.3 1946.9 3438.0 VIII 100.0 178.8 265.2 285.9 504.0 715.4 1275.1 2995.1 IR 2/ I 1.00 0.75 0.78 0.73 0.82 0.55 0.51 0.18 II 1.00 1.16 1.72 1.52 2.37 1.77 2.54 2.62 III 1.00 1.39 1.49 1.91 3.05 2.34 2.99 2.46 IV 1.00 1.04 1.46 1.36 1.55 1.15 1.32 1.18 V 1.00 1.11 0.96 0.95 1.09 1.01 1.22 1.39 VI 1.00 1.01 2.32 1.68 1.85 2.68 2.86 2.53 VII 1.00 1.54 1.81 2.06 1.95 2.36 2.73 2.42 VIII 1.00 1.32 1.53 1.11 1.47 1.64 1.79 2.11 Notes: (P Q) 1/ Estimated from : LPI = rs i . 100 where t = o represents 1955-57 rs(P Qi t=0 where P, Q represent'the Current Price and Area of each comuna i summed over each radial sector 2/ Index Ratio (IR) = Land Price Index (LPI) Consumer Price Index (CPl) F!G 4 LRND PRICE INDICES IN BOGOTR 1960 1964 196 1972 1376 3500.-0 - I I 3500.0 3250 0 3250,0 3000.0 - 3000.0 x 2750.0 ---C P8=. 2750.0 2500.0 ~RRD SECTOR "' 2500.0 2250.0 _-- x-RRD SECTOR "'2"' 2250.0 2000-0 A-RRD SECTOR "3 2000,0 1750.0 --RRD SECTOR "4: 1750.0 1500.0 1500.0 1250.0 1250.0 Q 1000.0 -1000.0 750.0 750.0 500.0 --- 500.0 250.0 - 250.0 1956 1960 1964 19658 1972 1976 YERPS 195J 1960 1962 1964 1966 1968 1970 1972 I974 1976 35û00.0 I3500.0 3250.0 3250.0 3000.0 '3000.0 2750.0 +-RRO SECTGR 5" 2750.0 S2750.0 _ W2500.0 5-RRD SECTOR "6 2500.0 S2250.0 z-RRDc SECTOR 7 2250.0 -2000.0 Y-RRD- SECTOR "8" 2000.0 1750.0 -1750.0 ILi1500.0 1500 .0 12.50.0 1250.0 ' 1000,-- 1000.0 750.0 750.0 500.0 500-0 250.0 -250.0 .0 0 1355 1960 1962 1964 1966 1963 1970 1972 1974 1976 YERR,S - 19 - The estimated land price index (LPI) is shown for each Ring (Table 4) and radial sector (Table 5) along with an index ratio (IR) which is defined as the ratio of the land price index (LPI) to the Consumer Price Index (CPI-White Collar workers Bogota). In the case of Rings, land prices in Ring 6 have kept well above the CPI, more than any other ring. Also ranked above the CPI are the indices of rings 5 and 4, and below the CPI are rings 3, 2, and 1, all in descending order as it is clearly shown in Figure 3. This in turn reinforces the idea of distinguishing between areas of the city, ring-shaped in this case, as areas of significantly different price behavior. In a similar way, the land price index was calculated for the eight radial sectors. The results are shown in Table 5. Table 5 and Figure 4 show that all sectors except Sector I (the CBD) pre- sent overall price increases above the level of the CPI. Although the relative differences are not very substantial, sectors II, VI, III and VII (South, south-east, and north-east) in that order, have been the sectors where land investors have made, on average, the largest profits. A word of caution is required about the land price index. As any other index or average (e.g., the Dow Jones Stock Average or the Bogota Stock Exchange Indicator) it reflects average variations in the entire portfolio. The fact that one's stock may in fact increase when the stock average falls is equally valid for the land price index case. As an illustration, many transactions have taken place, particularly in Sector VIII, the city north, in which the beneficiaries have reported - 20 - extraordinary profits in relatively short period of time. Our conclusions thus far pertain to sectoral averages, and not to specific transactions. In calculating the land price index, an estimate of the total land price index for the city by year was obtained. Table 6 reports both the LPI and the index related to the City's Consumer Price Index or IR (See Tables 4 and 5 for definitions). As one can see in the Table, land prices have been growing, on the average, at a rate greater than the CPI, thereby confirming that land can be regarded as a good investment or "shelter" against inflation. The rate of growth of land values increased steadily from 1955 to 1966. After that, as shown in the table, it was 12/ not until around 1972, the year of the establishment of the UPAC-2 system, that the 1966 level was achieved again. Since 1972, land prices have shown a strong momentum, to the point of being five and six times higher than the CPI particularly during 1973, 1974 and 1978, all years of great construction activity. A few comments on the nature of results shown in Table 6 must be made. On the one hand it is important to notice that the estimated Land Price Index for the city as a whole (LPI) reflects general business cycles such as those observed in 1961-62 and in 1975. On the other hand one must keep in mind some of the limitations of the index estimated. The most important of such limitations are: first, the fact that the total amount of land in each "comuna" gets an imputed average price independently of its actual use; (e.g., even if it is recreational park- land) and, second, that given that the area of each comuna is maintained constant over the entire period, the larger increases observed in the periphery, (i.e., outermost rings) get weighted more due to their larger areas. The periphery, which was largely uninhabited in earlier years - 21 - Table 6: Evolution of Land Values in Bogota (LPI), as Compared with the General Price Level 1955-1978 Year Index of 1/ Consumer Index Ratio Land Values - Price Index ((1) j (2)) 1 2 3 1955 100 100 1 1956 125 107 1.17 1957 150 121 1.24 1958 175 138 1.27 1959 213 148 1.44 1960 250 156 1.60 1961 275 167 1.65 1962 288 176 1.64 1963 425 222 1.91 1964 438 253 1.73 1965 488 269 1.81 1966 525 318 1.65 1967 638 349 1.83 1968 688 372 1.85 1969 813 402 2.02 1970 813 429 1.90 1971 912 470 1.94 1972 1075 528 2.04 1973 1450 631 2.30 1974 1675 768 2.19 1975 1913 940 2.04 1976 2338 1113 2.10 1977 3175 1396 2.27 1978 4525 1666 2.72 z (P Q ) 1/ Estimated from: LPI L 1 i t 100 where t=o represents 1955 and r it=o z where P,Q represent the nominal Price and Area of each. comuna summed over each ring; r = Ring, i = comuna, t = time period. 2/ Index Ratio =Land Price Index (LPI) Consumer Price Index (CPI) - 22 - registered large proportional increases (starting from low levels) and therefore has a large effect on total increases. Also as an intermediate result from constructing the land price index, an approximate estimate of the "total" land value of the city was obtained. The magnitude of the total land value is then compared to the series of gross domestic and recional products of Colombia and Bogota for the 1960-1975 period. The series of total land value of Bogota and their comparison to both GDP and GRP are presented in Tables 7 and 8. As pointed out above, the rate of growth of real prices in Table 7 comes to about 4 percent per year. The total annual g-.owth, in the terms of the nominal value of all land in Bogota (between 1955 and 1978) was about 18 percent per year, of which about 13 percent can be accounted for by inflation. Table 8 shows the total value of land in Bogota as compared with the gross domestic and regional productivity of Colombia and Bogota respectively for the 1960-1978 period. Average comuna values are computed for each year as before and aggregated for the city as a whole (see Appendix for aggregation procedure). It is found that the estimated value of the gross regional product of Bogota has