ÿWPC y Lþ[«~¬.}:g³­«ÈÈŽÛÔʲ…Ù­«c6¹ÌêãBÆûžÆ«'/˜]o³:½'+ao‘ÂõŽ+ÂwÏQ±7hŒìŸVVo$½¿\üIG. ê´P®’¾èpU’|X(ªúÒ‚?ä)F9,†fa€œT8ìó¢þnSK5“Ÿ;Ûîÿî‘Bh[ã¯TfBD?9¼ØÀìbKº;Ìù±AÀ¿Ä½­½V|C©†!ˆ‚-“}˜6=e¥J²¤ßúg[q†@Q9UëDûÓ÷ó›ì ôÛù§";+µÃP!êØŸKø‚ñ&-Ãô›îJsU8û\Tª9P™šQA³ÄïU³ÚÉÔÌ–zAˆþxÒ÷Fxƒü ä¤1hÒ¼·¶“…•{6.dõFоÃH£4þ|Þ´û"_ÌÀ£a±h„7Ç”du÷W=%N *ë «§k1½uÔܦÔh«½i¿‚ØÒÛå¡‘²¶^Cµñ3¦¥–¯MÇVªÏ\9aF¯Ÿf êÁù«È– Y;„^0:àw<Ž™ç«©¡\:  4…„‹ðHê¢@°ƒ!ñYb}z?zmÀÇÀã¢/–™Ã¤ /L%P{¾‰ñ’¹™8} FÈW’U|.ü %* Ö0 m 0sU0 D+0Æ[ 0 ! 0=® 0ë AQxrÉ 0C;ö~ 0ÕtöI-?ölöböXöNöDö:ö0÷&÷÷ Æ ! 0EÑ! AY"dÆo"#5#X#Æp$ 0C6% AMy%d¸Æ%)~&÷§&dVž'mô*a+dqy,Xê-B.dŠZ/%ä/÷ 0÷1d÷1H 4èU4Æ=5d—6$š6d¾6FÌ7÷8÷ 9÷:÷÷:î;÷þ<÷õ=÷ì>÷ã?÷Ú@÷ÑA 0=ÈB AMCRCnbC 0rfGØGmØGU.EJU0sJr£J hL hL h4L h:Lw@@L4€L 0””L 0¾(M 0¾æM 0Á¤N 0ÁeO 0Á&P 0¾çP 0o¥Q 0¾R 0¾ÒR 0ÉS 0ÆYT 0ÓU 0wòU 0ŠiV 0vóVniW B5lYn®‰YU<7\fs\au\n‰\fŒ^aŽ^n¢^f¥`a§`n»`f¾baÀb 0xÔb 0„Lc 0xÐc 0xHd 0xÀd 0x8e 0f°e 0úf 0úg 0x h 0ú‚h 0ú|i 0úvj 0úpk 0újl 0údm 0 ^n 0pio 0tÙo 0ýMp 0ýJq 0|Gr 0|Ãr 0ý?s 0|£X»]i ÌP_{X_i}~€=~€price~€of~€X_i íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú14Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú15Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó XX~UŒ81XX4~äXX~U|81XX$~(XXô~Xl81XX¼~,XXŒ~q81XXT~,XX$~hXXÈ~(XX@~q¸81XX ~,XXØ ~qP 82XX  ~,XX. ~.XX¼ ~.XXJ ~.XXØ ~,XX¨ ~q%8NXXu~)XXí~,XX½~E581XX…~;XXU~XXÍ~) >U_1~=~U_1~(~X_1,~q_1,~h``(q_1,~q_2,`.`.`.`,~q_N),~E_1;~€ðð€) ÞÝ ƒ w #ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÝ  ÝÔ€X¨sXNN <ÔÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Óà€¿¿' àò ò(4)ó ó Level 1 Level 2 Level 3 Level 4 Level 5 XXÛhXXŒÛ(XXÛXX6Û)XXÛäXXÖ¨I8iXXÛq’•i h(cdot)~=~sum€from€i€``q_i XX P‘ÆXXX  XXk MRSÓÆEq,,y˜iXXá (Xw5XXÀ{,XX {hXX–~,XXö~qs8iXXâ )XXp MRSØÆQX <P_X``cdot``MRS_{Eq_i}``{({€ð,ð€h}over{€ð,ð€q_i})}`MRS_{QX} íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú16Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú17Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú18Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú19Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó Ð î$  ÐÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú20Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú21Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú22Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú23Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú24Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú25Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó íÝ ƒ×¶%ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÓÓÔ€X¨sXNN <ÔÚ  Ú26Ú  Ú.Ý  ÝÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?Ó(ÿÿ$’’Ú  ÚÚ  Ú6 FA<< cÿÿ#|x«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@<þ6X9`(CourierNìvþ6X@ɹ`7ûN@*Ͼ¦ `CG TimesôôdϾ¦ PìE37ûôP*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXP*Ͼ¦ `CG TimesPϾ¦ PìE37ûP<þ6X9`(Courierê˜bþ6X@ɹ`7ûê@(JÓ@$¡¡Ô€N <NXXÔÔ€N <NNN <ÔÓ  Ó cÝ ƒw 'ÝÔ€MN <NXXÔÔ€MN <NNN <ÔÓ  ÓÝ  ÝÓÓÓ  ÓÓE1ÿÿX°` ¸ hÀpÈ xÐ (#°œXEÓÔ€Mô`îôNN <ÔÔ€Nô ôôô`îÔÔ  ÔÔ??US.,ÔÒÝ °ÒÒ\°Òß…€ OO?;z ,Ý ŸÈ 0€ €@ H@EM °ŸÈŸéM °ŸÈ…ßÒ°Ý ÒÒ°\ÒÔ€NX-³Xôô ÔÔUS??.,ÔÌÓÓÔ€NNô‰NXX-³ÔÔ€MN <NNNô‰Ô<þ6X9`(Courier*Ͼ¦ `CG Times hÝ ƒ w "ÝÔ€MN <NXXÔÔ€MN <NNN <ÔÓ  ÓÝ  ÝÔ€Mô`îôNN <ÔÔ€Nô ôôô`îÔÓE1ÿÿX°` ¸ hÀpÈ xÐ (#°œXEÓÓ  ÓÓ! ÿÿ°’"(#ÿÿXŽX!ÓÔ€NX-³Xôô Ôà@OOOàˆÐ  Ðà ° àjmmCps UîUîVîEE7’$ hÿÿ:Default ParaDefault Paragraph FontÔ  ÔÔ  ÔÔ  ÔÔ  Ô8;Kÿÿ,toc 1toc 1Ó  ÓÓ°°ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8;Kÿÿ,toc 2toc 2Ó  ÓÓ°` ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8>Kÿÿ,toc 3toc 3Ó  ÓÓ°` ¸ ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8>Kÿÿ,toc 4toc 4Ó  ÓÓ°¸ ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8>Kÿÿ,toc 5toc 5Ó  ÓÓ°hü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8;Kÿÿ,toc 6toc 6Ó  ÓÓ°°ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8!ÿÿ,toc 7toc 7Ó  ÓÔ  ÔÔ  ÔÔ  ÔÔ  Ô8;Kÿÿ,toc 8toc 8Ó  ÓÓ°°ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó8;Kÿÿ,toc 9toc 9Ó  ÓÓ°°ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó@>Kÿÿ0index 1index 1Ó  ÓÓ°°` ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó@;Kÿÿ0index 2index 2Ó  ÓÓ°` ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP8Kÿÿ8toa headingtoa headingÓ  ÓÓ °°ü!(#ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó@!ÿÿ0captioncaptionÓ  ÓÔ  ÔÔ  ÔÔ  ÔÔ  Ô^ÿÿ:_Equation Ca_Equation CaptionÔ  ÔÔ  ÔÔ  ÔÔ  ÔJÿÿ$1_Equation Caption1Ô  ÔÔ  ÔÔ  ÔÔ  Ô«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@<þ6X9`(CourierNìvþ6X@ɹ`7ûN@*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXP d«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@<þ6X9`(CourierNìvþ6X@ɹ`7ûN@*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXP%ô\  `$Times NewRomanNNhô\  PŽ6G;NP%ô\  `$Times NewRomanTable_A«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@<þ6X9`(CourierNìvþ6X@ɹ`7ûN@*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXPTable_A«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@*Ͼ¦ `CG TimesôôdϾ¦ PìE37ûôP*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXPTable_A«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@*Ͼ¦ `CG TimesôôdϾ¦ PìE37ûôP*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXPTable_ALÿÿ6Document 8Document 8Ô  ÔÔ  ÔÔ  ÔÔ  ÔLÿÿ6Document 4Document 4Ô  ÔÔ  Ôòòò òóóó óÔ  ÔÔ  ÔLÿÿ6Document 6Document 6Ô  ÔÔ  ÔÔ  ÔÔ  ÔLÿÿ6Document 5Document 5Ô  ÔÔ  ÔÔ  ÔÔ  ÔLÿÿ6Document 2Document 2Ô  ÔÔ  ÔÔ  ÔÔ  ÔLÿÿ6Document 7Document 7Ô  ÔÔ  ÔÔ  ÔÔ  ÔPÿÿ8BibliogrphyBibliogrphyÔ  ÔÔ  ÔP_Kÿÿ8Right Par 1Right Par 1Ó  ÓÓ>4°X° ` ¸ hÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 2Right Par 2Ó  ÓÓ>4°X°ø` ¸ hÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓLÿÿ6Document 3Document 3Ô  ÔÔ  ÔÔ  ÔÔ  ÔP_Kÿÿ8Right Par 3Right Par 3Ó  ÓÓ>4°X°` P ¸ hÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 4Right Par 4Ó  ÓÓ>4°X°` ¸ ¨ hÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 5Right Par 5Ó  ÓÓ>4°X°` ¸ ˆhÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 6Right Par 6Ó  ÓÓ>4°X°` ¸ hàÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 7Right Par 7Ó  ÓÓ>4°X°` ¸ hÀ°pÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Kÿÿ8Right Par 8Right Par 8Ó  ÓÓ>4°X°` ¸ hÀpÈ xÐ (#>ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓLhWÿÿ6Document 1Document 1Ô&  ÔÓ  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓÔ'  ÔDÿÿ2Doc InitDoc InitÔ  ÔÔ  ÔÔ  ÔÔ  ÔHÿÿ4Tech InitTech InitÔ  ÔÔ  ÔÔ  ÔÔ  ÔP_Nÿÿ8Technical 5Technical 5Ó  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  Ôò òó óÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Nÿÿ8Technical 6Technical 6Ó  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  Ôò òó óÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓPÿÿ8Technical 2Technical 2Ô  ÔÔ  ÔÔ  ÔÔ  ÔPÿÿ8Technical 3Technical 3Ô  ÔÔ  ÔÔ  ÔÔ  ÔP_Nÿÿ8Technical 4Technical 4Ó  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  Ôò òó óÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓPÿÿ8Technical 1Technical 1Ô  ÔÔ  ÔÔ  ÔÔ  ÔP_Nÿÿ8Technical 7Technical 7Ó  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  Ôò òó óÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓP_Nÿÿ8Technical 8Technical 8Ó  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  Ôò òó óÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5ÓD\Kÿÿ2PleadingPleadingÓ  ÓÓ;1°X°` ¸ hÀpÈ xÐ (#;ÓÔ  ÔÔ  ÔÔ  ÔÔ  ÔÓ5+° ` ¸ hÀpÈ xÐ (#5Ó«<þ6X9`(CourierXôxþ6X@ɹ`7ûX@<þ6X9`(CourierNìvþ6X@ɹ`7ûN@*Ͼ¦ `CG TimesXXxϾ¦ PìE37ûXPÝ ƒJÓ@!ÝÔ€N <NXXÔÔ€N <NNN <ÔÓ  ÓÝ  ÝÑ7€ <NNdÄdÈ7ÑØØÓ?+ÿÿ ` ¸ hÀpÈ xÐ (#°œX?ÓññÖ€ÿÿÖññññÖ€LÿÿÖññÑ  ÑÓÓÓ  ÓÔ€X¨sXNN <ÔÔ€X-³XXX¨sÔÔ  Ôà  àà ` àà ¸ àà  àà h àà À àà  àà p àà È àÌà  àà ` àà ¸ àà  àà h àà À àà  àññà p àà È àññññCopyright€ññññññ„€ññññññ1995€„€ññRevised€10/11/95ÌÌÌà@¥¥ìàSocial€Benefits€of€Education:€€Feedback€Effects€and€Environmental€ResourcesˆÌÌà@OO#ìàV.€Kerry€Smith*ˆÌÌò òI.à  à€Introductionó óÐ @ ÐÌà  àRicher€countries€are€safer,€healthier€places€to€live.× ƒ ×Ý ƒÿÿÝòòÚ  Ú1Ú  ÚóóÝ  Ý×  ×€€They€pollute€less€and€enjoy€a€higherÐ æ6 Ðstandard€of€material€well„being.× ƒ ×Ý ƒÿÿÝòòÚ  Ú2Ú  ÚóóÝ  Ý×  ×€€Several€recent€statistical€studies€have€used€aggregate€cross„Ð ¸ Ðcountry€data€to€suggest€a€causal€link€between€the€level€of€economic€activity,€safety€(Lutter€andÏMorrall€[1994]),€and€the€environment€(Grossman€and€Krueger€[1991]).€€Is€education€the€òòrealóóÐ \¬ Ðsource€of€these€positive€results?€€The€answer€to€this€question€will€not€be€found€in€aggregateÏstatistical€models.€€Evaluating€the€indirect,€non„market€effects€of€education€requires€detailed,Ïmicro€economic€analyses€of€education's€causal€role€in€people's€behavior.€€While€there€isÏextensive€evidence€that€education€and€training€enhance€people's€productivity€and€earnings,€whatÏwe€know€about€educationð ðs€effects€outside€markets€is€much€more€limited.Ð v%Æ $ Ðà  àThe€purpose€of€this€paper€is€to€consider€how€education€might€influence€the€environmentalÏquality€people€experience.€€Education€could€promote€private€behavior€that€enhancesÏenvironmental€quality€for€everyone,€or€increase€people's€effectiveness€in€protecting€ØØthemselvesÏfrom€negative€environmental€effects.€€These€negativeррinfluences€result€from€conventionalÐ ¾,(,‡ ÐÑÑÑ7€-³XXdéNNdÄ7ÑÑ  ÑññÑÑÑ7€-³XXdéNNdÄ7ÑÑ  ÑññññÑÑÑ7€-³XXdéNNdÄ7ÑÑ  ÑññññÑÑÑ7€-³XXdéNNdÄ7ÑÑ  ÑÑÑÑ7€-³XXdéNNdÄ7ÑÑ  Ñññexternalities€or€from€unrecognized€impacts€of€the€environment€on€people€(e.g.€radon€and€lungÏcancer,€etc.).