1# RECORD 416244dcac790a50e1100858c6071c54dd1650ba2# TITLE: Analysing location attributes with a hedonic model for apartment prices in Donetsk, Ukraine3# YEAR: 2007.0 VENUE: nan DOI: 10.1080/1648715x.2007.96375674# AUTHORS: Marko Kryvobokov; Mats Wilhelmsson5# PAGES: 22 CANDIDATE TABLE PAGES (1-based): [10, 13, 21, 22]6# ABSTRACT: A hedonic model is specified for asking prices for apartments in Donetsk (Ukraine). This model is used to determine statistically significant location attributes. These attributes can be used for land assessment in a city where data on the land market are lacking. Distance gradients for CBD accessibility are investigated in different geographical directions. Separate models are created for sub‐samples located inside and outside the city centre. A spatial weight matrix is used to detect spatial autocorrelation. The regression results are compared with the valuation of experts.789===== PAGE 1 (score 5, numeric-paren 0, table-words 0, var-words 5) =====1011International Journal of Strategic Property Management12ISSN 1648-715X print / ISSN 1648-9179 online © 2007 Vilnius Gediminas Technical University13http://www.ijspm.vgtu.lt14International Journal of Strategic Property Management (2007) 11, 157–17815ANALYSING LOCATION ATTRIBUTES WITH A HEDONIC16MODEL FOR APARTMENT PRICES IN DONETSK,17UKRAINE18Marko KRYVOBOKOV 1 and Mats WILHELMSSON 2191 Royal Institute of Technology, Section of Real Estate Planning and Land Law, Stockholm,20Sweden and Donetsk National Technical University, Section of Geoinformatics and Geodesy,21Artyoma Str. 58, 83000 Donetsk, Ukraine22E-mail: k_mark@ukr.net232 Royal Institute of Technology, Section of Building and Real Estate Economics, Brinellvägen 1,24S-100 44 Stockholm, Sweden25E-mail: matsw@infra.kth.se26Received 13 December 2006; accepted 22 June 200727ABSTRACT. A hedonic model is specified for asking prices for apartments in Donetsk (Ukraine).28This model is used to determine statistically significant location attributes. These attributes29can be used for land assessment in a city where data on the land market are lacking. Distance30gradients for CBD accessibility are investigated in different geographical directions. Separate31models are created for sub-samples located inside and outside the city centre. A spatial weight32matrix is used to detect spatial autocorrelation. The regression results are compared with33the valuation of experts.34KEYWORDS: Location attributes; Land assessment; Apartment prices; Regression; Donetsk351. INTRODUCTION36In Ukraine, land, whether urban or rural,37is taxed, while buildings and structures are38not. However, in most Ukrainian cities the39market for land is underdeveloped. There are40few land sales, and many of them involve land41sold by the city to the private sector. For ex-42ample, according to Thomas (2003), there is43effectively no secondary market for land in44Kiev. Therefore, the assessed value of urban45land in Ukraine is not based on market prin-46ciples (Thomas, 2003). Instead, according to47“Pattern of technical documentation of mon-48etary valuation of land” (1998), assessed value49depends on a weighted average of several lo-50cation attributes. Valuers, land managers, and51urban planners choose these location attributes52and weights of their relative importance based53on subjective judgements.54The process of property assessment can be55recognised as “quite subjective” even in coun-56tries with a long history of property taxation57(Cornia and Slide, 2005). The problem of sub-58jectivity in land assessment in Ukraine is de-59scribed in detail in Kryvobokov (2004). To in-60crease the degree of objectivity in the specifi-61cation of a model, he