risen faster than the total value of land. Gross regional product has grown at a rate of about 7 percent per year while land at about 3 percent since 1960. The two growth rates have been approximately equal since about 1970 as evidenced by the stability of the ratio of land values to gross regional product in column 5 of table 5. The ratio fell mainly in the sixties. It may be the case that since the whole periphery has been valued at - 22a - urban land value levels in earlier years as well as when it was largely uninhabited, the total land value for Bogota is overstated for the early years. If only the data for the seventies is considered when the city area has been relatively stable, total land values are seen to rise proportionally with gross product.11/ This is the result expected if the share of land in urban production is constant. When viewed in this context, the observed average price rises in Bogota are then not surprising. To summarize, if the share of land in urban production remains constant, as does the total urban area, then the average price can be expected to rise at the same rate as gross product and hence ahead of inflation in a growing economy. Last, a set of interesting policy questions may arise as a result of the estimation of total land values for Bogota or for any such city as well. What can be the implications of comparing estimated with assessed values? Can any distortions be corrected without affecting first, the income distribution structure; second, the provision of physical services in which user-charge rates are based on cadastral assessments; and third, the level of both government revenues and land use strategies? - 23 - Table 7: THE "VALUE" 1/ OF BOGOTA'S LAND 1955 - 1978 (Millions of Current Pesos) Year Total Value Year Total Value 1955 8,000 1967 51,000 1956 10,000 1968 55,000 1957 12,000 1969 65,000 1958 14,000 1970 65,000 1959 17,000 1971 73,000 1960 20,000 1972 86,000 1961 22,000 1973 116,000 1962 23,000 1974 134,000 1963 34,000 1975 153,000 1964 35,000 1976 187,000 1965 39,000 1977 254,000 1966 42,000 1978 362,000 Note: 1/ Calculated as the Sum of Products between the Average Nominal Price of each "comuna" and its own area in square meters. Table 8: COMPARISON BETWEEN THE TOTAL LAND VALUE OF BOGOTA AND BOTH GROSS DOMESTIC AND REGIONAL PRODUCT FOR SELECTED YEARS (1960-1978) (In millions of Col$ ,of 1970) Approx. Land Gross Regional Gross Domestic Value of Bogota Product Product Year (1) (2) (3) (1). (2) (1)+ (3) 1960 55,000 11,996 77,203 4.59 .71 1967 62,100 18,480 105,723 3.24 .59 1970 65,000 25,920 126,770 2.51 .51 1971 66,400 27,572 134,476 2.41 .49 1972 69,500 30.189 145,051 2.30 .48 1973 70,700 33,209 156,036 2.37 .45 1974 71,100 35,759 166,289 2.10 .43 1975 72,200 36,671 172,300 2.92 .42 1976 72,300 37,771* 179,992 1.91 .40 1977 78,200 40,415* 189,200 1.94 .41 1978 93,300 43,244* 204,591 2.16 .46 Notes: * projected with average annual rate of growth 1/ Calculated 2/ From "Cuentas Regional de Colombia 1960-1975", D.N.P., Nov. 1977, Bogota, p. 38 3/ From "Economic Position and Prospects of Colombia", Report 2535-CO, The World Bank, Vol. II, pp. 10. - 25 - Without intending to answer such policy relevant issues in this descriptive analysis of land-price behavior, some limited information found on cadastral assessed values for Bogota is compared with figures from Table 6 for similar years. Such a comparison appears in Table 9, along with an estimate of the level of "underestimation" of the appraised value from the cadastral record to the value of land itself. It must be pointed out that the cadastral estimates include not only the land value but the construction value as well for the Distrito EsDecial de Bogota (Bogota's Special District). Again a word of caution is necessary with respect to the figures presented in Table 9. According to the limitations already pointed out about the "derived" estimate of total land value of Bogota, the figures of Column (1) assume that all land of the city is assessed at the average price per unit of area in each comuna. The estimated average price for each comuna is based on the latest transactions and therefore reflects current market price for the land. Cadastral values are based on assess- ments done at infrequent intervals of time and are therefore always behind in times of high inflation. This indicates an intrinsic problem faced by municipal revenues because of high inflation. The estimated land value is also high because all land is valued including public use areas like roads, parks, open areas, etc. - 26 - Table 9: COMPARISON BETWEEN ESTIMATED TOTAL LAND VALUE OF BOGOTA, D. E. AND CADASTRAL 1/ ASSESSMENTS: 1974-1978 (Millions of Col$) Estimated Cadastral Land Value Appraisal Ratio Year (1) (2) (1) + (2) 1974 134,000 131160.5 1.02 1975 154,000 139715.2 1.10 1976 187,000 150231.4 1.25 1977 254,000 162867.9 1.56 1978 362,000 176800.7 2.05 Notes: 1/ Cadastr&l figures include both vacant land 2.us construction assessed values. Sources: (1) Calculated (2) Avalilos Catastrales, Cuadro Estadistico por predios y tarifas, Direcci6n Catastro Distrital- Investigaciones y Estadistica, April 10, 1979. - 27 - III. THE BEHAVIOR OF LAND PRICES IN BOGOTA: 1955-1978 A. The Estimation of Price Gradients for the City as a Whole The most common function used in the literature to represent urban population density and price patterns is the so-called negative exponential function. This function is as Mills points out, "a rarity in economics: a functional form derived from a respectable theory that can be estimated and tested with easily available data" (Mills and Tan, 1978). A large body of literature contains estimates of parameters for population density gradient functions.1-/ But as far as land price is concerned, the literature has been relatively scarce because the data has not been easily available. Figure 5, which is the graphical representation of the relation- ship between real prices of land and distance from the CBD, clearly supports the use of negative exponential functions in order to derive land-price gradients. Because much has been written about the function's mathematical properties, only a very brief description will follow; The exponential function is represented mathematically as: P(x).= P(o)e-bx (1) Where P(x) is the price of each comuna located at "x" distance from the CBD. P(o) is the price level of the CBD. e is the base of the natural logarithm and b measures the rate of decline of prices as a function of distance. b is usually referred to as the gradient of the price function. A special interest is focused on b as a measure of decentralization or the so-called suburbanization process. Both P(o) and b are usually used to summarize the structure of cities. Figure S: L RNND PR ICES IN GO TUR , 8 .0 2.0 4.0 6.0 ..0 10.0 12.0 325.0 l 325.0 300. - 300.0 275.0 275.0 250.0 250.0 UJ 225.0 225 . 200.0 :- - 2.00.0 C'175.0 175.0 150.0 1.50.0 J125 .0 ~ z25 . 0 .100.0 100 .0 Q-' 75.0 75.0 50.0 50.0 25.0 25.0 .0 .0 .0 2.0 4.0 6.0 8.0 10.0 12.0 DISTRNCE IN KMS .0 2.0 4.0 6.0 8.