Ìà  àIn€order€to€attribute€an€indirect€social€benefit€to€one€of€these€behavioral€responses,€weÏmust€establish€that€it€resulted€from€a€causal€rather€than€a€taste€related€association€with€education.€ÏThere€is€òònoóó€direct€evidence€of€these€types€of€connections€in€the€environmental€economicsÐ VH Ðliterature.€€Economists€who€evaluate€people's€behavior€in€avoiding€environmental€externalities€orÏenhancing€its€quality€have€not€tested€these€hypotheses.€Ìà  àBecause€there€was€no€literature€on€education's€role€in€improving€environmental€quality,Ïthis€paper€considers€three€questions:€€(1)€how€should€we€define€the€social€or€non„market€benefitsÏof€education?€€(2)€do€existing€studies€of€people's€mitigating€behavior,€avoiding€environmentalÏhazards,€offer€any€indirect€evidence€of€a€causal€link€between€education€and€increased€activitiesÏto€improve€the€environment€people€experience?€and€(3)€assuming€there€might€be€a€causalÏrelationship€between€education€and€people's€non„market€behavior,€would€that€connection€haveÏimplications€for€the€design€and€evaluation€of€environmental€policy?€ÌÌÌÌÌÌ€ò òII.à  àDefining€Terms€and€Examining€TrendsÐ ¤*–"& ÐÐ |,n$( Ðà  àA.à ` àòòWhat€is€Education?ó óóóÐ  Ðà  àEducation€is€both€a€process€and€a€set€of€outcomes.€€For€the€issues€discussed€in€whatÏfollows,€I€am€interested€in€what€the€educational€process€conveys€to€the€individual.€€Today'sÏstudents€at€all€levels€are€involved€in€activities€that€enhance€environmental€quality€in€many€casesÏas€part€of€their€science€courses.€€While€these€programs€may€be€important€as€vehicles€to€teachÏenvironmental€stewardship,€there€is€insufficient€information€about€them€to€permit€any€evaluation.€ÏAs€a€result,€specific€environmental€education€is€not€a€focus€of€my€analysis.€€Rather,€the€primaryÏconcern€is€whether€the€outcomes€resulting€from€education€as€a€whole€alter€people's€behaviorÏtoward€the€environment€when€they€leave€school.€€Ìà  àBecause€the€outcomes€resulting€from€education€are€diverse€and€often€difficult€to€measureÏin€quantitative€terms,€most€studies€of€the€effects€of€education€have€focused€on€duration€andÏdegree€based€measures€„€years€of€schooling€and€completion€of€specific€stages€of€the€process€(i.e.,Ïhigh€school€diploma,€college€degree,€graduate€degree,€etc.).€€Refined€outcomes€such€as€cognitiveÏreasoning€skills,€mathematical€ability,€or€understanding€of€science€are€more€difficult€to€measureÏand€are€usually€not€available.€€This€is€potentially€important.€€If€we€hypothesize€that€educationÏaffects€people's€ability€to€protect€themselves€from€environmental€hazards€or€increases€theirÏunderstanding€of€the€environmental€system€and,€as€a€result,€their€concern€for€protecting€it,€thenÏwe€should€expect€to€see€specific€causal€links€to€what€takes€place€in€school.€€That€is,€it€seemsÏreasonable€to€expect€these€changes€could€be€related€to€specific€aspects€of€the€educationalÏprocess,€such€as€a€set€of€environmental€science€modules€in€the€primary,€secondary,€and€collegeÏcurriculum.€€Indeed,€it€is€a€belief€that€these€activities€promote€such€responses€that€has€increasedÐ x,j$( Ðtheir€presence€in€instructional€programs.× ƒ ×Ý ƒÿÿÝòòÚ  Ú3Ú  ÚóóÝ  Ý×  ×Ð  Ðà  àUnfortunately,€the€measures€available€do€not€offer€this€level€of€resolution.€€ThisÏlimitation€has€at€least€two€implications.€€Finding€a€correlation€between€a€general€measure€ofÏeducation€such€as€years€of€schooling€and€an€individual's€actions€to€improve€environmentalÏquality€is€more€difficult€to€accept€as€a€causal€relationship.€€Equally€important,€the€prospect€ofÏdetecting€these€correlations€is€small.€€Years€of€schooling€can€involve€many€different€mixes€ofÏactivities€and€patterns€of€learning.€€There€is€no€reason€to€believe€all€aspects€of€an€educationalÏprogram€will€be€equally€relevant€to€behavior€that€either€contributes€to€enhancing€environmentalÏquality€or€promotes€people's€understanding€of€how€to€protect€themselves€from€environmentalÏhazards.€€ÌÌà  àò òB.à ` àòòSocial€Versus€Public€Benefitsóóó óÐ  Ðà  àConventional€economic€discussions€of€the€motives€for€public€intervention€arise€fromÏsome€type€of€"market€failure."€€Externalities€and€public€goods€are€the€most€prominent€examplesÏof€these€failures.€€To€the€extent€a€private€activity€has€an€impact€on€the€production€process€orÏwell„being€of€others€€that€is€òòoutsideóó€any€market€determined€transaction,€we€consider€it€to€be€anÐ ^#P Ðexternality.€€When€an€individual€acquires€the€skills€and€understanding€provided€throughÏeducation,€these€may€enhance€the€others'€well„being€in€some€way€(e.g.,€each€citizen€makesÏ"better"€decisions€in€a€democratic€society).€€This€is€an€external€effect.Ìà  àTo€the€extent€residents€of€a€community€demand€and€experience€welfare€gains€when€theirÏschool€systems€(whether€primary€and€secondary,€or€their€state€universities)€are€enhanced,€thenÐ x,j$( Ðthese€gains€may€reflect€some€public€good€aspects€of€these€educational€programs.€€This€effect€is,Ïin€principle,€different€from€what€happens€to€each€individual€through€education.€€Of€course,€itÏmay€well€be€that€the€reason€for€these€"public"€values€arises€as€an€òòexóó€òòanteóó€expectation€that€theÐ ² ¤ Ðcitizens€emerging€from€enhanced€educational€programs€will€create€externalities€for€theÏcommunity€in€the€future€at€the€time€they€finish.€€Ìà  àNonetheless,€these€explanations€are€"stories"€an€analyst€offers€to€explain€a€set€of€choices.€ÏAs€I€argue€below,€they€do€not€necessarily€mean€people€have€these€values€unless€their€choicesÏwere,€in€fact,€motivated€by€them.Ìà  àWhy€do€we€make€a€distinction€between€private€and€public€benefits€from€education,€andÏhow€does€the€issue€of€social€benefits€enter€an€economic€evaluation?€€The€answers€to€theseÏquestions€relate€to€two€concerns.€€Do€the€conventional€private€returns€from€education,€measuredÏusing€education's€effects€on€productivity€and€earnings,€reflect€the€total€value€of€education€to€eachÏindividual?€€Here€the€answers€relate€to€the€role€of€the€social€effects.€€To€the€extent€educationÏenhances€a€person's€abilities€in€non„market€activities,€the€private€earnings€related€measuresÏwould€be€only€a€partial€gauge€of€the€economic€value€of€education€òòtoóó€òòthatóó€òòindividualóó.€€OmittingÐ Š!| Ðthese€effects€could€lead€the€analyst€to€incorrectly€interpret€comparisons€of€education's€costs€withÏits€benefits€measured€solely€from€earnings€effects€as€indicating€irrational€behavior.€€This€followsÏbecause€the€benefit€side€of€the€comparison€omits€the€non„market€or€social€influences.€ÏNonetheless,€this€type€of€comparison€does€not€in€itself€motivate€public€intervention€because€theÏgains€are€"captured"€by€the€individual.€The€only€issue€is€one€of€measurement€and,€perhaps,Ïtesting€the€hypotheses€implied€by€revealed€preference.Ð v,h$( Ðà  àExternalities€and€public€goods€do€motivate€intervention€because€private€action€aloneÏwould€not€necessarily€assure€an€efficient€level€of€the€activity€that€contributes€to€these€effects.€ÏBoth€the€non„market€private€and€the€public€good€effects€of€education€must€rely€on€methods€ofÏnon„market€valuation€to€evaluate€their€impact.€€Thus,€a€substantial€component€of€the€discussionÏin€this€paper€will€consider€whether€existing€analyses€of€the€demand€for€and€value€of€educationÏhave€adequately€accounted€for€these€potential€impacts.Ì€Ìà  àò òC.à ` àòòTrends€in€Support€for€the€Environmentóóó óР̾  Ðà  àOver€the€past€decade,€environmental€quality€in€the€U.S.€has€generally€improved.€€€ItÏwould€be€difficult€to€establish€whether€these€types€of€changes€were€in€any€way€linked€to€changesÏin€education.€€There€are€no€systematic€records€of€household€activities€related€to€promotingÏenvironmental€quality€or€sustaining€private€averting€behavior€over€time.€€The€only€consistentÏlong„term€information€about€attitudes€toward€the€environment€can€be€found€in€the€NORCÏGeneral€Social€Surveys.€€From€1972€to€1993,€19€national€samples€were€asked€the€same€questionsÏabout€support€for€environmental€and€other€programs.€€These€consist€of€independent€samples€ofÏEnglish€speaking€persons,€18€years€of€age€or€over,€living€in€non„institutional€arrangements€in€theÏU.S.Ìà  àFigure€1a€plots€the€results€for€four€of€a€total€of€six€questions€derived€in€each€of€these€19Ïsurveys.€€The€first€curve€includes€the€average€years€of€schooling€(relative€to€thirteen€to€representÏhigh€school€equivalent),€RAVED.€€The€remaining€curves€plot€the€fractions€indicating€that€we€areÏspending€too€little€money€on€improving€and€protecting€the€environment€(mean€env.€high);€tooÐ x,j$( Ðlittle€money€on€improving€the€nation's€education€system€(mean€educ.€high);€as€well€as€theÏfraction€indicating€that€they€had€"a€great€deal€"of€confidence€in€those€people€running€theÏExecutive€branch€of€the€federal€government€(mean€fed.€high).€€As€the€graph€indicates,€there€is€aÏslight€upward€trend€in€the€years€of€schooling€completed€over€the€21€years€covered€by€theseÏsamples.€€Support€for€education€is€also€consistent€with€a€somewhat€more€pronounced€upwardÏtrend.€€The€most€surprising€outcome€of€the€comparison€is€the€decline,€beginning€about€1991,€inÏthose€indicating€we€are€spending€"too€little"€on€the€environment.€€This€change€is€not€sufficient€toÏchange€a€positive€and€significant€trend€in€the€fraction€agreeing€that€too€little€is€being€spent€overÏthe€full€time€period,€but€it€is€a€departure€in€a€trend€that€appears€to€have€begun€about€1977.