proposes that a list of62the most important location attributes that63influence values in well-developed foreign real646566===== PAGE 10 (score 17, numeric-paren 6, table-words 4, var-words 3) =====676816669M. Kryvobokov and M. Wilhelmsson70apartment attributes and all location at-71tributes, denoted LIN (the linear model) and72LOG (the log-linear model) in Table 3. Model73LOG has higher adjusted R2 than model LIN74(77 percent compared to 69 percent). In both75models, Build, Area, Cond6, Cond7, DCBD,76and Prest are statistically significant. Cond577is significant in model LIN, whereas FloorT78and Phone are significant in model LOG.79Where the location attributes are concerned,80DWater and DNuis are significant only in81model LIN, whereas DSCent and DRail are sig-82nificant only in model LOG. Among significant83location attributes measured in kilometers in84model LIN, the highest coefficients are for85DNuis and DWater; the lowest is that for86DCBD. At the same time the highest of the87significant location coefficients in model LOG88is that for DCBD.89To investigate the existence of multicolli-90nearity, we estimated the maximum of vari-91ance inflationary factors (VIF) (Table 3). The92principle that VIF in excess of 10 (or tolerance93lower than 0.10) indicates multicollinearity is94usually used in the literature, see e.g. Seiler95et al. (2001) and Des Rosiers et al. (2001). For96models LIN and LOG, the magnitudes of VIF97are lower than the threshold that indicates no98problem with multicollinearity.99Importantly, the values of the Moran’s I sta-100Table 3. Estimated OLS results for the whole data set101le102d103o104m105d106n107a108s109elb110air111a112V113s114citsir115e116tc117a118r119a120h121c122le123d124o125M126N127I128L129G130O131L132R133I134D135N136I137L138R139I140D141G142O143L144t145n146a147ts148n149o150C15151525.0153615401553156)15731589.0159(16071619.7162*163)16431651.21662167(16851699.6170217121725173)17451755.1176(177*178)17981800.21812182(18311849.71851186r187o188ol189F19021913.4192419341941195-196)19791986.1199-(20022010.0202-203)20402057.0206-(20772083.8209921042111212-213)21402158.1216-(217)21822196.0220-(22122220.0223-224T225r226o227ol228F22902303.0231723252331234-235)23672377.1238-(23962400.0241-242*243)24472452.2246-(24702486.1249525082511252-253*254)25552561.2257-(258*259)26032611.2262-(26362640.0265-266dli267u268B26942700.4271827222733274*275)27652777.4278(27952801.0281*282)28332845.6285(28602875.5288128912903291*292)29342946.4295(296*297)29882994.6300(30153021.0303e304n305o306h307P30863099.1310131133121313)31453155.1316(31783180.0319*320)32173228.2323(32473259.8326932703281329)33053313.1332(333*334)33513366.2337(33873390.0340a341er342A34343444.834573464347*348)34993505.7351(35253536.0354*355)35643572.7358(35903605.136163624363*364)36543662.7367(368*369)37063714.7372(37373746.0375cla376B37733785.837923805381-382)38323845.0385-(38613870.0388-389)39043913.0392-(39383946.539513969397-398)39924009.0401-(402)40344041.0405-(40674070.0408-4092410d411n412o413C41464150.2416141734181419)42074211.1422(42324240.0425)42604276.0428(42924301.4431143264331434)43564364.1437(438)43944406.0441(44224430.04443445d446n447o448C44984506.7451645294531454)45534564.1457(45824590.0460)46124624.0463(46464652.0466946794681469)47034714.1472(473)47434753.0476(47724780.04794480d481n482o483C48404857.6486248704882489)49054917.1492(49344940.0495)49614971.1498(49905005.4501050225032504)50545069.1507(508)50955102.1511(51245130.05145515d516n517o518C51935205.6521652225232524*525)52605271.2528(52955300.0531)53285335.1534(53555363.1537253835392540*541)54275431.2544(545)54685477.1548(54965500.05516552d553n554o555C55695571.0558055965604561*562)56365647.2565(56635671.0568*569)57005715.2572(57355747.5575157665774578*579)58095817.2582(583*584)58555867.2587(58845891.05907591d592n593o594C59585964.3597159875999600*601)60286031.6604(60506063.0607*608)60986100.6611(61286133.1614061546169617*618)61996207.5621(622*623)62466253.6626(62716283.0629)630d631e632u633nit634n635o636c637(638639640===== PAGE 