û 10.0 12.0 325.0 300.0 1 300.0 275.0 - 275.0 f 250.0 250.0 uJ 225.0-68-70 225.0 2000 Z-1971-73 200.0 175.0 Z-1974-78 - 175.0 150.40 - - 150 .0 1 125.0 125.0 100.-0 å 100 .0 ' 75.0 75.0 50.0 50.0 25-0 -25.0 -0 2.0 4.0 6.0 8 .C 10.0 12.0 DISTRNCE IN K," BRSE YERR=195 - 29 - Equation (1) was estimated using In P(x) = In P(o) - bD (2) i.e. In P(x) = a - bD (2') where a is a constant representing the price level in the urban center or the 'intercept' of the exponential function. The results obtained have been classified into two sets of results. The first set of results repre- sent parameter estimation of equation (2'), that is, estimates of a and b along with the estimate of the distance elasticity of land prices, for the entire city and for every year. And the second set of results represent the same parameter and elasticity estimates for a group of radial sectors, re-grouped in such a way as to coincide with relatively wide but similar land market areas. Results from the first set of regression estimates are shown in Table 10. As one can see in Table 10, there are several results of consider- able importance. One is the almost monotonic decrease in both gradient and elasticity values as well as the constant term. The declining gradient or coefficient of the exponential function, which measures the rate of change of prices with respect to distance, illustrates a "flattening" process of the price function as time passes. This has the effect of making prices "less responsive" to distance; that is, the function has been gradually be- coming more inelastic with respect to distance. The constant term (P(o)), which actually represents the average price at the CBD quite well, also declines as noted before. It is the combined effect of declining prices at the CBD with the relative increases at the periphery discussed in section II which are reflected here by the tendency of the price function to get "flatter" over time. - 30 - Table 10: ESTIMATION OF PRICE GRADIENTS AND ELASTICITIES FROM NEGATIVE EXPONENTIAL PRICE FUNCTIONS OF BOGOTA: 1955-1978 (1) (2) (3) (4) (5) (6) Year Constant Gradient Elasticity R2 SE -RHO 1955 4.83 -.259 1.770 .523 .968 .589 1956 4.80 -.213 1.448 .441 .938 .517 1957 4.93 -.209 1.421 .464 .878 .562 1958 4.85 -.195 1.326 .359 .102 .594 1959 4.97 -.182 1.237 .563 .632 .632 1960 4.85 -.132 .898 .447 .573 .599 1961 4.91 -.149 1.013 .516 .564 .549 1962 4.90 -.135 .918 .405 .641 .599 1963 4.61 -.066 .449 .142 .633 .250 1964 4.75 -.097 .660 .228 .699 .080 1965 5.01 -.154 1.047 .553 .541 .676 1966 4.63 -.114 .775 .356 .601 .604 1967 4.89 -.127 .863 .539 .461 .624 1968 4.93 -.139 .945 .496 .551 .608 1969 4.79 -.101 .687 .478 .414 .564 1970 4.57 -.076 .517 .379 .379 .546 1971 4.77 -.105 .714 .394 .514 .560 1972 4.58 -.069 .469 .228 .493 .349 1973 4.74 -.077 .524 .390 .379 .532 1974 4.59 -.053 .360 .204 .415 .375 1975 4.19 -.011 .075 .089 .361 .263 1976 4.45 -.051 .347 .205 .391 .400 1977 4.66 -.071 .483 .354 .378 .535 1978 4.47 -.025 .170 .071 .349 .241 Notes: (1) The constant term, or "intercept" related to the CBD price. (2) Estimated coefficient of the distance variable or gradient of the function. (3) The elasticity estimated at the mean values of P and D: AP D AD P (4) Coefficient of multiple determination (5) Standard error of the equation (6) Simple correlation coefficient between real price and distance to the CBD. All coefficients are statistically significant at the .001 two-tailed test of significance. - 31 - Both the "goodness of fit" as represented by the coefficient of multiple determination (R 2) and the correlation coefficient (RHo) tend to get worse as time passes. Since the decline in R2 is reflecting large deviations from the curve as time passes, it means that increasingly more important "disturbances" associated with both the dependent as well as the independent variable have appeared in the market. Among such "disturbances" are the increasing importance of other determinants different from the distance variable and the rise of other alternative business districts or commercial areas. In fact it was around the mid-nineteen-fifties when the "Chapinero" area started becoming important as a new commercial district followed later by "centro internacional", "Restrepo", "San Victorino", "Sears", "Siete de Agosto" and more recently, "Lago-Gaitan" and "Unicentro". Such areas of commercial activity, however, can only be analyzed more closely on individual basis and at a higher level of disaggregation than that used so far. The patterns described above are illustrated in Figs. 6 and 7. The former illustrates the steady decline in both gradient and elasticity values for the city as a whole. The latter shows the relationship P(o) vs. b which is quite illustrative: it shows a relative decentralization pro- cess resulting from the combination of declining CAD's prices and decreasing gradients (decreasing the surface slope). That has been due to an increasing importance (i.e., higher price levels) of non-CBD areas, particularly those located far out from the CBD. The forces behind the changes in both P(o) and b - CBD prices and gradient - are numerous and quite important to recognize given that they are - 32 - Figure 6 GRRDIENTS RND ELRSTICITIES e--RRDIENT ý*-ELRSTICITY LRND PRICES IN BOGOTR 1956 190 1964 196p 1972 1976 3.00 30 2.75 -~ 2.75 2,50 ~ 2.50 i- 2.25 *2.25 2 00 2.00 1 .75 - - 75 1.50 - I 1.50 1.25 1,25 1.00 -1.00 'D .50 E- .50 195i 1960 1964 1968 1972 1976 YERRS - 33 - Figure 7 PM0 VS B .00 -02 -04 .06 .08 .10 .12 -14 -16 -18 .20 .22 30 01 C27 5- n525 0 22 5 20 0 C 17 5 i5 0 V- 12 5 10 0 LJ7 5 75 5 0 0 2 5 -00 -02 .04 .05 .08 -10 -12 .14 .15 -18 .20 .22 GRRDIENT (B) - 34 commonly used to represent the structure of the city. In relation to the observed changes in P(o), one can identify one important supply factor, that is, the-negative effect on P(o) of increasing the supply of land within a fixed u,rban boundary. And among the forces that affect the b coefficient are: income, population, transportation costs and amenities. Ceteris-Paribus, the gradient coefficient (b's) will decrease with in- creases in income or population, with decreases in transportation costs and with increases in amenities (e.g., low traffic congestion, low density, clean environment, low noise, more green areas, etc.). B. The Estimation of Price Gradients for Individual Radial Sectors Along with estimates of land-price gradients for the entire city, gradient functions were also estimated for radial sectors and time periods of four years each. The estimates are presented in Table 11. There, it is shown that the gradient estimated for each sector decreases - as for the city as a whole - from period to period although in a slightly different fashion in each sector. All sectors [without exception] show steadily de- creasing gradients throughout the 24-year period. However, sectors II, III and IV decrease their gradients [by much larger proportions]. These three sectors include areas which have developed rather recently or still remain undeveloped. Such is the case of the "Salitre" Park and undeveloped area, the "Kennedy-city" and the Provivienda residential complexes. This fact may in part explain their gradient behavior given that prices elsewhere have not increased as much as in these ssctors as shown in Fig. 4 thus reflecting rising gradient values. - 35 - Table 11. ESTIMATION OF PRICE GRADIENTS FROM NEGATIVE EXPONENTIAL FUNCTIONS FOR DIFFERENT RADIAL SECTOR OF BOGOTA: 1955- 1978 Sector. Period. Constant Gradient "t stat. R2 II 55-58 5.25 -0.481 11.01 .82 i 59-62 4.99 -0.274 8.12 .71. 