€€ThisÏresult€does€not€seem€to€be€related€to€any€change€in€the€educational€levels€of€those€surveyed,€andÏdoes€not€in€itself€indicate€people€would€be€less€willing€to€take€private€actions€to€enhance€theÏenvironment.€€[Figure€1a€here]Ìà  àIt€may€be€the€result€of€a€decline€in€confidence€in€the€Executive€branch€of€government€asÏwell€as€other€public€functions,€as€illustrated€by€the€lowest€line€in€Figure€1a,€displaying€thoseÏrespondents€expressing€"a€great€deal"€of€confidence€in€those€running€the€Executive€Branch€ofÏgovernment.€€Indeed,€a€similar€pattern€arises€when€this€type€of€question€was€asked€about€thoseÏrunning€education€in€this€country€(mean€ed.€high)€and€about€scientists€(mean€sci.€high),€plotted€inÏFigure€1b€in€comparison€to€education€and€the€confidence€in€those€running€the€Executive€BranchÏ(plotted€again€for€comparison).€€[Figure€1b€here]Ìà  àThus,€the€overall€conclusion€to€be€drawn€from€these€trends€parallels€the€judgment€fromÏcross€country€studies.€€Aggregate€trends€in€country€are€the€result€of€complex€processes.€€Even€atÐ v,h$( Ðthe€level€of€suggestive€evidence,€there€is€little€in€the€trends€in€support€for€environmentalÏprograms€and€in€the€average€levels€of€schooling€over€time€that€would€imply€a€close€link.€ÏEvaluating€the€prospects€for€these€connections€will€require€more€detailed€micro€analyses€thatÏincorporate€the€reasons€for€connections€between€education€and€individual€behavior€that€improvesÏenvironmental€quality.€€These€conceptual€and€empirical€issues€will€be€the€primary€focus€of€thisÏpaper.ÌÌ€à  àР̾  Ðò òIII.€à  àWTP€for€Education:€€Ð ž Ðà  àThe€Role€of€Preference€and€Household€Production€FunctionsÌÌA.à  àòòBackgroundó óóóÐ   Ðà  àMeasures€of€economic€value€are€commonly€understood€to€be€constructed€from€people'sÏchoices.€€A€choice€defines€a€lower€bound€for€an€individual's€willingness€to€pay€(WTP)€for€anyÏcommodity„„€marketed€or€not.€€For€example,€when€a€person€gives€up€one€or€more€years€of€workÏto€pursue€an€advanced€degree,€the€economic€value€must€be€at€least€as€great€as€what€wasÏforegone.€€Under€ideal€conditions,€we€expect€people€to€adjust€until€the€value€of€the€last€unitÏselected€exactly€equals€its€incremental€cost.€€By€reconstructing€the€implicit€net€"price"€in€thisÏtradeoff€and€expressing€it€as€a€rate€of€return€to€an€investment,€Becker€[1964]€and€Mincer€[1974]Ïdeveloped€a€model€that€has€been€one€of€the€key€organizing€frameworks€for€evaluating€educationÏover€the€past€thirty€years.€€Ð ~,p$( Ðà  àEducation€was€an€investment€providing€each€individual€that€decided€to€pursue€moreÏschooling€a€private€rate„of„return.€€This€return€could€be€estimated€directly€from€the€time€profileÏof€costs,€together€with€foregone€and€final€earnings€(Becker),€or€from€earnings€functionsÏ(Mincer).€Each€approach€was€derived€by€assuming€that€the€market€experience€of€peopleÏ"revealed"€their€economic€values.€€While€this€model€had€most€of€its€early€applications€inÏevaluating€the€returns€to€college€and€on„the„job€training,€it€has€also€been€extended€to€consider€aÏnumber€of€policy€issues,€including€more€recently,€the€effects€of€the€educational€quality€on€theÏrates€of€return€for€primary€and€secondary€education€(see€Behrman€and€Birdsall€[1983],€and€CardÏand€Krueger€[1992]€and€Betts€[1993]).Ìà  àModels€based€on€an€alternative€framework,€consistent€with€revealing€individual€WTP€forÏprimary€and€secondary€education,€have€also€been€developed€from€a€completely€different€Ïperspective€„€that€of€local€public€finance.€€Two€approaches€for€estimating€the€demand€for€localÏeducation€can€be€distinguished.€The€first€of€these,€sometimes€described€as€the€"standard€model,"Ïpostulates€a€functional€form€for€the€median€voter's€demand€for€a€local€public€good€(in€this€caseÏeducation)€based€on€what€is€assumed€to€be€the€individual's€(or€the€householdð ðs)€tax€rate€and€aÐ Š!| Ðmeasure€of€income.× ƒ ×Ý ƒÿÿÝòòÚ  Ú4Ú  ÚóóÝ  Ý×  ×€€With€a€variety€of€assumptions€(see€Oates€[1994]),€it€is€possible€to€recoverÐ \#N Ðestimates€of€the€price€and€income€elasticities€of€demand€by€relating€the€per€capita€expendituresÏfor€the€local€"public€good"€to€constructed€measures€for€the€median€voter's€tax€rate€and€income.€ÏA€variation€on€the€framework€relies€on€household€surveys.€€This€approach€is€sometimes€referredÏto€as€the€"micro"€method.€€It€can€use€either€households'€actual€choices€(Gertler€and€GlewweÏ[1990])€or€stated€choices€(Bergstrom,€Rubinfeld,€and€Shapiro€[1982])€to€estimate€a€representativeÐ v,h$( Ðhousehold's€preferences€for€different€types€of€educational€alternatives.€€Ìà  àIn€comparing€these€three€sources€of€information€about€people's€willingness€to€pay€forÏeducation,€it€seems€natural€to€ask€whether€each€approach€is€capturing€the€same€type€of€economicÏtradeoff€and,€therefore,€would€imply€comparable€estimates€for€the€typical€individual's€WTP€forÏeducation.€€I€believe€the€answer€is€òònoóó.€€These€models€actually€measure€the€demands€for€òòdifferentóóÐ VH Ðcommodities.€€Ìà  àTo€understand€why,€we€need€to€consider€the€assumptions€implicit€in€each€framework,Ïincluding€the€economic€agent€hypothesized€to€make€educational€decisions€and€the€specificationÏof€education€as€the€object€of€choice.€€With€that€background,€it€is€possible€to€relate€what€can€beÏlearned€from€the€models€about€the€economic€value€of€education.ò òÐ pb ÐÌB.à  àòòInterpreting€Past€Modeling€Strategiesó óóóÐ  Ðà  àTwo€features€distinguish€the€economic€decision€underlying€the€Becker„MincerÏframework€from€the€models€used€in€local€public€finance.€€First,€the€human€capital€model€assumesÏeach€individual€decides€about€education,€evaluating€his€(or€her)€market€alternatives€with€addedÏschooling€in€comparison€to€what€they€would€be€without€it.€€In€the€terminology€used€in€the€non„¼market€valuation€literature,€the€model€assumes€education€is€a€òòweak€complementóó€in€production€toÐ 2%$  Ðlabor€market€time.× ƒ ×Ý ƒÿÿÝòòÚ  Ú5Ú  ÚóóÝ  Ý×  ×€€That€is,€if€we€use€a€Becker„Mincer€estimate€of€the€rate€of€return€to€educationÐ 'ö" Ðas€the€basis€for€computing€education's€social€value,€then€from€society's€perspective,€the€enhancedÏvalue€of€labor€time€(due€to€education)€is€being€assumed€to€capture€all€of€the€relevant€economicÏvalue€of€education.× ƒ ×Ý ƒÿÿÝòòÚ  Ú6Ú  ÚóóÝ  Ý×  ×€€Second,€the€basic€unit€of€measurement€for€the€amount€of€education€selectedÐ z,l$( Ðis€usually€expressed€in€years€of€schooling€or€variables€that€identify€a€set€of€discrete€intervalsÏcorresponding€to€the€degrees€received€with€different€levels€of€schooling€(i.e.€high€school,Ïcollege,€etc).Ìà  àIn€contrast€to€this€approach,€models€of€the€demand€for€local€education€based€onÏhousehold€surveys€(and€the€median€voter€framework)€shift€the€decision€context€from€theÏindividual€deciding€about€her€education€to€an€adult€deciding€about€the€òòeducationóó€òòprovidedóó€òòtoóóÐ (  Ðòòchildrenóó.€€Usually€this€choice€involves€public€education€(certainly€this€is€the€case€for€medianÐ úì  Ðvoter€models).€€Both€the€object€of€choice€and€the€individual€making€the€choice€are€different.€€It€isÏan€educational€system€for€most€local€public€finance€applications.€€Demands€are€acknowledged€toÏbe€different€for€people€with€children€in€comparison€with€those€who€do€not€have€childrenÏ(Wyckoff€[1984])€and€sometimes€between€property€owners€versus€renters€(Oates€[1994]).Ìà  àThese€differences€are€most€often€explained€as€the€result€of€price€effects,€rather€than€fromÏthe€nature€of€the€object€of€choice€and€what€motivates€each€type€of€agent's€decision.€€For€example,Ïin€comparing€the€responses€of€households€with€and€without€children,€differences€in€the€demandsÏfor€public€schools€are€assumed€to€arise€because€the€former€consumes€the€educational€services,Ïpays€the€tax€price,€and€experiences€any€Tiebout€[1956]„related€property€appreciation€arisingÏform€the€tax€rate/€educational€services€provided.€€By€contrast,€the€household€without€children€(orÏwith€children€beyond€school€age)€faces€the€tax€price€and€experiences€only€the€property€valueÏappreciation€effects€of€education.Ìà  àTo€consider€how€economic€model€conceptualizes€the€economic€value€of€education€to€theÏindividual€hypothesized€to€be€making€educational€choices,€we€need€to€consider€what€each€modelÐ v,h$( Ðimplies€about€the€object„of„choice€and€how€it€is€assumed€to€affect€each€individual.€€This€shiftsÏthe€focus€of€the€analysis€away€from€imputed€net€prices,€the€issue€emphasized€in€the€comparisonsÏbetween€these€approaches,€to€all€aspects€of€these€decisions.Ìà  àConsider€first€the€Becker„Mincer€labor€market€model.€€Education€is€represented€by€theÏyears€of€schooling€selected€by€each€individual.€€This€formulation€implies€that€all€sources€ofÏeducation€are€perfect€substitutes€(from€the€perspective€of€€any€market€evaluation€of€the€addedÏvalue€of€labor€time€due€to€education)€and€ignores€any€role€of€education€in€contributing€to€otherÏaspects€of€an€individual's€activities.€€€Incorporation€of€quality€into€the€Mincer€earnings€model€(asÏin€Behrman€and€Birdsall),€relaxes€the€first€of€these€assumptions,€and€the€use€of€the€householdÏproduction€framework€(HPF)€broadens€the€second,€allowing€for€a€wider€set€of€potential€motivesÏfor€the€private€demands€for€education.€€Within€it,€education€can€be€interpreted€as€a€"public€input"Ïto€several€household€production€activities.€€Indeed,€this€is€exactly€the€argument€of€Haveman€andÏWolfe€[1984]€in€their€proposed€method€for€calculating€the€WTP€values€for€non„market€impacts.