11 (score 5, numeric-paren 1, table-words 1, var-words 2) =====641642167643Analysing Location Attributes with a Hedonic Model for Apartment Prices in Donetsk, Ukraine644tistic are quite large; this suggests that nei-645ther model captures the underlying spatial re-646lationship very well. As an alternative, we con-647sider spatial models to incorporate the spatial648structure into the analysis (Anselin, 1988 is649the standard reference documenting these650models). We have estimated two different kinds651of spatial regression models to correct for the652underlying spatial relationship. These are a653spatial autoregressive model and a spatial er-654ror model. The results for the spatial models655are presented in the Appendix. Even though656the coefficient for spatial dependency is highly657significant, none of the estimates for the loca-658Note: White heteroskedastic consistent estimates. t-statistics are in parentheses below the coefficients. The asterisk659indicates significance at the 5% level.660le661d662o663m664d665n666a667s668elb669air670a671V672s673citsir674e675tc676a677r678a679h680c681le682d683o684M685N686I687L688G689O690L691R692I693D694N695I696L697R698I699D700G701O702L703)704d705e706u707nit708n709o710c711(712D713B714C715D71617172.871827198720-721*722)72327247.6725-(72647272.0728-729*730)73147322.07331734-(73577367.5737373897391740-741*742)74367446.7745-(746*747)74827498.6750-(75137522.0753-754t755n756e757C758S759D76027616.77624763-764)76537662.0767-(76857690.0770-771*772)77317749.4775-(77677776.677897791780-781)78277838.0784-(785*786)78777884.3789-(79047910.0792-793p794ot795S796D79707985.4799680088012802-803)80468053.1806-(80718080.0809-810)81148123.0813-(81448159.5816781798185819-820*821)82208239.2824-(825)82668272.0828-(82908300.0831-832lia833R834D83528364.78373838-839)84018414.0842-(84318441.0845-846*847)84848494.5850-(85138525.385308543855*856)85708583.2859(860*861)86258632.4864-(86518661.0867-868reta869W870D87108726.2873087478751876-877*878)87998807.3881-(88218830.0884-885)88678873.0888-(88928908.989198922893-894)89588965.0897-(898)89939009.0901-(90219030.0904-905si906u907N908D90959103.6911191289132914*915)91609178.3918(91939200.0921)92249235.1924(92549269.4927892819293930*931)93289331.4934(935)93699372.1938(93939400.0941ts942er943P94479455.1946694779484949*950)95179529.5953(95499550.0956*957)95849597.2960(96189621.3963596449652966*967)96879697.2970(971*972)97319740.2975(97689770.0978D979B980C981D982·983N984-985-98679874.798839897990*991)99219939.2994(995)99659973.0998(999110000.01001D1002B1003C1004D1005·1006S1007-1008-1009610107.1101161012-1013)1014210155.01016-(1017)1018610195.11020-(1021310220.01023-1024D1025B1026C1027D1028·1029W1030-1031-1032210336.610344103571036*1037)1038110390.51040(1041)1042410437.11044-(1045410460.01047-1048s1049n1050oit1051a1052v1053r1054e1055s1056b1057O105851059210603106151062210633106451065210663106751068210693107021071R1072d1073e1074ts1075u1076j1077d1078A1079910806.01081710827.01083110847.01085710867.01087e1088ula1089v-1090F1091610927.9109331094710954.7109651097610987.7109931100011012.0110251103x1104a1105M1106F1107I1108V11096.41110(1111D1112B1113C1114D1115)11167.31117(1118D1119B1120C1121D1122)11232.0112421125(1126D1127B1128C1129D1130)1131)1132D1133B1134C1135D1136(11371.61138I1139s'n1140a1141r1142o1143M1144611453.71146011476.31148911494.61150411517.31152tion variables