63.-66 4.97 -0.267 6.32 .60 II 67-70 4.86 -0.203 5.35 .52 71-74 4.71 -0.157 4.84 .47 iI 75-78 4.76 -0.150 5.62 .54 III 55-58 5.33 -0.377 9.03 .75 59-62 5.32 -0.288 9.70 .78 63-66 5.37 -0.237 11.4.4 .83 III 67-70 5.04 -0.171 6.52 .62 iii 71-74 5.01 -0.154 7.63 .69 75-78 4.15 -0.013 1.63 .09 IV-V 55-58 5.24 -0.244 8.33 .67 59-62 5.19 -0.194 9.78 .73 63-66 5.21 -0.201 8.83 .69 IV-V 67-71 5.17 -0.162 10.55 .76 71-74 5.02 -0.153 10.68 .77 IV-v 75-78 4.70 -0.080 5.44 .46 55-58 5.71 -0.370 11.31 .85 59-62 5.59 -0.234 12.84 .88 VI 63-66 5.43 -0.156 6.75 .67 67-70 5.63 -0.179 9.06 .78 VI 71-74 5.40 -0.128 8.23 .75 VI 75-78 5.14 -0.101 7.38 .71 VII 55-58 5.89 -0;287 12.40 .83 59-62 5.79 -0.209 17.71 .91 VII 63-66 5.40 -0.146 11.33 .82 VII 67-70 5.29 -0.135 14.37 .87 71-74 5.18 -0.097 9.19 .73 VII 75-78 4.94 -0.080 8.89 .72 VIII 55-58 5.16 -0.205 14.92 .91 59-62 5.59 -0.163 11.54 .85 VIII 63-66 5.70 -0.159 11.88 .86 VIII 67-70 5.63 -0.161 16.05 .92 71-74 5.40 -0.115 8.50 .76 75-78 5.12 -0.069 5.20 .55 - 36 - By comparing overall gradient levels among sectors it is not sur- prising to find that in the last period all gradients have stabilized to a great extent but sector II and VI show higher relative levels. This comes as no surprise given that population densities are higher in these two sectors as shown elsewhere and as confirmed by estimates of population density gradients. 151/ The land price surface not only presents irregularities, but changes take place over time in a non-uniform fashion.. In fact sectors III and VI, which are located on each side of the so-called Industrial Corridor, show the highest gradient changes during the entire period. Gradient changes in sectors II and VII have been also quite significant over the entire period, but no sector shows a more pronounced change in its price surface than sectors III, IV, V and VIII. C. Comparison of Land Price Gradients With Other Cities As noted in the introduction very few studies have estimated land price gradients mainly because of data availability problems. However, there are two studies that estimated land price gradients based on time-series and cross-sectional data bases. One is for Chicago from 1836 to 1928 and the other is for a group of 70 Japanese cities.1-6/ The results from the two studies are shown in Tables 12 and 13. There are some remarkable similari- ties in relative terms between the Bogota results and some individual Japanese cities in terms of gradients (although in general Japanese cities tend to have much "flatter" surfaces at similar CBD levels). The overall comparison shows that the two parameter values, namely the constant and the gradient of the negative exponential function, can be very useful in terms of summarizing the structure of the city. Furthermore, the evolution of such parameters reveal similarities between the "dynamics" of the urban land prices in Bogota and Chicago. In each case the evolution of gradients and elasticities show a very similar pattern of decline over time. - 37 - Table 12: CHICAGO LAND VALUE GRADIENTS: 1836-1928 1 Year Regression Constant Gradient ElasticitY2/ R2 N 1836 Exponential 5.632 -0.403 -3.12 0.781 208 (44.832) (27.168) Maximum likelihood 3.247 -0.197 -4.53 0.834 208 (62.550) (36.800) 1857 Exponential 8.748 -0.513 -4.04 0.858 211 (70.886) (35.742) Maximum likelihood 6.791 -0.365 -4.69 0.866 211 (79.751) (36.800) 1873 Exponential 9.980 -0.344 -2.66 0.682 207 (71.655) (21.011) Maximum likelihood 6.386 -0.161 -2.97 0.684 206 (98.374) (21.105) 1892 Exponential 10.043 --0.246 -1.97 0.418 173 (52.558) (11.169) Maximum likelihood 6.611 -0.134 -2.42 0.401 173 (72,400) (12,678) 1910 Exponential 10.584 -0.319 -2.35 0.566 123 (52.018) (12.678) 1928 Exponential 11.736 -0.220 -1.74 0.497 139 (72.390) (11.735) 1/ t-values are in parentheses 2/ This is the elasticity of land value with respect to distance. 31 The exponential equation is the maximum likelihood estimate for 1910 and 1928. Distance is measured in miles. Source: "Urban Land Value Functions and the Price Elasticity of Demand for Housing" by J. B. Kau and C. F. Sirmans (Journal of Urban Economics,.Vol. 6, No. 1, (1979); pp. 117, Table 1. - 38 - Table 13: LAND PRICE GRADIENTS OF SELECTED JAPANESE CITIES Sample Price Gradient 2 Size at CBD (b) t R 1 105 4,071 0.02885 (12.309) 0.533 2 21 3.183 0.00469 (0.486) 0.013 3 38 2.689 0.02652 (2.905) 0.190 4 19 3.608 0.02039 (1.582) 0.128 5 26 3.381 0.02882 (3.555) 0.345 6 24 3.380 0.02145 (1.535) 0.097 7 59 4.176 0.02458 (6.445) 0.421 8 89 3.363 0.04470 (7.471) 0.601 9 13 3.787 0.07484 (3.243) 0.489 10 29 3.291 0.03514 (3.091) 0.262 11 40 3.670 0.04453 (3.224) 0.215 12 60 3.950 0.02557 (6.436) 0.416 13 25 3.872 0.03604 (4.342) 0.451 14 27 4.569 0.02435 (3.472) 0.325 15 28 4.732 0.03642 (6.356) 0.608 16 31 4.413 0.02644 (5.096) 0.472 17 68 4.323 0.02160 (4.891) 0.266 18 26 4.677 0,02915 (4.709) 0.480 19 42 4.390 0.01706 (3.759) 0.261 20 53 4.135 0.01329 (2.720) 0.126 21 33 3.977 0.00386 (1,463) 0.063 22 29 4.386 0.01393 (5.038) 0.485 23 33 3.796 0.00759 (1.682) 0.083 24 51 4.369 0.03266 (8.955) 0.621 25 37 3.813 0.05468 (9.938) 0.738 26 38 4.102 0.04337 (7.456) 0.607 27 16 4.188 0.05541 (7.169) 0.786 28 16 3.574 0.03955 (1.823) 0.192 29 21 3.601 0.04981 (5.193) 0.587 30 22 4.345 0.05479 (4.387) 0.490 31 39 4.227 0.03535 (6.837) 0.558 32 45 3.651 0.02132 (5.498) 0.413 33 21 3.905 0.02044 (2.146) 0.196 34 27 3.641 0.01403 (1.650) 0.098 35 28 3.416 0.01450 (3.644) 0.338 36 90 3.525 0.00676 (0.564) 0.045 37 21 3.170 0.01646 (3.567) 0.402 38 32 3.218 0.01519 (2.144) 0.133 39 26 4.133 0.01476 (3.929) 0.391 40 92 5.102 0.02534 (11.754) 0.607 41 64 4.295 0.01091 (4.352) 0.234 42 28 4.530 0.00563 (0.847) 0.028 43 31 4.383 0.00066 (0.139) 0.001 44 24 4.418 0,01934 (3.241) 0.324 Note: Only the first 44 out of 71 cities were selected. The names were omitted for simplicity,. Distance is measured in kilometers. Source: N. J. Glickman and Y. Oguri, "Modeling Urban Markets: The Case of Japan." Journal of Urban Economics, 5, 505-525 (1978), pp. 512; Table 2, - 39 - IV. CORRIDOR ANALYSTS It is believed that the analysis of land-price variations along selected streets which resemble both radial and ring-shaped corridors, throws some additional light in the overall picture illustrated so far.17/ Such an exercise provide a close up look at cross-sections or "cuts" which show certain characteristics of the price surface not shown in the analysis based on comuna averages. As before, the analysis is bi- dimensional in terms of both time and space variation of average prices. Four basic conclusions are derived when analyzing price varia- tions along specific corridors: (1) Although