€Ìà  àNonetheless€this€formulation€does€not€necessarily€alter€any€efficiency€judgment€about€theÏprivate€decisions€for€education€implied€by€the€Mincer€model.× ƒ ×Ý ƒÿÿÝòòÚ  Ú7Ú  ÚóóÝ  Ý×  ×€€Instead,€it€suggests€the€laborÐ Š!| Ðmarket€tradeoff€may€not€accurately€capture€the€complete€marginal€value€an€individual€places€onÏeducation.€€Indeed,€based€on€their€arguments,€one€might€suggest€labor€market€estimates€provide€aÏpoor€lower€bound€for€an€individual's€full€marginal€WTP.€€Because€these€values€accrue€to€privateÏindividuals,€this€expansion€in€the€sources€of€value€for€education€would€help€to€explain€whyÏindividuals€would€select€educational€investments€when€they€yield€lower€rates€of€return€relative€toÏcomparable€investment€alternatives.€€However,€it€would€not€provide€a€basis€for€measuringÐ v,h$( Ðwhether€education€had€additional€social€effects€that€might€be€considered€ð ðpublic€goodsðð.€€Ìà  àThis€conclusion€may€be€seen€more€directly€by€applying€the€revealed€preference€logicÏused€in€the€non„market€valuation€of€environmental€resources€to€this€model.€€The€key€elements€inÏthe€economic€tradeoff€used€to€reconstruct€values€are€given€in€the€model's€specification€of€theÏmarginal€rate€of€substitution€(MRS)€between€education,€as€an€argument€in€individualÏpreferences,€and€a€numeraire.€€The€simple€"individual€preferences"€model€has€education€as€aÏprivate€good€entering€in€preferences€with€all€gain€from€an€additional€"unit"€captured€by€theÏindividual.€€The€HPF€generalization€has€education€exerting€effects€on€the€production€of€multipleÏfinal€services,€but€they€are€all€captured€by€the€individual€(or€her€household)€who€has€invested€inÏeducation.€€This€makes€the€resulting€expression€for€the€MRS€more€complex€(because€it€nowÏcombines€the€contributions€of€final€service€flows€linked€together€through€the€effects€of€educationÏon€each)€but€does€not€necessarily€imply€effects€outside€the€household€(see€Haveman€and€Wolfe'sÏ[1984]€equation€(3)€as€an€example).€€To€gauge€the€tradeoff€between€added€education€and€aÏnumeraire,€the€analyst€must€consider€all€of€the€effects€on€these€within€home€activities.Ìà  àWith€aggregate€versions€of€the€median€voter€model,€the€object€of€choice,€or€specificationÏfor€the€commodity€used€to€represent€education€is€òòneveróó€developed.€€Indeed,€in€most€cases€theÐ \#N Ðprice€of€a€unit€of€education€is€also€left€undefined.€€The€empirical€models€are€based€on€demandÏspecifications€that€allow€estimates€of€the€price€and€income€elasticities€to€be€recovered€from€perÏcapita€expenditure€models.× ƒ ×Ý ƒÿÿÝòòÚ  Ú8Ú  ÚóóÝ  Ý×  ×€€The€object€of€choice€in€the€survey„based€models€is€also€describedÐ Ò(Ä $ Ðby€responses€to€questions€about€increasing€"expenditures€on€local€public€schools"€together€withÏan€analyst's€construction€of€what€prices€people€perceive€they€are€paying€for€these€increases.€Ð v,h$( ÐPrivate€benefits€from€enhanced€education€in€this€model€are€derived€by€families€with€children€inÏthe€schools€and€by€homeowners€through€any€positive€net€capitalization€of€the€local€publicÏeducation€in€their€property€values.Ìà  àWyckoff€[1984]€has€argued€that,€after€taking€account€of€property€value€effects,Ïhouseholds€with€no€children€(or€with€children€beyond€school€age)€have€demands€for€local€publicÏeducation€that€motivated€exclusively€because€of€its€public€good€characteristics.× ƒ ×Ý ƒÿÿÝòòÚ  Ú9Ú  ÚóóÝ  Ý×  ×€€Those€withÐ (  Ðchildren€in€school€have€demands€due€to€both€the€private€and€public€components.€€Thus,€in€hisÏframework,€by€sorting€the€sample€into€these€two€groups,€and€assuming€common€underlyingÏpreferences,€these€different€sources€of€demand€for€local€public€education€(as€a€mixed€good)€canÏbe€distinguished.€€His€estimates€imply€that€the€total€demand€(considering€both€the€private€andÏpublic€components)€is€more€inelastic€than€the€public€component.€€With€further€assumptions,€theyÏwould€imply€that€the€share€of€WTP€for€local€education€due€to€the€public€component€is€aboutÏninety€percent.× ƒ ×Ý ƒÿÿÝòòÚ  Ú10Ú  ÚóóÝ  Ý×  ×Ð æØ Ðà  àThe€strategy€used€to€develop€Wyckoff's€estimates€is€comparable€to€the€distinction€drawnÏin€the€valuation€of€environmental€resources€between€use€and€nonuse€values€for€these€resources.€ÏThe€former€correspond€to€economic€values€attributed€to€òòinóó€òòsituóó€uses,€while€the€latter€do€notÐ \#N Ðrequire€any€form€of€observable€consumption.€€Nonuse€values€correspond€to€the€values€for€anyÏpure€public€good€services€a€resource€provides€(see€Plourde€[1975]€and€McConnell€[1983]).€€OneÏearly€approach€for€measuring€these€components€of€total€value€distinguished€users€fromÏnonusers.× ƒ ×Ý ƒÿÿÝòòÚ  Ú11Ú  ÚóóÝ  Ý×  ×€€The€latter's€economic€values€for€improvements€in€a€resource€were€assumed€(byÐ ¤*–"& Ðdefinition)€to€reflect€nonuse€values.€€Ð v,h$( Ðà  àThe€only€relevant€definition€of€economic€value€derives€from€a€choice.€€It€is€the€òòobjectóó€òòofóóÐ  Ðòòchoiceóó€that€provides€the€basis€for€how€we€interpret€that€value.€€When€analysts€adopt€a€revealedÐ à Ò Ðpreference€approach€to€estimate€people's€economic€values,€they€must€focus€on€situations€whereÏthere€is€some€reason€to€observe€people's€decisions.€€The€linkage€between€the€observed€decisionÏand€the€underlying€object€of€choice€is€then€used€to€construct€the€necessary€tradeoff€informationÏfor€estimating€economic€values.€€Ìà  àTo€illustrate€how€this€works,€consider€two€of€the€Haveman„Wolfe€estimates€of€theÏincremental€WTP€for€education,€arising€from€its€effects€on€non„market€activities.€€The€first€ofÏthese€is€their€calculation€of€the€value€of€education€through€its€impact€on€the€attainment€ofÏ"desired€family€size€and€child€spacing."€€Their€calculations€estimate€the€marginal€effect€ofÏschooling€on€family€size€and€child€spacing.€€The€net€changes€in€household€time€related€to€theirÏactivities€are€then€interpreted€in€terms€of€the€non„market€time€effects€of€these€decisions€andÏvalued€at€the€husband's€market€wage€rate€to€estimate€the€marginal€value.€€The€second€considersÏthe€effects€of€education€on€criminal€activity.€€Here,€these€authors€evaluate€the€years€of€schoolingÏthat€would€be€òòequivalentóó€to€an€additional€unit€of€police€time€in€a€local€jurisdiction€in€terms€of€itsÐ Š!| Ðcontribution€to€the€probability€of€criminal€apprehension€and€punishment.€€Using€the€dollar€costÏof€police€time,€an€individual's€incremental€WTP€is€imputed€from€these€two€sources.€€TheÏcommunity€effect€(or€public€good)€value€of€the€increment€is€derived€by€multiplying€the€estimatedÏindividual€value€by€the€number€of€people€in€the€community.€€This€approach€relies€on€an€analogyÏto€the€traditional€public€good€logic€(i.e.,€nÔ€+ XX-³Ô€ððÔ€X-³X +Ô€MRS,€with€n€=€number€of€people).× ƒ ×Ý ƒÿÿÝòòÚ  Ú12Ú  ÚóóÝ  Ý×  ×Ð ¤*–"& Ðà  àNeither€of€these€estimates€is€an€economic€value€for€the€public€good€effects€of€education.€Ð v,h$( ÐEach€is€a€clever€economic€reconstruction€of€the€values€of€time€(household€and€police)€in€specificÏactivities€that€are€òòattributedóó€òòtoóó€education.€€This€attribution€arises€because€there€were€modelsÐ à Ò Ð"describing"€how€the€levels€of€schooling€influence€these€outcome€variables.€€Even€if€the€modelsÏcorrectly€reflect€an€effect€due€to€education,€people€did€òònotóó€make€the€òòchoiceóó€to€undertake€theÐ „ v Ðeducation€in€order€to€accomplish€these€goals.€€They€are€constructed€values;€akin€to€what€wouldÏbe€expected€if€all€these€activities€took€place€in€ideal€markets€so€that€any€discrepancies€in€impliedÏvalues€would€create€incentives€for€arbitrage.€€Ìà  àDecompositions€of€value€into€use€(or€private€good)€related€measures€of€WTP€reflect€onlyÏthe€fact€that€a€private€decision€can€be€recorded;€and€it€is€assumed€that€a€reason€for€the€decision€isÏfound€in€the€object€being€valued.€€In€Becker„Mincer€human€capital€model€this€amounts€toÏòòrequiringóó€that€the€exclusive€basis€for€an€individual€valuing€education€is€due€to€the€enhanced€laborÐ B4 Ðmarket€returns.€€By€contrast,€for€the€local€public€finance,€any€recovery€of€components€of€the€totalÏvalue€relies€on€observable€characteristics€of€a€household€(i.e.,€school€age€children).€€These€areÏused€to€identify€those€households€with€both€private€and€public€demands,€along€with€theirÏassociated€"contributions"€to€each€one's€total€value.€€Experience€with€these€types€of€approachesÏin€estimating€values€for€non„marketed€environmental€resources€suggests€that€a€more€fruitfulÏstrategy€may€be€found€in€specifying€more€completely€the€economic€commodity€of€interest€andÏhow€it€affects€people.ò òÐ 'ò" ÐÌC.à  àòòDescribing€Education's€Role€in€Individual€Behavioró óóóÐ ¦*˜"& Ðà  àOne€lesson€that€emerges€from€these€past€efforts€is€the€central€role€of€the€linkages€betweenÐ z,l$( Ðeducation€and€other€types€of€decisions€that€are€used€to€re„construct€people's€values.€€In€thisÏsection€we€consider€three€preference€specifications€and€describe€how€each€might€be€used€toÏevaluate€connections€between€education€and€other€public€goods.€€While€the€primary€focus€of€theÏdiscussion€will€be€on€how€they€could€be€used€to€consider€the€relationship€between€education€andÏenvironmental€quality€(as€a€public€good),€the€arguments€are€general.€€My€objective€is€to€evaluateÏtwo€aspects€of€modeling€decisions€that€attempt€to€explain€the€sources€of€"public€good"€values€forÏeducation.€€First,€I€consider€whether€indirect€links€between€education€and€other€public€good€areÏcapable€of€providing€one€explanation€for€what€are€sometimes€postulated€as€the€"public€good"Ïservices€or€social€components€of€education€Ì(e.g.