change dramatically. None of1153them are significantly different from the re-1154sults of the OLS model. Hence, we use the OLS1155estimates in our econometric analysis.1156As in Söderberg and Janssen (2001), we1157analyse whether the CBD accessibility affects1158the asking price differently in different geo-1159graphical directions. We divide Donetsk into1160four parts, North, East, South, and West, with1161the CBD as the origin, to carry this out. Con-1162sidering that the area to the east of the CBD1163is the most prestigious area outside the city1164centre, area East is used as the default area.1165So, we introduce three dummies for location116611671168===== PAGE 13 (score 13, numeric-paren 11, table-words 1, var-words 0) =====116911701691171Analysing Location Attributes with a Hedonic Model for Apartment Prices in Donetsk, Ukraine1172le1173d1174o1175m1176d1177n1178a1179s1180elb1181air1182a1183V1184s1185citsir1186e1187tc1188a1189r1190a1191h1192c1193le1194d1195o1196M1197N1198I1199L1200G1201O1202L1203R1204I1205D1206N1207I1208L1209R1210I1211D1212G1213O1214L1215t1216n1217a1218ts1219n1220o1221C1222912233.9122441225712261122761228*1229)1230112312.21232(1233512340.91235*1236)1237712380.71239(1240912415.4124211243512440124511246*1247)1248012495.31250(1251812528.81253*1254)1255512566.8125711258(125911260r1261o1262ol1263F1264412656.91266312677126871269-1270*1271)1272812732.21274-(1275012762.01277-1278*1279)1280412812.21282-(1283012841.012854128621287-1288)1289312903.01291-(1292112930.01294-1295)1296212972.01298-(1299T1300r1301o1302ol1303F1304413053.41306313079130821309-1310)1311813124.11313-(1314013151.01316-1317)1318413198.11320-(1321013223.713238132491325-1326)1327813283.11329-(1330513310.01332-1333)1334013356.11336-(1337dli1338u1339B1340313415.71342913438134461345*1346)1347513486.21349(1350613511.01352*1353)1354213552.21356(1357213588.11359413607136131362*1363)1364713657.51366(1367013682.01369*1370)1371113727.71373(1374e1375n1376o1377h1378P1379513801.613818138211383-1384)1385613860.01387-(1388413890.01390)1391013925.01393(1394913958.813969139791398)1399014005.11401(1402614030.01404*1405)1406814071.21408(1409a1410er1411A1412214130.614146141551416*1417)1418014198.41420(1421614228.01423*1424)1425114268.41427(1428314291.314306143131432*1433)1434714350.61436(1437314384.01439*1440)1441314425.31443(1444cla1445B1446514478.71448814492145021451-1452)1453814540.11455-(1456514570.01458-1459)1460014618.01462-(1463014641.714659146651467-1468)1469614708.01471-(1472014730.01474-1475)1476814770.01478-(147921480d1481n1482o1483C1484614855.514860148721488-1489)1490614910.01492-(1493114940.01495-1496)1497114981.01499-(1500415014.815022150381504)1505715067.01507(1508215090.01510)1511615124.01513(151431515d1516n1517o1518C1519315202.21521915227152311524)1525815263.01527(1528415290.01530)1531315323.01533(1534315353.91536415373153811539)1540215419.01542(1543215440.01545)1546115473.01548(154941550d1551n1552o1553C1554715555.91556315572155811559)1560815613.01562(1563115640.01565)1566715670.01568(1569915708.315717157241573)1574615754.01576(1577415780.01579)1580415810.11582(158351584d1585n1586o1587C1588515897.115905159171592-1593)1594315952.01596-(1597415980.01599-1600)1601316024.01603-(1604216057.91606916078160811609)1610016119.11612(1613916140.01615*1616)1617116182.21619(162061621d1622n1623o1624C1625816267.51627816288162991630*1631)1632816332.21634(1635216362.01637)1638516398.11640(1641416427.31643016446164521646*1647)1648116490.21650(1651216521.01653*1654)1655516562.21657(165871659d1660n1661o1662C1663416645.71665416661166771668)1669416704.11671(1672516731.01674)1675716761.11677(1678116792.91680116812168291683*1684)1685216864.61687(1688316893.01690*1691)1692916937