the shape and dynamics of the price 'surface' around principal radial streets (e.g., "Carrera 7", "Carrera 13", "Caracas", "El Dorado" and "Jimenez" - see Map 4) are similar to the shape and dynamics of the entire surfaces of the city, individual price "peaks" are not necessarily located in the CBD. (2) The three main radial corridors of Bogota: "Carrera 7" 18 / or K7- , "Carrera 13" or K13 and "Avenida Caracas" or K14 show a displacement of their surface peaks towards the north, while the principal ring-shaped corridors, have remained relatively static, with "peaks" around rings 1 and 3 and "valleys" around rings 2 and 4. - 40 - (3) At the corridor level of disaggregation there are two areas that show the highest level of absolute prices, namely, the area located between "Calle" 45 and "Calle" 60 and "Carreras" 13 and 14; and also the area located between "Carreras" 8 and 30 within "Ave Doralo" and "Av. Jimenez". These two areas within the corridor perspective play the role of the CBD in and around their surroundings. (4) Sections north and south of ring-corridors illustrate quite clearly the pattern already stated: places around the CBD tend to decrease in price while those located at the periphery increase their level more than proportionally. As a result the surface covering such areas get flatter over time as pointed out before. The relevance of these four conclusions is that they illustrate quite nicely that there are certain costs associated with looking at the price 'surface' as a whole; that is, the level of generality does not allow one to 'detect' specifi7 'peaks' or sharp irregularities than can only be observed when a more rigorous and detailed analysis is made. A rather similar problem occurs when one focuses attention on average stock indicators for the stock exchange market. Such indicators represent the net effect of all changes and may be pointing a different direction from that followed by a specific stock that deviated signifi- cantly from the average. In terms of land prices there are not only one but many peaks, some of which even challenge the primary level of the CBD, but which get overshadowed by analyses in terms of averages. - 41 - Table 14 and Figures 8 and 9 show the evolution of land prices along the corridors selected for periods of three years. As with the previous analysis, the "radial" corridors are selected streets that run from the CBD or thereabouts towards the periphery. Similarly, 'ring' corridors are set of streets sections that follow the pattern of pre- viously defined rings. Table 14 is a summary table from both Table 1-8 and 1-9 of Appendix 1. There one can see the 'shifting' pattern of the price 'peak' for each corridor along with the price increases registered at the periphery. It is not difficult, however, to associate some of the significant price variations with the "birth, rise and fall" of some commercial areas like Restrepo, San Victorino, Centro Administrativo located along the west section of rings 1, 2 and 3.respectively; and others like Centro Internacional, Chapinero and Lago located between K7 and K13-K14 between C26-C32, C45-C60 and C72-C80. Summarizing, Radial corridors show a consistent displacement of peak land values away from the CBD together with a relative decrease of such values around such peaks and a noticeable increase in peripheral values. As far as ring corridors are concerned, Table 1-8 shows a clear tendency for real prices to decrease around "inner" rings (i.e., rings 1 and 2) and to increase in the "outer" rings (rings 3 and 4). Such pattern, however, is more emphasized along sections north and south of these rings than in section west or namely the ineustrial a:ea of the city. 42- Table 14: FOUR-YEAR REAL- AVERAGE LAND PRICES ALONG THE MAIN "CORRIDORS" 2/ OF BOGOTA, 1955-1978 (in Col$) Corridor. 1955-58 1959-62 1963-66 1967-70 1971-74 1975-78 TOTAL Radial Carrera 7 241 196 203 198 149 157 196 Carrera 13 246 295 255 236 208 122 244 Av.Caracas 121 137 131 124 80 143 128 Av.Dorado 117 108 115 -101 112 89 109 Av.Jimenez 178 171 240 347 102 73 189 Other 3/ 95 92 84 77 84 72 84 Rig Ring 1 254 160 119 266 -- 139 197 Ring 2 77 60 69 80 96 94 77 Ring 3 149 142 133 118 130 103 133 Ring 4 44 28 49 48 62 112 53 Other 3/ 99 99 91 94 81 72 89 Notes: 1/ Real (1955 = 100) weighted average prices 2/ See Map 4 and Notes for Tables A.8 and A.9 3/ "Other" refers to the rest of the city not including the above radial or ring corridors. -43- EVOLUTION OF LAND PRICES ALONG MAIN "RADIAL" CORRIDORS Carrera 14 Carrera 13 300 - - - - -C4"-0 200 CS1-72 800 20 13-26 46-60 1 C1326 400 10 C2-45 27-45 1000- 55-58 59-62 63-66 67-70 71-74 75-78 CALLE 26 200 10K 8-30 -C13-26 1,0 -1C46-60 K11-7 100 b C27-45 0 5-58 59-62 63-66 67-70 71-74 75-78 0 ~ Carrera 7a, 55-58 59-62 3-66 67-70 71-74 75-78 800 -r7 7-13 Ave. Jiménex 0 K1-7 700 6006 500 101- 400-[ f400 -300 [-C l14-26 300 200 -C27-45 C46-60 -- 1.772, 100 - --30 0 C73- 100 K31-501 C- 2K1 -7 -C1-105 5 59 6 55-58 59-62 63-66 67-70 7 1-74 75-78 07 55-58 59-62 SU-6 67-70 71-74 75-78 44 - Figure 9 EVOLUTION OF LAND PRICES ALONG MAIN "RING" CORRIDOR SECTIONS NORTH 900 CS61(K 1,3R) 800 . 300 200 C26 K1-14 100 44(K 1-30) C100 ( 7-45 56.58 59-62 63-66 67-70 71.74 75-78 SOUTH WEST 600 400 13(K1-4 14____ ____ ____ 500 300 14(Cl'1 200 200 KS0(C13-26 00 100 100- . Kw(CS-26) 55-o8 9-62 6A6-C1-81 5r55a 59-62 63-66 777 7-4 5-78 558 59-62 63-66 67-70 174 75-78 - 45 V. CONCLUSIONS At the aggregate level Bogota still remains very much a mono- centric city according to the evidence of land-price levels, although it is usually referred to as a multi-center city. Average land-prices in the CBD, although decreasing since 1955, are still the highest prices registered in the city even as late as 1978, the last year for which observations were available. Again, this conclusion runs somewhat against a wide-spread belief that Bogota has several alternative business districts . ompeting with the traditional downtown. However, this does not contradict a well-established fact that prices in non-CBD areas have increased con- siderably in the last decade. Indeed there have been relatively large price increases along the periphery of the city. This, expressed in different words, means that the land price surface continues to be its highest "peak" around the CBD and that the entire surface has tended to get 'flatter' as a result of this dual set of forces: decreases in the level of prices of the CBD and increases in prices (relative to those of the CBD) along the peripheral ring. The ring sectorization seems to be appropriate in terms of iden- tify±ng different land market performance. It reflects well-defined areas with significantly different price behavior. However, the Radial sector sectorization becomes very useful in terms of areas with different socio- economic conditions and different residential and industrial characteristics. In terms of radial sectors there are three sectors, namely sectors II, III and VI, which show the large price increases, all well above the inflation rate level. Other sectors, although showing