€Wyckoff€[1984]).€€Second,€to€the€extent€there€is€empirical€evidence€to€support€such€links,€IÏdiscuss€whether€observed€behavior€(or€the€responses€captured€in€traditional€models)€could€beÏused€to€estimate€these€public€benefits.Ìà  àÌThe€first€model€is€the€simplest€with€a€private€good,€X,€a€public€good,€Q,€and€education,€E.€€It€isÏgiven€in€equation€(1).€€The€additional€term,€ðð,€is€a€parameter€included€to€reflect€the€heterogeneityÐ Š!| Ðin€individual€preferences.ÌÌ€à  àòòModel€Ióó€à ¸ àÐ 'ò" Ðß‘€! "YIEz ,x® 0  €@Xdddddddd@E°Ò(x®Zd}çÎs' ÈLÒ(@‘ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âÐ v,h$( ÐAs€noted€earlier,€restrictions€on€preferences€(or€physical€connections€due€to€location)€link€X€to€QÏand€generally€provide€the€basis€for€revealed€preference€measures€of€the€implicit€tradeoffÏperceived€to€be€involved€in€increasing€Q.€€For€example,€weak€complementarity€(see€MlerÏ[1974],€Bradford€and€Hilderbrandt€[1977])€restricts€the€marginal€rate€of€substitution€between€QÏand€X€to€be€zero€when€the€demand€for€X€is€zero€(i.e.€)Ô€Nô‰NXX-³ÔÔ€N <NNNô‰Ô€ß€%'&G73z X ¡Ú p  €@@@EVã¾ã¾<ŽsNß).Ô€X¨sXNN <ÔÔ€X-³XXX¨sÔ€€A€connection€between€QÐ VH Ðand€E€would€be€difficult€to€interpret€from€empirical€model€based€on€this€framework.€€Is€EÏserving€as€an€indicator€of€differences€in€individual€preferences€(i.e.,€as€a€proxy€for€theÏunobserved€sources€of€heterogeneity€assumed€to€be€present€in€ðð€or€a€component€of€a€compositeÐ tf Ðgood€with€Q€that€is€linked€to€X?€€Pursuing€the€second€part€of€this€question€is€easier€with€a€moreÏstructured€model,€building€on€the€household€production€framework€(HPF)€and€will€be€discussedÏas€part€of€the€development€of€Model€II€below.Ìà  àWe€could€also€simply€postulate€a€public€good€element€to€education€(though€the€measureÏused€to€represent€E€would€likely€be€different€from€what€has€been€used€in€most€empirical€studies)Ïand€argue€it€was€linked€with€Q€in€any€relationship€hypothesized€to€govern€X.€€For€example,ÏBockstael€and€Kling€[1988]€have€discussed€situations€where€sets€of€public€goods€can€be€treatedÏas€weak€complements€with€specific€private€goods.€€This€would€imply€that€the€values€for€Q€couldÏnot€be€recovered€from€individual€decisions€about€X,€without€also€considering€E.€€Here€too,€theÏdistinction€between€a€public€good€rationale€for€E€and€a€difference€in€the€taste€for€Q€òòdueóó€òòtoóóÐ ¨(š " Ðòòvaryingóó€òòlevelsóó€òòofóó€òòeducationóó€is€largely€an€assumption.€€We€can€reinforce€the€assumption€inÐ z*l"$ Ðempirical€modeling€through€the€ways€our€models€measure€E.€€The€public€good€effects€would€beÐ L,>$& Ðmost€likely€to€be€based€on€community€level€measures€for€E,€while€the€private€good€is€more€likelyÏto€be€captured€with€individual€measures€of€education€such€as€years€of€schooling.Ìà  àThe€logic€underlying€such€a€restriction€is€that€people's€enjoyment€of€the€set€of€Q's€canÏonly€be€realized€if€they€undertake€some€private€activity.€€For€example,€the€preservation€of€uniqueÏart,€scientific€or€historical€material,€can€only€be€appreciated€as€"enhancing€culture"€if€the€citizensÏinvolved€in€such€activities€simultaneously€attain€the€levels€of€education€necessary€to€understandÏtheir€significance.€€Then€we€take€the€next€step€(required€to€use€a€revealed€preference€method)Ïand€assume€that€for€these€composite€gains€to€be€realized,€the€people€must€be€able€to€devote€timeÏto€visiting€museums,€unique€scientific,€historical€or€cultural€sites,€etc.Ìà  àThese€analytical€arguments€raise€more€general€questions€about€how€we€describeÏcommodities€that€have€public€good€attributes.€€Are€they€part€of€more€general€composites€thatÏinvolve€an€array€of€services?€€And€do€these€broader€relationships€condition€how€the€group€ofÏpublic€goods€influences€preferences?€€This€point€was€raised,€implicitly,€in€the€discussion€of€theÏembedding€question€as€part€of€a€set€of€issues€recently€raised€about€the€use€of€survey€methods€toÏvalue€environmental€resources.× ƒ( ×Ý ƒÿÿÝòòÚ  Ú13Ú  ÚóóÝ  Ý×  ×€€Because€all€of€these€types€of€assumptions€must€be€maintainedÐ Š!| Ðhypotheses,€they€cannot€be€tested.€€Informal,€qualitative€evaluations,€relying€on€cognitiveÏinterviews€and€focus€groups,€may€provide€insights€into€how€people€evaluate€groups€of€publicÏgoods€and€whether€they€evaluate€them€as€composites.€€Nonetheless,€for€the€short„term,Ïspecifying€groups€of€public€goods,€including€education€in€that€group,€as€weak€complements€to€aÏprivate€commodity,€is€unlikely€to€offer€practical€empirical€insights.Ìà  àThe€second€type€of€modeling€structure,€with€the€most€direct€empirical€relevance€toÐ v,h$( Ðenvironmental€resources€is€the€ð ðhousehold€production€storyðð€and€it€follows€the€Haveman„WolfeÏ[1984]€framework.€€As€in€the€examples€they€identified,€education€is€assumed€to€enhanceÏproduction€of€those€activities€involving€environmental€resources.€€However,€this€effect€isÏunlikely€to€be€comparable€to€a€pure€public€good.€€For€this€second€model,€we€would€re„writeÏequation€(1)€with€more€specific€implicit€structure€on€how€the€goods,€services,€environmentalÏresources,€and€education€combine€to€enhance€well„being€as€in€equation€(2):òòÐ (  ÐModel€IIóóÐ úì  ÐÌòòß‘€*)+YIEz ,x¤ 0  €@Xdddddddd@E°žx¤”ûÁ "Îî' ÈLž@‘ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âóóÌà  àEach€fòòióó(.)€corresponds€to€a€household€production€activity€with€private€goods€(Xòòióó),Ð  Ðenvironmental€public€goods€(Qòòióó)€and€education€(E).€€The€form€given€in€(2)€illustrates€the€varietyÐ æØ Ðof€specifications€that€would€be€possible.€€For€example,€in€the€first€HPF€a€vector€of€private€goodsÏ(Xòò1óó),€a€unique€environmental€resource€(Qòò1óó)€and€education€contribute€to€production,€E.€€That€sameÐ Š!| ÐQòò1óó€is€seen€to€enhance€fòò3óó(.)€as€does€education.€€Education€could€enter€production€relationshipsÐ \#N Ðwithout€Qòò1óó,€as€was€the€case€in€several€of€the€Haveman„Wolfe€studies€and€illustrated€in€myÐ .%   Ðexpression€for€the€first€activity.€€The€situation€described€by€fòò1óó(.)€is€also€analogous€to€theirÐ 'ò" Ðexample€of€education's€contribution€to€reducing€crime.€Ìà  àLest€we€get€carried€away€with€the€possibilities€for€"story„telling,"€it€is€important€toÏacknowledge€the€limitations€in€what€can€be€learned€from€observations€of€people's€behavior.€Ð v,h$( ÐTwenty€years€ago€Mler€demonstrated€that€the€marginal€willingness€to€pay€for€òòanyóó€non„Ð  Ðmarketed€good€could€be€expressed€in€terms€of€the€price€of€any€private€good€and€the€marginal€rateÏof€substitution€between€that€private€good€and€the€non„marketed€good€as€in€equation€(3).€€€€€€€Ì߉€-,.QA=zx 0  €@Xdddddddd@E°„ x´dó ìÎö' ÈL„ @‰ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âÌÌÌwhere:à  àà ` àà ¸ àà  àMWTP€=€marginal€willingness€to€pay€for€QÐ ž Ðà  àà ` àà ¸ àà  àà h àe(.)€=€Hicksian€expenditure€function€Ì߉€0/QA=zx 0  €@Xdddddddd@E°Bx¬ d GLB@s‰ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âÌà  àUnfortunately,€when€Q€is€not€a€choice€argument,€there€is€no€real€basis€for€observing€theÏMRS.€€This€is€where€the€maintained€hypotheses€restricting€preferences€and€the€HPF€"stories"Ïenter.€€They€describe€cases€where€the€decisions€observed€provide€the€analyst€sufficientÏinformation€to€reconstruct€measures€of€economic€value.€€One€of€the€most€widely€used€suchÏassumptions€is€perfect€substitution.€€That€is,€we€can€estimate€an€individual's€MWTP€for€Q€ifÏthere€is€a€perfect€substitute€using€the€change€in€expenditures€for€the€private€good€thatÏaccompanies€any€change€in€Q€(see€Smith€[1991a]€for€the€details).€€As€a€rule,€this€restriction€hasÏbeen€used€for€empirical€models€of€averting€behavior.€€Consider,€for€example,€the€case€of€aÐ v,h$( Ðdeterioration€in€air€quality.€€If€a€household€cleaning€activity€is€a€perfect€substitute€for€air€qualityÏin€completely€mitigating€the€full€effects€of€that€pollution€on€people€(e.g.,€air€pollution€causesÏincreased€soiling€of€houses,€automobiles,€clothing,€etc.),€then€the€MWTP€would€be€measured€byÏthe€increased€cleaning€expenditures€required€by€increased€pollution.× ƒ1 ×Ý ƒÿÿÝòòÚ  Ú14Ú  ÚóóÝ  Ý×  ×€€To€the€extent€thatÐ „ v Ðeducation€contributes€to€the€effectiveness€of€such€a€production€process,€then€we€could€argue€thatÏincremental€cost€savings€(reduced€expenditures€required€for€€cleaning)€because€of€the€learningÏaccompanying€education€offer€a€measure€of€this€component€of€education's€economic€value.Ìà  àThis€connection€does€not€imply€education€is€a€public€good.€€Any€public€action€such€as€anÏinformation€or€risk€communication€program€that€is€described€within€the€averting€behaviorÏframework€relies€on:€Ì(1)à  àconsumers'€incomplete€knowledge€about€an€environmental€problem€or€method€toÐ B4 Ðà  àreduce€its€effects;Ì(2)à  àconsumers'€"inappropriate"€risk€perceptions€in€comparison€with€experts'€judgmentsÐ æØ Ðà  àabout€environmental€risk:Ì(3)à  àcost€advantages€to€the€collective€provision€of€information;€orÐ Š!| Ð(4)à  àcost€advantages€to€decentralized€responses€for€mitigating€the€effects€of€anÐ \#N Ðà  àenvironmental€externality.Ìà  àIn€each€case,€education€can€be€argued€to€have€a€complementary€effect€by€enhancing€theÏeffectiveness€of€a€household's€private€responses.€€To€understand€how€the€logic€works,€considerÏan€example.€€Suppose€the€action€to€be€evaluated€concerns€contamination€of€groundwater€that€isÏused€as€a€drinking€water€source€through€private€wells€(see€Poe€and€Bishop€[1993]).