.51694(1695D1696B1697C1698D1699517005.01701617022170361704-1705*1706)1707517085.21709-(1710317111.01712-1713*1714)1715617160.21717-(1718917193.217201172161722-1723*1724)1725417266.51727-(1728117293.01730-1731*1732)1733517346.0173511736-(1737t1738n1739e1740C1741S1742D1743217441.61745317461174711748-1749)1750817512.01752-(1753217541.01755-1756)1757417584.01759-(1760017618.417629176371764-1765*1766)1767517687.31769-(1770417710.01772-1773*1774)1775617764.31777-(1778p1779ot1780S1781D1782917838.31784217855178631787)1788017894.01790(1791117920.01793)1794917950.01796(1797317984.71799518007180121802)1803818045.11805(1806118070.01808)1809418105.01811(1812lia1813R1814D1815618165.51817818181181951820-1821)1822418238.11824-(1825318268.01827-1828)1829818304.11831-(1832118338.718341183521836-1837*1838)1839318404.21841-(1842818430.01844-1845*1846)1847518489.31849-(1850reta1851W1852D1853418546.71855118560185781858)1859018600.11861(1862718630.01864)1865318669.01867(1868018692.21870418714187211873-1874*1875)1876818772.31878-(1879418800.01881-1882*1883)1884318850.21886-(1887si1888u1889N1890D1891618921.91893318948189541896-1897)1898318998.01900-(1901219021.01903-1904)1905419065.01907-(1908119095.419102191171912)1913919149.01915(1916319170.01918)1919219207.11921(1922ts1923er1924P1925-1926-1927519281.21929819302193121932*1933)1934119350.31936(1937619380.01939*1940)1941219420.21943(1944s1945n1946oit1947a1948v1949r1950e1951s1952b1953O195451955719565195771958019595196021961019625196321964R1965d1966e1967ts1968u1969j1970d1971A197221973719744.01975719764.01977919786.01979319807.01981e1982ula1983v-1984F1985319866.41987019887.41989419907.0199131992019938.6199431995F1996I1997V1998x1999a2000M20019.82002(2003lia2004R2005D2006)20070.92008(2009lia2010R2011D2012)20130.42014(2015lia2016R2017D2018)20194.32020(202152022d2023n2024o2025C2026)2027Table 4. Estimated OLS results for split data sets2028Note: In models LIN_Cent and LOG_Cent, Prest is equal to 1 for all apartments.202920302031===== PAGE 14 (score 24, numeric-paren 0, table-words 10, var-words 4) =====203220331702034M. Kryvobokov and M. Wilhelmsson2035respectively, both are larger than the critical2036value of 1.57 at the 5% significance level.2037WSE for sub-samples in total is calculated2038using the formula2039∑2040∑2041=2042=2043−2044−2045⋅2046−2047−2048=2049m2050i2051i2052i2053m2054i2055i2056i2057i2058k2059n2060SE2061k2062n2063WSE20641206512066)12067(2068)2069)12070((2071,2072where: n is the number of observations; k is2073the number of explanatory variables; i is the2074count for sub-samples; m is the number of sub-2075samples; SEi is the standard error in the i-th2076sub-sample model.2077The results of WSE estimations are pre-2078sented and compared with the initial models2079in Table 5. For linear specifications, the reduc-2080tion in the standard error is 17 percent,2081whereas for log-linear specifications it is only20828 percent. According to literature review by2083Watkins (2001), either 5 percent or 10 percent2084cut off is employed without a strict guidance2085on the size of the reduction. In our case, the2086sub-market specifications in a linear form sat-2087isfy both thresholds, while the log-linear mod-2088els satisfy only the former one.2089Thus, Chow-test and WSE estimation indi-2090cate that two delineated sub-markets are dif-2091ferent and cannot be described by one hedonic2092model. There are some interesting findings2093when it comes to the location variables. DCBD2094is the only significant location variable in mod-2095els LIN_Cent and LOG_Cent and the coefficient2096for DSCent is smaller than that for DCBD. For2097the linear model the former is 5.5 times smaller2098than the latter. In models