relatively good performance in - 46 - terms of price increases, lag behind these three sectors. This is even the case in sector VIII (the city north) regarded to be the "best investment alternative for land in Bogota". The patterns stated above are confirmed when certain specific corridors are selected for analysis. Similair to the shape and dynamics of the city price surface, the individual corridor's surface shows a tendency to get flatter over time with the greatest increases around the periphery. The highest values of prices in corridors, however, do not necessarily co- incide with the Central Business District (CBD). These higher points or 'peaks' remain 'hidden' when a more aggregate analysis is chosen thus becoming overshadowed by the CBD. As such, it appears, even as late as 1978, that the price surface can be characterized by a single really impor- tant peak which, although losing relative height as time passes, has not been seriously "challenged" by other price 'peaks' around alternative or secondary business districts such as San Diego or "Centro Internacional", Chapinero, Sears, Siete de Agosto, Lago, Unicentro, or any other. Two structural parameters that are commonly used to describe land- price surfaces, namely, the gradient and price elasticcy with respect to distance, show remarkable consistency in illustrating the trend stated above. The city's overall gradient decreases systematically, thereby show- ing an increasingly flatter surface, which becomes more inelastic with respect to the distance to the CBD. That is, as the surface gets "flatter" over time, price changes respond less and less to changes in distance. The former effect has a dual cause: while prices at the CBD, i.e., the surface 'peak', decrease in absolute terms (i.e., real terms) the periphery - 47 - increases more, relatively speaking, than the intermediate areas. The latter effect appears to be cause, among other things, of the increasing complexity in the determinants of the land prices. In other words, as the land markets have become relatively more complex, other factors such as neighborhood quality, environmental consideration, infrastructure, relative density, etc., may play an increasingly important role in affect- ing price changes. Finally, from a comparison of the structural parameters mentioned above with those of some other cities like Chicago and several Japanese cities, it is clear that there are striking similarities in the behavioral pattern of such parameters. Without wanting to be conclusive about such similarities it is tempting to believe that urban land markets present two basic characteristics: regularity and a similar dynamic trend toward de- centralization of prices tat reflect the universal decentralization trend of business activity, jobs, commercial centers and population. - 48 - FOOTNOTES 1/ See Muth (1969); Mills (1972); Mills and Song (1977); Mills and Tan (1978); Berry et al (1963); Brush (1968); Cassetti (1967, 1969); Clark (1951, 1957); Greene and Barnbrock (1978); Guest (1975); Ingram and Carroll (1980); Newling (1964, 1966, 1969); Papageorgiou (1971); Mohan (1979). 2/ The only studies known to the author are those done by Glickman and Oguri (1972) and by Kau and Sirmans (1979). 3/ The basic unit of analysis used here is the so-called comuna, a grouping of barrios defined by DANE (Departamento Administrativo Nacional de Estadistica), the National Institute of Statistics. Bogota has 38 comunas which contain about 700 barrios. 4/ Guillermo Wiesner, (1980). 5/ A second document is in process which reports on a similar analysis on individual observations rather than comuna averages. Appendix 2 illustrates how weighted moving-average prices were estimated in the corridor analyses. 6/ The latest estimate for Bogota's current area is approximately 30,800 hectares. Appendix 2 explains the procedure followed to calculate average prices weighted by area. 7/ It has been a popular belief that comunas 83 and 84, which form part of the Sector VIII, are where the highest rates of price increases have taken place. Indeed, this may have happened. However, because Radial sector VIII also includes 'comunas' 81, 82 and 85 (see Map 3), sector VIII's average shows that increases have been moderate. - 49 - 8/ Departamento Administrativo Nacional de Estadistica, DANE, "Estratificacion Social de los Barrios: Bogota"; 1976. 9/ A second phase of this analysis includes a document on "Determinants of Land Prices in Bogota", currently underway. 10/ The highest transaction ever registered up to mid-1979 was, according to Guillermo Wiesner, a lot located at the "Avianca" Building (K7-C16) for which the unit price paid was Col. $21,500 per sq. meter. 11/ Because all transactions refer'to empty lots, the number of observa- tions in the CBD declines considerably over time. Thus prices at the CBD reflect the average price of increasingly fewer transactions. 12/ The UPAC system consists of indexing financial instruments related to construction activities giving additional incentives to housing construction investment. 13 Interestingly enough the ratio of flow (i.e., GRP) to the stock (e.g., the value of land) is roughly the same as that estimated for the U.S. and Korea. See Mills and Song (1977). 14/ See Footnote 1. 15/ See Mohan, R., (1979) for figures on Population and Density distri- butions by ring and sector. 16/ See Kau and Sirmans (1979) and Glickman and Oguri (1978). 17/ The idea of doing corridor analysis was first suggested by Oscar Borrero of CENAC who has been doing extensive work along these lines. The author is also grateful to him for other related comments. 18/ As a general rule for following paragraphs the letter "K" stands for "carrera" the north-south streets and the letter "C" for "Calle" the east-west streets. - 50 - APPENDICES Appendix 1 Maps 1 - 4 Tables 1-1 to 1-10 Appendix 2 Averaging Land Prices - 51 - APENI MP i BOGOTA, D.E. > 31 32 Cra. 15 "aa 84 Av. Ca1c2 gade62 &6 to22 e 55 t-ogon i 三 MA P MAP OF BOGOTA, D.E. "RADIAL" AtID RIN15` CORRIDORS 31 3 al 13 Cra. 15 /' 1 12. 74 702 . . f 63 21 p,( z 52 22 8 \03 (D3 Av 55 44 e24 Suba Aarqluorio Eldorado N Rio n ogotå - 56 - Table 1-1: FREQUENCY DISTRIBUTION OF LAND PRICE DATA SET BY COMUNA IN BOGOTA: 1955-1978 No. of Data No. of Data Comuna Points % Comuna Points % 11 54 0.87 43 81 1.31 21 120 1.94 53 133 2.15 31 585 9.44 63 85 1.37 41 374 6.03 73 235 3.79 51 153 2.47 83 400 6.45 61 324 5.23 14 63 1.02 71 334 5.39 24 39 0.63 81 117 1.89 44 96 1.55 91 . 127 2.05 54 73 1.18 12 120 1.94 64 129 2.08 22 51 0.82 74 123 .198 32 146 2.35 84 208 3.35 42 185 2.98 25 42 0.68 52 114 1.84 45 78 1.26 62 264 4.26 55 124 2.00 72 310 5.00 65 52 0.84 82 420 6.77 85 195 3.15 92 7 0.11 56 83 1.34 13 17 0.27 66 4 0.06 23 95 1.53 - 57 - Table 1-2: DEFINITION OF "RING" AND "RADIAL" SECTORS 1/ Ring Comuna (See Maps 1 and 2) 1 (CBD) - 31 2 32, 41, 61, 71, 81 3 11, 12, 21, 42, 62, 72, 82 4 13, 14, 22, 23, 43, 44, 51, 52, 63, 63, 74, 83 5 24, 25, 45, 56, 64, 53, 54, 55, 65, 85 6 85, 91, 92 Radial I 31, II 11, 12, 13, 14, 25, 32 III 21, 22, 23, 24, 44, 45 IV )41, 42, 43, 65 v )/ 61, 62, 63, 64 VI 51, 52, 54, 55, 56 VII 53, 71, 72, 73, 74, 91, 92 VIII 81, 82, 83, 84, 85 1/ DANE's zone system. 