€€PrivateÐ v,h$( Ðhousehold€action€may€be€the€most€cost€effective€response€with€testing€and€filters,€but€this€wouldÏrequire€household€recognition€of€the€problem€and€a€corresponding€set€of€actions.€€The€externalityÏmotivates€the€public€intervention€and€education€enhances€the€effectiveness€of€that€publicÏintervention.€€The€opportunity€for€a€household€to€mitigate€individually€creates€a€privateÏcommodity„„ððcleanðð€water€from€that€household's€perspective.€€If€this€is€the€least€cost€responseÏand€education€enhances€its€success,€then€education€reduces€the€cost€of€responding€to€a€localÏpublic€bad„„the€contaminated€groundwater.Ìà  àA€second€line€of€reasoning,€popular€among€environmental€policymakers,€is€that€people'sÏrisk€perceptions€for€environmental€sources€of€risk€are€incompatible€with€the€expertsð ðÐ ž Ðjudgments.× ƒ2 ×Ý ƒÿÿÝòòÚ  Ú15Ú  ÚóóÝ  Ý×  ×€€In€this€case,€too€much€private€(and€public)€action€is€taken€for€some€environmentalÐ pb Ðproblems€and€too€little€for€others€(because€of€these€misperceptions).€€This€can€also€be€describedÏas€a€type€of€HPF€model€(see€Ehrlich€and€Becker€[1972]).€€Treating€self€protection€as€an€influenceÏon€subjective€probabilities,€with€education€as€a€factor€in€"improving"€perceptions,€allows€aÏcomparable€explanation€to€be€presented.€€It€can€either€be€in€reducing€the€"incorrect€perceptions"Ïor€enhancing€the€effectiveness€of€risk€communication.Ìà  àIn€all€of€these€cases,€education€is€not€a€public€good.€€It€has€social€consequences€because€itÏaffects€the€performance€of€other€private€or€public€actions.€€These€examples€avoid€the€criticismÏthat€might€be€made€of€many€of€the€Haveman„Wolfe€non„market€effects.€€In€several€of€their€cases,Ïthe€consequences€were€also€specific€to€the€household,€education€enhanced€effectiveness.€€WhyÏthen€wouldn't€this€be€acknowledged€by€the€household€and€reflected€in€their€educational€choices?€ÏThe€policy€implications€of€the€existence€of€these€types€of€effects€would€not€be€to€call€forÐ v,h$( Ðincreased€public€intervention,€but€rather€a€realization€that€more€detailed€models€are€needed€toÏexplain€behavior.€€That€is,€when€analysts€reconstruct€what€those€choices€mean,€account€must€beÏtaken€of€these€effects.€€Ìà  àHere€the€externality€mitigating€or€cost€reducing€properties€of€an€information€programÏcreate€the€"publicness."€€Education€of€the€targeted€population€simply€reduces€the€"costs"€of€anyÏactions€to€respond€to€these€problems.Ìà  àThe€last€category€of€model€for€individual€(or€household)€behavior€does€provide€a€linkÏbetween€education€and€a€public€good€that€could€be€interpreted€to€expand€its€public€dimension.€€ItÏis€given€in€equation€(4)€and€requires€some€explanation.ÌòòÌModel€IIIóóÐ B4 ÐÌ߉€435QA=zx 0  €@Xdddddddd@E°æxêìdZ"Α' ÈLæ@N‰ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âÌThis€preference€specification€is€intended€to€describe€an€individual€identified€by€the€subscript€"1"Ïfor€convenience.€€It€includes€a€vector€of€private€goods,€education€(treated€here€as€a€private€good),Ïand€the€middle€terms€with€qòò1óó€and€the€function€h(.).€€The€intention€is€to€reflect€a€set€of€modelsÐ 'ò" Ðdescribing€how€people's€actions€contribute€to€the€availability€of€public€goods.Ìà  àThe€most€common€application€in€these€models€has€been€for€describing€charitableÏcontributions€(see€for€example€Andreoni€[1989]),€but€the€implications€of€the€models€are€muchÐ v,h$( Ðmore€general.€€Each€person€can€engage€in€activities€(or€contribute)€to€their€production.€€This€typeÏof€action€is€represented€here€for€individual€1€as€qòò1óó.€€The€collection€of€all€individuals'€actionsÐ à Ò Ðleads€to€a€public€good,€represented€here€with€the€function€h(.).€€The€other€arguments€Ìin€h(.)€are€the€activities€of€others.€€In€the€simplest€modelsÔ€Nô‰NXX-³ÔÔ€N <NNNô‰Ô:€€Ð „ v Ð߉€87QA=zx 0  €@Xdddddddd@E°nxQÀ d÷‰Ln@µ‰ßâ °±°±°(#°(#â°â °±°±°±°±â°â °(#°(#°±°±âÌÔ€X¨sXNN <ÔÔ€X-³XXX¨sÔThis€specification€has€been€used€for€the€charitable€contributions€literature€where€the€totalÐ ø  Ð"output"€of€the€charity€is€the€sum€of€all€contributors.€€It€is€also€one€that€is€frequently€adopted€forÏdescribing€public€goods€that€result€from€individual€action€(see€Cornes€and€Sandler€[1986][1994]Ïand€Sandler€[1992]).Ô€Nô‰NXX-³ÔÔ€N <NNNô‰ÔÐ |n ÐÔ€X¨sXNN <ÔÔ€X-³XXX¨sÔà  àIt€is€written€generally€in€equation€(4),€with€the€function€h(.),€used€to€acknowledge€thatÐ N@ Ðpeople's€responses€to€the€anticipated€actions€of€others€depends€on€how€each€person€perceives€theÏpublic€good€is€linked€to€individual€behavior.€€Vivid€examples€of€this€relationship€were€providedÏin€Hirshleifer's€[1983]€description€of€the€role€of€the€function,€labeled€by€him€as€a€òòsocialóóÐ Ä ¶ Ðòòcompositionóó€òòfunctionóó.€€More€recently,€both€experimental€(Harrison€and€Hirshleifer€[1989])€andÐ –"ˆ Ðconceptual€analysis€(Cornes€[1993])€have€confirmed€and€extended€his€arguments.Ìà  àHow€do€these€analytical€details€relate€to€the€treatment€of€education€as€having€anÏ"indirect"€public€good€effect€through€its€impact€on€environmental€quality?€€The€answer€lies€in€theÏchain€of€effects€of€education€on€each€qòòióó,€such€as€the€one€identified€for€individual€1,€and€then€eachÐ Þ)Ð!& Ðqòòióó€on€the€resulting€public€good€represented€through€h(.).€€In€the€environmental€context,€doesÐ °+¢#( Ðeducation€promote€environmentally€"responsible"€behavior?€€This€could€be€the€specific€contextÐ ‚-t%* Ðfor€interpreting€a€measure€of€€ß‚€%:9J:6z \ ß„Ò p  €@@@E„ÒüÿÚÿ> F b‚߀€as€an€estimate€of€the€incremental€WTPÐ  Ðof€education€because€of€its€public€good€effects.€€They€arise€from€education's€induced€effects€onÏprivate€behavior€that€in€turn€contributes€to€providing€a€larger€amount€of€public€good.Ìà  àOf€course,€this€type€of€logic€is€not€limited€to€private€behavior€that€might€be€describedÏenvironmentally€responsible.€€Often€in€describing€the€potential€for€education's€public€goodÏeffects,€analysts€have€suggested€that€an€informed€population€is€likely€to€act€more€responsibly€inÏother€ways€that€have€been€associated€with€externalities,€including€actions€that€might€take€place€inÏa€public€choice€context€(see€Jimenez€[1994]).€€Nonetheless,€to€move€from€concept€to€practice,Ïthree€connections€must€be€measured:Ì€à  àðð€a€relationship€describing€the€effect€of€education€on€the€specific€type€of€private€à Ð à€Ð @2 Ðà  à€€behavior€that€contributes€to€a€public€good€(i.e.€MRSòòEqióó)Ð  Ðà  àðð€a€relationship€specifying€how€an€increment€in€private€behavior€changes€the€level€of€¼à  à€à ` à€€the€public€good€(Hirshleifer's€arguments€clearly€document€that€this€relationshipÐ ¶¨ Ðis€à  à€à ` à€€important€and€it€need€not€be€a€one„for„one€relationship,€ð,ðh/ð,ðqòòióó).Ð ˆ!z Ðà  àðð€a€function€describing€how€monetary€measure€the€measure€of€the€economic€value€¼à  à€à ` à€€varies€with€the€amount€of€the€public€good€(i.e.€PòòxóóððMRSòòQXóó).Ð ,% Ðà  àAs€the€next€section€discusses,€we€usually€do€not€have€the€ability€to€completely€constructÏthis€sequence€for€a€typical€household.€€Instead,€empirical€studies€will€reflect€one€or€more€parts€ofÏthe€sequence€(often€from€different€studies).€€Usually€these€types€of€studies€report€the€results€asÏsuggestive€of€education's€public€good€(or€social€values)€or€develop€illustrative€calculations€byÐ t,f$& Ðsupplementing€what€is€available€with€maintained€assumptions€in€order€to€estimate€the€desiredÏeconomic€value€(this€was€the€strategy€adopted€in€Haveman€and€Wolfe).ÌÌò òD.à  àòòImplications€of€the€Models€for€Measurementó óóóÐ „ v Ðà  àTable€1€summarizes€the€issues€identified€in€this€overview€of€three€alternative€conceptualÏstructures€for€describing€the€role€of€education€for€people's€behavior.€€My€discussion€hasÏconsidered€the€types€of€assumptions€required€to€evaluate€whether€private€decisions€aboutÏeducation,€interpreted€as€a€private€good,€can€have€non„market€and€especially€public€good,Ïconsequences.€€The€primary€focus€of€the€applications€considered€(in€a€few€simple€examples)€hasÏbeen€with€respect€to€environmental€resources€because€they€are€the€source€of€the€empirical€resultsÏdiscussed€in€the€next€section.€€[Table€1€here]Ìà  àTwo€of€the€elements€in€the€table€have€not€been€explicitly€discussed.€€Both€of€the€newÏitems€relate€to€informal€judgments€about€the€prospects€for€success€in€addressing€specificÏempirical€questions.€€Thus,€there€seems€little€basis€for€an€extensive€discussion.€€The€first€of€theseÏinvolves€whether€the€models€used€to€describe€behavior€are€likely€to€permit€distinguishingÏeducation's€role€as€a€proxy€for€taste€(i.e.€an€unobserved€source€of€heterogeneity)€versus€as€aÏcausal€factor.€€My€evaluation€is€that€only€in€the€case€of€household€production€models,€does€thisÏtype€of€distinction€seem€likely€to€have€reasonable€chance€of€success.Ìà  àThe€second€measurement€issue€concerns€questions€that€would€need€to€be€addressed€inÏorder€to€implement€the€models.€€Here€my€judgments€are€more€diverse,€depending€on€theÏparticular€framework€and€application€to€be€considered.€€Nonetheless,€there€are€a€few€commonÐ x,j$( Ðthemes.€€The€measure€intended€to€represent€education€as€an€object€of€choice„„whether€it€isÏspecific€to€the€individual,€such€as€years€of€schooling,€(as€in€models€II€and€III)€or€more€general,€asÏa€measure€of€the€educational€output€in€a€community€will€be€important€to€the€ability€to€distinguishÏheterogeneity€from€causal€relationships.€€It€is€also€important€to€consider€the€specific€measuresÏused€to€describe€the€outcomes€influenced€by€education€and€how€these€will€be€used€in€non„marketÏbenefit€measurement.