LIN_Out and2099LOG_Out, all the location variables except2100DStop and DNuis are significant. LIN_Out is2101the only linear model where DSCent is signifi-2102cant. In this model, the coefficient for DSCent2103is higher than that for DCBD; the coefficient2104for DWater is 2.4 times higher than that for2105DCBD. The remarkable point is the compari-2106son of the coefficients for DCBD in LIN_Cent2107and LIN_Out. For central locations, it is more2108than 10 times higher than that for locations2109outside the centre. At the same time, in model2110LOG_Out the coefficient for DCBD is the high-2111est among coefficients for the location vari-2112ables. For central locations, the insignificant2113variables DWater and DNuis have unexpected2114signs. This might be explained by the high den-2115sity of buildings, which are barriers for the2116influence of water, green areas and nuisance2117objects.2118Though there are dissimilarities between2119the linear and log-linear models concerning the2120importance of secondary centres, it seems that2121these centres in Donetsk are important. There-2122fore, Donetsk can be considered as a non-2123monocentric city. The difference between non-2124monocentric and polycentric urban models is2125described e.g. in Sivitanidou (1997). Here we2126do not investigate this difference, but conclude2127that a monocentric model is not appropriate2128for Donetsk.2129Table 5. Weighted standard error tests2130le2131d2132o2133M2134r2135o2136rr2137e2138d2139r2140a2141d2142n2143a2144t2145S2146n2147oitc2148u2149d2150e2151r2152t2153n2154e2155c2156r2157e2158P2159:s2160n2161oit2162a2163cific2164e2165p2166s2167r2168a2169e2170ni2171L2172N2173I2174L217592176721774217852179t2180u2181O2182_2183N2184I2185L2186d2187n2188a2189t2190n2191e2192C2193_2194N2195I2196L2197821983219952200422017220212203:s2204n2205oit2206a2207cific2208e2209p2210s2211r2212a2213e2214nil-2215g2216o2217L2218G2219O2220L222142222022239222452225722261.02227t2228u2229O2230_2231G2232O2233L2234d2235n2236a2237t2238n2239e2240C2241_2242G2243O2244L224532246322473224822249622501.022518225222532254===== PAGE 21 (score 24, numeric-paren 24, table-words 0, var-words 0) =====225522561772257Analysing Location Attributes with a Hedonic Model for Apartment Prices in Donetsk, Ukraine2258APPENDIX2259Spatial Autoregressive Models (SAR) and Spatial Error Models (SEM)2260d2261n2262a2263s2264elb2265air2266a2267V2268le2269d2270o2271m2272s2273citsir2274e2275tc2276a2277r2278a2279h2280c2281le2282d2283o2284M2285G2286O2287L2288R2289I2290D2291G2292O2293L2294R2295A2296S2297M2298E2299S2300R2301A2302S2303M2304E2305S2306t2307n2308a2309ts2310n2311o2312C2313723149.82315*2316)2317423180.92319(2320523211.82322*2323)2324623257.1232622327(2328223291.62330*2331)2332823331.92334(2335923360.82337*2338)2339423407.1234122342(234312344r2345o2346ol2347F2348223490.02350-2351)2352223538.02354-(2355323560.02357-2358)2359923609.02361-(2362223630.02364-2365)2366723677.02368-(2369223700.02371-2372)2373023749.02375-(2376T2377r2378o2379ol2380F2381623820.02383-2384*2385)2386123875.22388-(2389623900.02391-2392*2393)2394323953.22396-(2397623980.02399-2400*2401)2402124034.22404-(2405524060.02407-2408*2409)2410424112.22412-(2413dli2414u2415B2416424171.02418*2419)2420124211.62422(2423624241.02425*2426)2427924285.62429(2430424311.02432*2433)2434424351.62436(2437624381.02439*2440)2441824425.62443(2444e2445n2446o2447h2448P2449724500.02451*2452)2453724547.22455(2456624570.02458*2459)2460424615.22462(2463724640.02465*2466)2467724685.22469(2470624710.02472*2473)2474324753.22476(2477a2478er2479A2480024816.02482*2483)2484024854.62486(2487024886.02489*2490)2491524922.62493(2494224956.02496*2497)2498624995.62500(2501125026.02503*2504)2505