2/ Central Business District 3/ Industrial Corridor Table 1-3: DISTRIBUTION OF NUMBER OF OBSERVATIONS BY RADIAL SECTORS AND BY THREE-YEAR PERIODS: 1955-1978 Radial Sector 1955-57 1958-60 1961-63 1964-66 1967-69 1970-72 1973-75 1976-78 Total I 75 91 123 69 88 47 55 37 585 II 27 69 98 64 49 27 45 63 442 III 31 59 80 59 36 50 78 86 479 IV 72 95 96 36 70 81 95 96 641 V 106 145 138 77 93 61 92 90 802 VI 15 44 74 70 56 65 116 107 547 VII 165 207 243 153 103 140 139 119 126 VIII 177 186 208 163 205 122 135 139 133 Total 668 896 1060 691 700 593 755 737 6100 Ln co Table 1!4: DISTRIBUTION OF NUMBER OF OBSERVATIONS BY RING AND BY THREE YEAR PERIODS: 1955-1978 RING. 1955-57 1958-69 1961-63 1964-66 1967-69 1970-72 1973-75 1976-78 TOTAL 1. 75 91 123 69 88 47 55 37 585 2 208 433 223 139 137 101 125 119 1,285 3 193 243 240 194 149 116 151 161 1,447 4 135 213 224 200 176 162 181 206 1,497 5 40 73 132 118 104 128 185 178 958 6 17 39 37 37 46 39 58 55 328 TOTAL 593 801 856 688 612 546 700 719 6,100 , ' � ,7 4 TыLIe I_5г 810N19iLY SERCES OF GON5UCIER YNiCE 1NUFdC FОК BUCOTA (Ч1+1Се Co1lnr Workвca� YEAR JhN. FEB. tlARCП �РКТ4 ИАУ JONE У1;ПR .TULY Аtьг. SCP'Р. г1С�Г, 1�l�V 1ьf_С. ].954 - - - - - - 1954 9П.5 � 98.3 37.7 97.6 99.1 99.'1 1955 100.0 100.9 101.0 102.Э 101.9 101.2 1Э55 10П.5 100.5 1ПП.2 101.2 ]01.7 1П2.0 195b 103.0 103.6 10Ч.6 105.5 106.7 10G.9 195G 107,4 107.5 107.G 100.2 109.3 109.5 1957 110.5 1]1.7 11Ч.4 116.0 117.3 119.9 195`1 122.+[ 123.9 12G.Э 127.8 128.� 129,0 195U 129.8 130.1 1Э2.9 135.И 139.0 1Э9.7 1958 1Э9.5 139.G 1ц0.5 1�ы0.G ]+ь2,0 142.4 1959 1+U1.7 145.4 147.7 1+i9.7 148.б 14д•2 1959 '1+ьII.1 147.3 1Ч7.8 1+ь0.2 14Э.2 1SП.1 19G0 150.7 150.3 15Э.2 153.9 154.6 155.2 1960 156.3 156.8 15G.7 157.D 15D,8 159.9 1961 159.9 162.D 1Gц.П 167.5 170.4 169.0 19б1 1G9.5 1G8.II 1G0,7 5.69.П 169.4 2%0,3 19G2 170.i[ 17D.0 17Э.0 175.�i 175.0 176.1 ]9G2 177.4 1,77,3 178.6 180.П 100.G 101,4 1963 106.5 199.0 2]0.3 21Е.Ч 221,G 226.7 1963 226.9 220.Е 2Эh.Ч 2Э4.h 237.7 2)9.G 1064 2И1.5 2+12.1 247.4 254.'д 261.7 261.1 196k 25G.6 253.7 25'L.9 252.5 '15?.5 254.D 1965 255.2 256.9 258,9 262.7 2G+ь.9 260.5 19G5 260.4 271.3 27+ь.2 277.9 2fl].G 289.7, 196G 292.5 297.ц 307.8 31Ч.5 321.6 Э20.4 19G6 321.2 Э22,5 327.0 33b.2 3Э1.б Э33,1 1967 336.'1 ЭЭ7,И 34].Э 342.6 3+17.U 351,4 1967 350,9 352.5 353,4 355.у 35lt.0 J59.3 1960 3G2.6 363.0 Зб?.3 З?'.Э Э?2,3 373.5 1960 373.2 Э72.0 373.5 37G.9 ЭОП.] 33D.7 19G9 3G3.4 30+1.6 30II.11 39+1,0 402,4 405.3 19б9 +107.2 +ь07,9 410.2 h12.0 ц15.0 +s17.U 19'1П 4i5.1 4]5.9 Ч21.6 Ч2.�.,Э h30,5 h33.0 1970 q32.0 h31.0 h32.8 ц3б.2 43Е.Ч 4a9.5 1971 +ь4Э.5 k50.4 457.9 цБЭ.Ч ц67.+ь 4Ge.h 1971 47Э.0 h70.G 479,i 4В5.� 4[1G:fl 409.+ь 1972 493.Э 501.1 504.,5 51D.2 513.7 523.G 1972 ya3.q 536.Э SqИ,Ч 551.5 562,.', 56q,7 19'13 569.1 57'l.1 59"1.2 б22,8 G37.2 б42.9 1973 641.3 Gh3.1 699.5 GS2.П 659.3 L71.0 1974 G04,9 7П2.4 72tЭ.2 '141.1 7+17.D 756.7 1974 763.1 770.2 789.4 B18.G Н37.5 8h9.5 1975 В77.+1 092.0 9]7.7 93G.5 940.0 95ц.1 -1975 9ч6.7 9Э9.9 949.5 96б.5 973.2 .90i.П 197G 1001,ц 1035.G 1044.8 1056,2 1075.0 1097.Э 197б 12П.8 ]142.3 1170.G 1179.2 120П.4 122ц.9 1977 1220,8 126G.П 1ЭОЧ.9 1373.2 1425.4 1447.5 1977 1+[55.6 1437.1 1437.J Лl41.G ]+ьSП.3 1Ч73.2 ь 1J/0 1507.7 15GЭ.б 1G[1G.9 1G4II,9 1б77.6 1692.1 19Т0 1G04.6 1fi79.7 1693.6 1714.3 1748.9 1783.2 + и � Snurco: 11ЛtдЕ xnd U u+cn де 1а BepuLlLca: 19?0-1у79 61 låble 06 AMMUAL CONSUMER PRICE INDEX ( White-Collar workers) FOR BOGOTA ( July 1950 1955- 1978 1955 100.0 1956 106.7 1957 120.7 1958 137.6 1959 147.9 1960 155.3 1963, 167.4 1962 176.4 1963 221 .9 1964 252.6 1965 269 .1 1966 318.3 1967 343.9 1968 372. 3 1969 402.4 1970 429,2 1971 470.3 1972 528.3 103 630,5 1974 765.8 1975 940.2 1976 1. 113,0 1977 1.396.0 1978 1.656.8 The figure correspond to the monthly averages used to deflate all nominal land prices. SOURCE: Banco de la RepÖblica. Boletin Mensual 1975-79 Table 1-7 MOVING2/ WEIGHTED3/ REAL 31LAND PWICES IN BOGOTA BY COMUNA" AND SELECTED PERIODS 11 12 1.3 14 21 22 23 24 25 31 82 4.1 1956-1960 33.3 57.2 20.1 27.5 58.0 43.9 28.6 20.0 11.2 379.5 71.5 112.1 1961-1964 38.8 62.1 29.9 45.2 76.7 60.3 43.4 33.1 29.4 333.9 62.4 106.1 1965-1967 41.8 64.1 30.0 48.9 76.2 65.5 49.3 13.6 20.3 242.6 64.9 140.7 1968-1970 44.8 57.4 37.2 52.2 58.3 46.6 48.8 46.0 40.2 341,4 43.8 136.1 1) 1971-1973 50.0 55.4 43.9 44.8 54.8 55.3 68,2 47.2 41.5 329.0 49.5 98.3 1974-1976 55.4 63.7 55.0 44.6 69.7 53.9 63.2 46.6 37.8 256.2 78.7 100.1 1977-1978 50.8 60.5 64.0 42.4 69.0 51.0 49.1 43,2 39.0 259.2 87.0 82.5 42 43 44 45 51 52 53 54 55 56 61 6k 63 64 1956-1960 59.1 40.8 18.9 11.9 72.9 44.6· 42.3 22.0 50.0 3.0 146.0 82.0 36.0 26.0 1961-1964 62.3 65.9 38.7 11.9 74.9 53.6 54.0 70.0 42.0 35.0 151.0 85.0 26.0 31.0 2) 1965-1967 68.8 77.2 40.1 31.2 75.6 57.2 55.8 45.0 47.0 14.0 140.0 74.0 32.0 21.0 1968-1970 61.8 53.7 42.6 33.2 58.6 56.3 56.0 64.0 48.0 31.0 123.0 81.0 47.0 26.0 1971-1973 64.5 48.2 49.0- 32.2 62,4 58.2 69.5 60.0 46.0 49.0 115.0 73.0 42.0 34.0 1974-1976 62.1 50.9 48.4 41.1 68.2 56.9 78.0 63.0 50.0 49.0 107.0 81.0 54.0 43.0 1977-1978 55.2 52.6 46.3 36.9 62.0 53.2 76.6 63.0 51.0 48.0 107.0 80.0 56.0 58.0 Table 1-7 (Continued) ... MOVING 2/ CIGlTED3/ REAL 3/ LAND PRICES IN BOGOTA BY " COMUNA" AND SELECTED PERIODS 65 71 .72 173 74 81 82 83 84 A5 91 92 1956-1960 16.0 122.0 112.0 102.0 86.0 151.0 132.0 89.0 29.0 13.0 9.0 10.0 1961-1964 57.0 122.0 116.0 106.0 94.0 128.0 147.0 99.0 50.0 15.0 23.0 11.0 1965-1967 32.0 100.0 129.0 90.0 83.0 133.0 110.0 83.0 43.0 10.0 33.0 17.0 1968-1970 46.0 90.0 137.0 88.0 73.0 157.0 135.0 101.0 53.0 19.0 24.0 31.0 1971-1973 50.0 99.0 150.0 105.0 80.0 160.0 130.0 114.0 59.0 28.0 42.0 44.0 1974-1976 55.0 87.0 99.0 105.0 72.0 166.0 127.0 150.0 80.0 24.0 60.0 48.0 1977-1978 57.0 86.0 115.0 98.0 67.0 210.0 11,6.0 145.0 98.0 39.0 50.0 30.0 Notes: (1) Three-year moving averages (2) Weighted by area (3) 1955=100 ( See consummer price index attached ) -64- Table 1-8:AVERAGE I REAL 2/LAND PRICES OF "RING CORRIDORS" IN BOGOTA FOR FOUR-YEAR PERIODS BETWEEN 1955 AND 1978 Rinq 4/ Sections. 1955-58 1959-62 1963-66 1967-70 1971-74 1975-73 Ring 1 C.26 (K 1-14) 343 203 170 231 -- 75 K.14 (C.13-26) 228 332 264 --- --- 269 0.13 (K.1-14) 581 109 215 232 --- 75 R.i.ng 2 C.34 (K.1-30) 90 --- --- -- - K.30 (C.6-26) 68 97 --- 77 166 126 C.6 (K.1-30) 84 60 69 81 73 66 Ring 3 C.61 (K 1-39) 175 331 119 66 878 -- K 50 (C 13-26) 147 142 140 120 113 109 C 1 ( K 1-49) 52 71 66 52 63 49 Ring 4 C 100 (K 7-45) 25 19 41 21 --- 159 A.68 (C 1-81) 65 48 50 57 62 C 26 (K 1-55) 21 -- 64 -- -- 65 NOTES: 1/ Average for individual observation en each corridor and 2 adjacent streets. 1 2/ Real Prices ( 1955=100) 3/ See Map 4 for illustration of corridors selected. 