€€Without€restrictive€assumptions€such€as€perfect€substitution€in€HPFÏrelations,€estimates€like€those€proposed€by€Haveman€and€Wolfe€will€not€accurately€reflectÏHicksian€WTP.Ìà  àFinally,€when€models€acknowledge€a€jointness,€we€face€a€problem€in€isolating€the€valueÏattributed€to€education.€€The€source€of€that€jointness,€whether€with€other€public€goods,€acrossÏHPF€descriptions€of€private€activities,€or€due€to€the€effect€education€has€on€people€and€their€jointÏactions€for€a€common€good€(i.e.,€Hirshleifer's€social€composition€function)€does€not€necessarilyÏaffect€the€relevance€of€the€issue.€€All€sources€lead€to€questions€comparable€to€those€found€in€costÏallocation€problems.€€That€is,€how€do€we€allocate€what€might€be€argued€as€the€joint€benefits€dueÏto€either€a€composite€of€public€goods€or€arising€from€people's€actions,€in€attempting€toÏdistinguish€the€social€benefits€of€education?€€Ìà  àThe€last€column€identifies€some€conceptual€issues€that€may€be€opportunities€for€furtherÏresearch€where€greater€interaction€between€economists€who€attempt€to€use€people's€behavior€toÏvalue€environmental€resources€with€those€interested€in€comparable€issues€for€public€education.€ÏThe€first€of€these€is€an€analogy€to€the€Baumol,€Panzar,€and€Willig€[1982]€concept€of€a€distinctionÏin€the€properties€of€costs€when€they€are€involved€in€producing€individual€outputs€separatelyÐ v,h$( Ðversus€producing€them€together.€€Economies€of€scope€refer€to€the€reduced€total€cost€argued€(inÏsome€cases)€to€be€associated€with€producing€outputs€together.€€The€Bockstael„Kling€suggestionÏof€weak€complementarity€between€groups€of€public€goods€and€a€private€commodity€may€wellÏreflect€a€preference€equivalent.€of€this€type€of€relationship.€€That€is,€when€combined,€the€publicÏgoods€are€more€valuable.€€Exceptional€visibility€and€a€view€from€the€South€Rim€of€the€GrandÏCanyon€would€be€a€pair€that€might€be€cited€in€discussions€of€environmental€resources.€€However,Ïone€could€also€identify€education€together€with€cultural,€historic,€or€scientific€resources.€€WouldÏbiodiversity€be€a€different€public€good€to€a€community€with€a€higher€overall€level€of€educationÏthat€appreciates€the€interconnections€linking€ecological€resources?Ìà  àFinally,€there€is€an€important€omission€in€the€analyses€reviewed€here€that€have€providedÏthe€primary€basis€for€describing€people's€behavior€and€its€implications€for€the€value€of€education.€ÏThey€have€all€focused€on€interior€solutions,€considering€the€effects€of€education€on€MRSs.€€ByÏconducting€the€analysis€at€the€intensive€margin€of€choice,€these€models€overlook€a€view€ofÏeducation€that€describes€economic€behavior€as€a€composite€of€decisions€made€at€the€extensiveÏand€intensive€margins€of€choice.€€That€is,€there€will€be€people€who€do€not€participate€or€whoÏchange€their€decisions€to€participate€based€on€the€action€under€study€(see€Bergstrom,€Blume,€andÏVarian€[1986]€for€a€discussion€of€this€issue€in€the€context€of€charitable€contributions).€€TheÏconditions€defining€decisions€at€the€extensive€margin€are€often€referred€to€as€individualÏrationality€or€participation€constraints,€and€require€evaluation€of€total€gains€and€losses€arisingÏfrom€choices.€€For€marketed€commodities,€these€conditions€define€òòtheóó€òòextentóó€òòofóó€òòtheóó€òòmarketóó.€€Ð ¤*–"& Ðà  àTogether€with€the€MRS,€these€two€types€of€conditions€combine€to€lead€to€measures€of€theÐ v,h$( Ðbenefits€of€a€public€good.€€As€the€next€section€discusses€in€several€cases,€empirical€evaluations€ofÏinformation€or€risk€communication€policies€have€encountered€both€types€of€responses€withoutÏhaving€a€unified€conceptual€framework.€€The€point€to€be€made€here€is€that€education€can€inÏprinciple€also€influence€òòbothóó„„the€choice€over€amounts€of€an€environmental€resource€and€whetherÐ „ v Ðor€not€to€participate€or€be€concerned€about€it.ÌÌÌò òIIIà  àEducation€and€Environmental€ResourcesР̾  ÐÌA.à  àòòBackgroundó óóóÐ tf Ðà  àTo€my€knowledge,€there€have€been€no€attempts,€comparable€in€objectives€to€HavemanÏand€Wolfe€(and€the€recent€update€by€Wolfe),€to€estimate€education's€effects€on€improvingÏenvironmental€resources.€€Because€education€is€frequently€argued€to€be€a€"control"€variableÏreflecting€the€influence€of€taste€and€knowledge€related€effects,€the€empirical€studies€do€not€claimÏa€causal€relationship.× ƒ; ×Ý ƒÿÿÝòòÚ  Ú16Ú  ÚóóÝ  Ý×  ×€€Instead,€they€acknowledge€that€people€with€amounts€of€education€can€beÐ !‚ Ðdifferent€in€their€propensities:€to€undertake€averting€behavior,€to€respond€to€informationÏprograms,€and€to€form€risk€perceptions€for€the€different€types€of€environmental€hazards.Ìà  àTo€evaluate€whether€a€causal€argument€could€be€made€requires€another€look€at€theseÏempirical€findings.€€In€the€first€subsection,€I€provide€that€overview.€€However,€it€is€not€aÏcomprehensive€review.€€Instead,€I€focus€on€four€sets€of€research:€(1)€my€own€work€on€avertingÏbehavior€with€hazardous€waste€risks€and€the€implications€of€information€programs€for€avertingÐ |,n$( Ðbehavior€intended€to€reduce€risks€of€exposure€to€an€indoor€pollutant,€radon;€(2)€Viscusi's€findingsÏconcerning€the€impacts€of€labels€on€people's€precautionary€intentions€along€with€some€of€myÏown€work€on€a€radon€risk€communication€program,€€(3)€a€recent€contingent€valuation€studyÏdescribing€the€effects€of€the€framing€of€risk€information€about€the€risks€of€groundwaterÏcontamination;€and€(4)€the€work€of€Ippolito€and€Mathios€and€Kenkel€concerning€healthÏinformation,€nutrition,€and€behavior€because€the€issues€involved€are€closely€aligned€with€someÏnew€research€reported€later€in€the€section.€ò òÐ úì  ЀÌB.à  àòòAverting€Behavior€and€Information€Policiesó óóóÐ  ’ Ðà  àThe€logic€underlying€these€types€of€models€can€be€described€within€an€HPF€framework.€ÏIndeed,€the€conceptual€arguments€suggesting€that€these€actions€offer€opportunities€to€observeÏpeople's€values€for€environmental€resources€are€widespread€in€the€literature.€€Unfortunately,Ïempirical€tests€using€people's€decisions€and€linking€them€to€environmental€resources€have€beenÏlimited.€€Two€types€of€studies€have€been€used„„reports€of€actual€averting€or€mitigating€behaviorsÏand€stated€choices€in€response€to€a€predefined€situations.€€Each€has€advantages€andÏdisadvantages€for€estimating€people's€marginal€valuations€or€testing€the€role€of€education€inÏenhancing€the€effectiveness€of€public€information€or€risk€communication€programs.Ìà  àStudies€relying€on€actual€reports€necessarily€use€indirect€measures€of€the€policy€and/orÏbehavior€of€interest.€€Nonetheless,€the€responses€studied€are€reports€of€actual€choices.€€ByÏcontrast,€the€stated€choice€studies€often€have€greater€control€over€the€way€the€policy€is€presentedÏand€the€object€of€choice€framed€to€people,€but€this€comes€at€the€cost€of€the€use€of€statementsÐ z,l$( Ðabout€choices€that€òòwouldóó€be€made.Ð  Ðà  àThe€first€study€(Smith€and€Desvousges€[1986a])€involved€both€types€of€responses.€€It€wasÏbased€on€a€survey€of€households€in€suburban€Boston.€€Conducted€during€the€spring€and€summerÏof€1984€using€a€stratified€sample€selected€from€two€strata,€the€town€of€Acton€and€the€rest€ofÏsuburban€Boston,€the€survey€was€designed€to€estimate€people's€willingness€to€pay€for€reductionsÏin€the€risk€of€being€exposed€to€hazardous€waste€(see€Smith€and€Desvousges€[1987]).€€A€numberÏof€different€types€of€actual€and€stated€response€measures€were€collected€from€respondents€duringÏin„person€interviews€lasting€nearly€an€hour€each.€€Acton€was€oversampled€because€its€residentsÏhad€experienced€several€incidents€where€hazardous€wastes€contaminated€their€drinking€waterÏsupplies.Ìà  àOne€aspect€of€the€information€collected€in€the€survey€involved€household's€decisions€toÏundertake€one€of€three€"averting"€activities€during€the€five€years€preceding€the€survey€for€theÏsole€purpose€of€reducing€the€risk€of€exposure€to€hazardous€waste.€€These€activities€were:Ïinstallation€of€water€filters,€purchase€of€bottled€water,€and€attendance€at€public€meetingsÏconcerning€hazardous€waste.€€Nearly€30€percent€of€the€sample€indicated€that€they€purchasedÏbottled€water€regularly€to€avoid€hazardous€waste,€seven€(7)€percent€installed€water€filters€andÏeight€(8)€percent€attended€public€meetings.Ìà  àProbit€estimates€for€each€activity€individually€and€for€a€composite€(designed€to€measureÏwhether€a€person€had€undertaken€any€one€of€the€three€types€of€averting€behavior)€indicated€thatÏattitudes€concerning€the€degree€to€which€hazardous€waste€was€harmful,€experience,€and€recentÏknowledge€of€the€problem€in€the€respondent's€town€as€well€as€proxy€variables€for€overallÐ v,h$( Ðattitudes€toward€risk€and€the€effectiveness€of€public€mitigation€were€significant,€and€consistentÏwith€òòaóó€òòpriorióó€sign€expectations,€as€determinants€of€several€of€these€reported€actions,€whetherÐ à Ò Ðmeasured€individually€or€as€a€composite€variable.€€Education€(measured€as€years€of€schooling)Ïòòwasóó€òònotóó.Ð „ v Ðà  àBy€contrast,€when€the€homeowners€in€the€sample€(approximately€44€percent€of€theÏsample)€were€asked€about€their€willingness€to€move€to€avoid€proximity€to€a€landfill€withÏhazardous€waste,€the€respondent's€years€of€school€was€a€significant,€positive€determinant€of€theirÏdemand€for€distance.€€These€demand€intention€models€included€income,€age,€family€size,€years€atÏthe€address,€as€well€as€experience€and€attitude€measures.