425064.62507(2508cla2509B2510125110.02512-2513)2514125153.02516-(2517125180.02519-2520)2521825220.02523-(2524125250.02526-2527)2528425291.02530-(2531125320.02533)2534125351.02536(253722538d2539n2540o2541C2542125430.02544)2545425463.02547(2548125490.02550)2551625522.02553(2554125550.02556)2557825583.02559(2560125610.02562)2563025643.02565(256632567d2568n2569o2570C2571225720.02573)2574925752.02576(2577125780.02579)2580125812.02582(2583125840.02585)2586525872.02588(2589125900.02591)2592625931.02594(259542596d2597n2598o2599C2600326010.02602)2603226048.02605(2606326070.02608)2609126107.02611(2612326130.02614)2615126169.02617(2618326190.02620)2621026228.02623(262452625d2626n2627o2628C2629526300.02631)2632126333.12634(2635526360.02637)2638326393.12640(2641526420.02643)2644426454.12646(2647526480.02649)2650626514.12652(265362654d2655n2656o2657C2658226591.02660*2661)2662626635.22664(2665326661.02667*2668)2669526706.22671(2672326731.02674*2675)2676026778.22678(2679426801.02681*2682)2683826848.22685(268672687d2688n2689o2690C2691026923.02693*2694)2695226969.52697(2698026993.02700*2701)2702327038.52704(2705127063.02707*2708)2709127100.62711(2712027133.02714*2715)2716427179.52718(2719D2720B2721C2722D2723927241.02725-2726*2727)2728127291.92730-(2731327322.02733-2734*2735)2736827371.1273812739-(2740827411.02742-2743*2744)2745527462.72747-(2748127492.02750-2751*2752)2753227547.72755-(2756t2757n2758e2759C2760S2761D2762527630.02764-2765*2766)2767627682.42769-(2770527710.02772-2773*2774)2775427760.42777-(2778427790.02780-2781*2782)2783627842.32785-(2786427870.02788-2789*2790)2791427928.22793-(2794p2795ot2796S2797D2798127990.02800-2801)2802228033.02804-(2805128060.02807-2808)2809228106.02811-(2812128130.02814-2815)2816228172.02818-(2819128200.02821-2822)2823228245.02825-(2826lia2827R2828D2829828300.02831-2832*2833)2834228353.42836-(2837128381.02839-2840*2841)2842628438.42844-(2845828460.02847-2848*2849)2850628515.32852-(2853028541.02855-2856*2857)2858128590.42860-(2861reta2862W2863D2864128650.02866-2867)2868628693.02870-(2871128720.02873-2874)2875828766.02877-(2878128790.02880-2881)2882728838.02884-(2885228860.02887-2888)2889728902.12891-(2892)2893d2894e2895u2896nit2897n2898o2899c2900(290129022903===== PAGE 22 (score 10, numeric-paren 8, table-words 1, var-words 0) =====290429051782906M. Kryvobokov and M. Wilhelmsson2907d2908n2909a2910s2911elb2912air2913a2914V2915le2916d2917o2918m2919s2920citsir2921e2922tc2923a2924r2925a2926h2927c2928le2929d2930o2931M2932G2933O2934L2935R2936I2937D2938G2939O2940L2941R2942A2943S2944M2945E2946S2947R2948A2949S2950M2951E2952S2953)2954d2955e2956u2957nit2958n2959o2960c2961(2962si2963u2964N2965D2966229670.02968)2969629701.12971(2972229730.02974)2975829761.12977(2978229790.02980)2981029820.12983(2984229850.02986)2987129889.02989(2990ts2991er2992P2993929940.02995*2996)2997729988.22999(3000030011.03002*3003)3004230058.23006(3007730080.03009*3010)3011730121.23013(3014830150.03016*3017)3018930190.23020(3021D3022B3023C3024D3025·3026N3027-3028-3029130300.03031)3032530333.03034(3035130360.03037)3038230391.03040(3041D3042B3043C3044D3045·3046S3047-3048-3049230500.03051-3052)3053030540.13055-(3056430570.03058-3059)3060530615.13062-(3063D3064B3065C3066D3067·3068W3069-3070-3071330720.03073-3074)3075030766.13077-(3078430790.03080-3081*3082)3083730840.23085-(3086ρ3087130882.03089*3090)3091430927.33093(3094-3095930961.03097*3098)3099131003.33101(3102-3103λ3104-3105231062.03107*3108)3109231103.33111(3112-3113131142.03115*3116)3117831181.33119(3120s3121n3122oit3123a3124v3125r3126e3127s3128b3129O3130531312313233133531342313533136531372313833139531402314133142d3143e3144ts3145u3146j3147d3148A3149R23150731517.03152831537.03154731557.03156831577.03158