4/ Each " Ring section-" correspond to the "calle" or carrera portion of the ring, "calles (c) ", run from east to west and " carreras (k)" run from south to north. -65- Table 1-9: AVERAGE 1/ REAL LAND PRICES OF RADIAL CORRIDORS IN BOGOTA FOR FOUR YEAR PERIODS BETWEEN 1955 AND 1978 RADIAL SECTOR 1955-58 1959-62 1963-66 1967-70 1971-74 1975-73 Carrera 7 C 1$-10S 31 54 38 --- 72 66 C 1 6 58 48 41 55 55 77 C 7 13 799 375 560 542 139 150 C 14 26 408 334 274 314 257 343 C 27 45 181 115 112 188 299 214 C 46 60 93 125 101 100 124 1-3 C 61 72 92 131 105 115 140 166 C 73 100 72 105 77 92 91 140 Carrera 13 C 13 26 315 291 259 193 253 136 C 27 45 169 223 331 145 885 80 C 46 60 278 429 291 872 151 133 C 61 72 115 154 . 89 336 91 133 Carrera 14 C 13 26 161 150 131 --- 76 133 C 27 45 115 135 129 93 33 114 C 46 60 79 140 150 158 78 253 C 61 72 115 130 105 111 81 192 Calle 26 K 1 7 111 94 79 49 .74 49 K 8 30 121 116 133 128 146 103 K 31 50 95 104 109 94 107 64 Avenida Jimenez K 1 7 334 132 494 685 88 50 K 8 30 197 266 222 227 114 98 K 31 20 82 82 73 77 57 69 K 51 72 35 42 38 61 61 82 Notes: For explanation of 1/, 2/ and 3/ see Table 4/ Each "R2dial Section" corresponds to parts of the radial corridor and adjacent streets that show significant number of observations. - 66 - A P P E N D I X 2 ___________________ -67- Averaging Land Prices In this appendix I outline the method used to obtain real, weighted average-moving land prices used i1n the preceding sections. 1. Real Prices: All nominal prices were deflated by the monthly series of the consumer price index of white collar workers for Bogota between 1955 and 1978, according to the month of the transaction observed. The monthly series appears in Tablel.5 and sumary Table 1.6 showing yearly averages. 2. Weighted Averages: A staged method was followed to obtain weighted land price averages for comuna, rings and sectors. i) Comuna average by year: the average price for each comuna for each year was calculated as follows: ct i it it E ait i where P is the average price for comuna c fo: year t. ct Pit zd ait are the price and area respective.y of lot i observed in comuna c and year t. -68- ii) Ring and sector averages: the average price for each ring or sector was calculated as follows: rt c Pct Ac C where Prt is the average price for ring (or sector) r at time t. Pct is the average price for comuna a at time t calculated in (i) above (for comuna s in ring (or sector) r) and Ac is the area for comuna c. iii) City wide averages: the city wide averages were also calculated using comuna wide averages: S= ZP A t c ct c c Ac where Pt is the average price for the whole city. 3. Moving Averages: From the averages calculated above, three year moving averages were calculated for each "comuna". As such two observations were lost in the process, namely the end point observations. The result of this process is illustrated in Table 1-7. -69 - REFERENCES Alonso, William, "Location and Land Us-", Harvard University Press, Mass. (1964). Berry, Brian J.L., James W. Simmons and Robert J. Tennant, "Urban Population Densities, Structure and Change," Geographical Review 53 (1963), 389-405. Brush, John E., "Spatial Patterns of Population in Indian Cities," Geographical Review 58 (1968), 362-391. Casetti, E., "Urban Population Density Patterns: An Alternate Explanation," Canadian Geographer 11 (1967), 96-100. . "Alternate Urban Population Density Models: An Analytical Comparison of Their Validity Range," in Studies in Regional Science, A.J. Scott (editor), (1969) 105-113. Chinitz, B.C., City and Suburb, (ed.) (1964). Clark, Colin, "Urban Population Densities," Journal of the Royal Statistical Society 114 (1951), 490-496. , "Urban Population Densities, "Bulletin de L'Institute International de Statistique 36 (1957), 60-68. Follain Jr., Renaud B. and Lim, G.C., "Economic Forces Underlying Urban Decentralization Trends: A Structural Model of Density Gradients Applied to Korea", Environment and Planning A. 11, 541-51 (1979). Glicknan, Norman, The Growth and Management of the Japanese Urban System, New York: Academic Press (1979). , and Oguri, Y., "Modeling the Urban Land Market: The Case of Japan", Journal of Urban Economics, 5, 505-525 (1978). Greene, David and John Barnbrock, "A Note on Problems in Estimating Exponential Urban Density Models," Journal of Urban Economics 5 (1978), 285-290. Guest, Avery M., "Population Suburbanization in American Metropolitan Areas 1940-1970," Geographical Analysis 7 (1975), 267-283. Hansen, W., "How Accessibility Shapes Land Use", Journal of American Institute Planners, 25, 73-76 (1959). Harrison, D., and Kain J.F., "Cumulative Urban Growth and Urban Density Functions", Journal of Urban Economics Vol. 1, 67-98 (1974). Hoyt, T., "One Hundred Years of Land Values in Chicago", University of Chicago Press, Chicago (1933). -70- Ingram, Gregory and Alan Carroll, "The Spatial Structure of Latin American Cities," Journal of Urban Economics, (forthcoming). (1980). Kau, J.B., and Lee, C.F., "The Functional Form in Estimating the Density Gradient: An Empirical Investigation", Journal of the American Statistical Association, Vol. 71, 326-327 (1976). and C. Sirmans, "Urban Land Value Vunction and the Price Elasticity of Demand for Housing," Journal of Urban Economics, 6 112-129 (1979). Latham, Robert F. and Maurice H. Yeates, "Population Density Growth in Metropolitan Toronto," Geographical Analysis 2, 177-185 (1970). Mills, Edwin S., "The Value of Urban Land", in The Quality of the Urban Environment (Perloff and Wingo, eds.), Baltimore: Johns Hopkins University Press (1971). , and Tan, J.P., "A Comparison of Urban Population Density Function in Developed and Developing Countries", paper presented at the Annual Meeting of the American Economic Association, Chicago, Ill, August 1978. , Urban Economics. Glenview: Scott, Foresman & Co., (1972). , Studies in the Structure of the Urban Economy. Baltimore: Johns Hopkins Press for Resources for the Future (1972). , and Byung Nak Song, Urbanization and Urban Problems. Cambridge: Harvard University Press (1979). Mohan, Rakesh, "Population, Income, and Employment in a Developing Metropolis: A Spatial Analysis of Bogota, Colombia," Washington: World Bank, City Study Project Paper No. 6 (1979). Muth, Richard, Cities and Housing. Chicago: University of Chicago Press, (1969). Newling, B.F., "Urban Growth and Spatial Structure: Mathematical Models and Empirical Structure," Geographical Review 56, 213-225 (1966). ,, "Urban Population Densities and Intraurban Growth," Geographical Review 54, 440-442 (1964). , "The Spatial Variation of Urban Population Densities," Geographical Review 59, 242-252 (1969). Papageorgiou, G.J., "A Theoretical Evaluation of the Existing Population Density Gradient Functions," Economic Geography 47, 21-26 (1971). Pines, D., and Weiss, Y., "Land Improvement and Land Values", Journal of Urban Economics 1, 1-13 (1976). 71- Solow, R.M., and W. S. Vickrey, "Land Use in a Long Narrow City", Journal of Economic Theory 3, 340-47 (1971). Wheaton, W.C., "A Comparative Static Analysis of Urban Spatial Structure", Journal of Economic Theory 9, 223-37 (1974), Oct. Wiesner, Guillermo, "Cien Anos de Historia de los Precios de la Tierra en Bogota: 1879-1978". Bogota: Corporacion Centro Regional de Poblacion. CCRP City Study Paper No. 3 (1980). Wilson, A.G., "Entropy in Urban and Regional Modelling", Pion Limited, London (1970).