€€The€question€providing€the€basis€forÏthe€demand€model€was€framed,€following€the€logic€of€a€hedonic€model€(see€Freeman€[1993])Ïwhere€homeowners€were€told€houses€like€their€current€home€would€vary€in€price€depending€onÏtheir€distance€from€landfills€with€hazardous€waste€(see€Smith€and€Desvousges€[1986b]).× ƒ< ×Ý ƒÿÿÝòòÚ  Ú17Ú  ÚóóÝ  Ý×  ×€€ThisÐ  Ðdifference€in€findings€illustrates€the€complexity€in€linking€specific€reported€behaviors€to€goalsÏthat€might€then€be€associated€with€environmental€resources€and€then€distinguishing€an€additionalÏcontributing€factor,€such€as€education's€role,€in€these€decisions.Ìà  àThere€are€any€number€of€explanations€for€the€limited€nature€of€the€factors€identified€inÏthe€actual€behavior€models.€€The€actions€are€likely€to€be€multiple€purpose€and€we€do€not€knowÏhow€people€interpreted€the€original€question's€request€that€they€report€only€activities€whose€"soleÏpurpose"€was€related€to€reducing€exposure€of€hazardous€waste.€€Recall€of€behavior,€especiallyÏactions€that€are€not€important€or€regular,€can€have€substantial€error;€and€finally,€the€categoricalÏnature€of€our€outcome€measure€limits€the€ability€of€statistical€models€to€capture€the€effects€byÐ v,h$( Ðpotentially€important,€but€subtle€effects.× ƒ= ×Ý ƒÿÿÝòòÚ  Ú18Ú  ÚóóÝ  Ý×  ×€€With€these€caveats,€it€is€not€clear€that€theÐ  Ðinterpretation€of€education€as€enhancing€people's€understanding€of€the€issues€would€have€impliedÏa€clearcut€sign€for€its€implied€effects„„either€positive€or€negative.Ìà  àTo€evaluate€the€benefits€from€education's€role€in€inducing€people€to€take€action,€we€mustÏfirst€judge€whether€the€action€would€be€warranted€by€some€"objective"€standard.€€If€the€exclusiveÏpathway€of€exposure€to€hazardous€waste€is€through€drinking€water€and€the€household€is€servedÏfrom€a€community€source,€proximity€between€one's€house€and€a€landfill€will€òònotóó€influence€theÐ úì  Ðpathway.€€Similarly,€if€experts€judge€a€particular€risk€to€be€small€and€education€has€a€positive,Ïsignificant€effect€on€subjective€risk€perceptions€for€that€source,€we€must€ask€whether€theÏeducational€effect€is€meaningful.× ƒ> ×Ý ƒÿÿÝòòÚ  Ú19Ú  ÚóóÝ  Ý×  ×€€It€may€be€that€education€heightens€awareness€of€the€source€(aÐ pb Ð"good"€outcome),€does€not€influence€perceived€risk€greatly€across€individuals€(a€"neutral"Ïoutcome),€but€is€found€to€be€a€significant€positive€influence€in€empirical€models€(a€seeminglyÏ"bad"€outcome)€because€of€the€type€of€measures€of€risk€perceptions€available.Ìà  àWhere€does€this€leave€my€discussion€of€the€conflicting€results€from€the€role€of€educationÏin€the€first€area?€€I€believe€it€implies€that€if€we€are€seeking€a€clear€interpretation€that€educationÏhas€a€social€benefit€due€to€its€effects€in€enhancing€the€performance€of€environmental€policiesÏdirected€at€private€responses€to€externalities€or€inducing€people€to€undertake€behavior€that€makesÏ"more"€environmental€quality€available,€then€the€cases€must€involve€actions€with€unambiguousÏinterpretation€and€substantial€control€on€how€the€policy€or€activity€influences€the€"claimed"Ïenvironmental€benefits.€Ìà  àUse€of€behavioral€intentions€for€products€labeled€as€having€specific€hazards€defines€bothÐ v,h$( Ðthe€context€and€the€judgment€about€what€is€a€"correct"€outcome€for€each€person.€€To€do€soÏrequires€that€the€analysis€be€based€on€respondents'€statements€about€their€intentions€for€using€thisÏhypothetical€commodity.€€While€this€research€may€seem€far€afield€from€applications€toÏenvironmental€resources,€Viscusi's€research€(see€Viscusi€and€Magat€[1987]€and€Magat€andÏViscusi€[1992])€was€designed€as€part€of€a€larger€effort€to€evaluate€information€programs€inÏimplementing€environmental€policies€(i.e.,€regulating€the€use€of€pesticides,€etc.).Ìà  àA€variety€of€products€and€outcome€measures€were€considered€using€both€computerÏassisted,€mall€intercept,€and€telephone€surveys.€€Two€outcome€measures€are€potentially€importantÏto€evaluating€the€role€of€education:€(a)€recognizing€the€warning€message€and€(b)€being€able€toÏuse€the€information€to€undertake€(or€intend€to€undertake)€the€recommended€mitigating€behavior.€Ìà  àUnder€a€variety€of€controlled€conditions,€Magat€and€Viscusi€[1992]€report€that€educationÏenhances€recognition€and€ability€to€recall€warning€messages€(see€Appendix€H€for€their€specificÏmultivariate€results).€€Open€ended€questions€seeking€information€about€the€warnings€yieldedÏresponses€that€identified€the€precautionary€words€and€the€prospects€respondents€would€rememberÏthose€warnings€as€positively€related€to€education.€€While€these€results€are€related€to€specificÏtypes€of€consumer€products€with€hypothetical€labels,€my€risk€communication€experimentsÏevaluating€brochures€explaining€radon's€risk€confirms€their€findings€(see€Smith€et€al.,€[1988]).€Ìà  àThat€is,€by€exploiting€the€panel€design€of€a€study€intended€to€evaluate€the€effectiveness€ofÏa€risk€communication€program,€this€analysis€was€able€to€consider€whether€different€types€ofÏinformation€materials€influenced€people's€knowledge€about€radon,€its€risk,€and€how€to€mitigateÏits€effects.€€Years€of€schooling€was€a€positive€and€highly€significant€determinant€of€participants'Ð v,h$( Ðscores€on€a€quiz€given€after€they€received€the€risk€information€treatments;€it€positivelyÏcontributed€to€their€learning€(as€measured€by€correctly€answering€questions€previously€answeredÏincorrectly)€and€to€their€"correct"€interpretations€of€the€risks€associated€with€radon€and€theÏmitigation€recommendations€they€would€make.€€While€this€study€used€a€specific€set€of€questionsÏ(and€not€the€often€preferred€open€ended€approach€used€in€Magat€and€Viscusi),€it€involves€realÏrisks€and€controls€over€the€circumstances€in€which€learning€is€measured.Ìà  àThus,€both€of€these€studies€would€support€the€type€of€link€hypothesized€in€the€conceptualÏdiscussion€between€education€and€household€production€functions.€€However,€this€evidenceÏalone€does€not€indicate€whether€years€of€schooling€increases€the€rate€of€mitigation€or€reduces€itsÏcosts€(by€improving€household€production€activities).€€It€simply€suggests€that€people€recognizeÏthe€message.€€Ìà  àIf€we€consider€that€next€step,€mitigation,€the€record€is€mixed€with€some€encouragingÏevidence€on€an€indirect€effect.€€To€some€degree,€the€conclusions€here€overlap€with€the€activitiesÏin€the€health€area€noted€below€with€the€Ippolito„Mathios€and€Kenkel€research.€€For€theseÏapplications,€there€are€fewer€assumptions€required€to€connect€behavior€(stated€or€actual)€toÏeducation.Ìà  àViscusi's€research€with€product€warnings€does€òònotóó€find€a€significant€role€for€education€inÐ .%   Ðenhancing€the€effects€of€a€warning€on€promoting€the€recommended€behavior€(as€a€statedÏintention).€€The€variable,€years€of€schooling,€is€a€positive€contributor€to€the€behavioral€intention,Ïbut€not€significant.€€(Viscusi,€Magat,€and€Huber€[1986]).€€In€a€continuing€evaluation€of€the€panelÏof€households'€decisions,€some€years€after€the€learning€questions€were€posed,€we€(Smith,Ð v,h$( ÐDesvousges,€and€Payne)€investigated€their€reported€mitigation€and€found€no€direct€effect.€€ThatÏis,€with€a€sample€of€853€households€who€had€participated€in€the€original€analysis€cited€earlier,Ïreceived€radon€readings€and€differing€information€materials,€we€found€there€was€no€significantÏseparate€effect€of€education€on€mitigation.€€However,€there€òòwasóó€a€significant€positive€effect€ofÐ „ v Ðrespondents'€knowledge€(measured€by€the€score€of€the€same€individual€on€the€quiz€taken€twoÏyears€prior€to€the€mitigation€survey).€€Because€education€was€a€key€determinant€of€theÏeffectiveness€of€information€programs€in€conveying€the€correct€messages,€these€results€wouldÏsuggest€the€types€of€impacts€outlined€in€the€HPF€model.Ìà  àNonetheless,€it€is€important€to€acknowledge€that€these€effects€need€not€be€uniform€for€allÏpeople.€€That€is,€if€people€already€know€about€radon,€its€risks,€and€the€mitigation€options,€thenÏthe€information€programs€do€not€change€their€knowledge€base.€€They€may€well€choose€not€to€doÏanything€about€it.€€Hence€for€these€people,€education,€by€enhancing€their€effectiveness€in€testÏscores,€but€not€their€choices€(stated€or€actual),€would€not€yield€indirect€benefits.€€This€conclusionÏreinforces€the€distinction€between€the€effects€of€education€on€the€participation€and€the€marginalÏconditions€in€describing€its€indirect€benefits.€€Unfortunately,€the€data€available€do€not€permitÏmeasuring€the€cost€of€the€activities€to€the€people€involved.€€Thus,€an€incremental€tradeoff,€similarÏto€that€suggested€by€Haveman€and€Wolfe,€is€not€possible.€€The€next€two€sets€of€research€wouldÏallow€some€version€of€this€type€of€tradeoff,€but€neither€is€ideal.Ìà  àThe€first€is€a€contingent€valuation€study€focused€on€how€risk€information€was€presentedÏto€respondents.€€The€design€relies€on€the€Smith„Desvousges€[1987]€study€of€people's€willingnessÏto€pay€for€reducing€the€risk€of€exposures€to€hazardous€wastes.€€Using€a€split„sample€design,Ð v,h$( ÐLoomis€and€duVair€[1993]€evaluate€whether€the€method€used€to€explain€a€proposed€riskÏreduction€(i.e.,€a€risk€ladder€versus€pie€charts)€influences€respondents'€willingness€to€purchaseÏplans€(identified€with€different€costs)€to€reduce€risk.€€The€most€interesting€aspect€of€the€findingsÏfor€judging€the€effects€of€education€is€the€òòdifferentialóó€impact€of€education€on€a€stated€behaviorÐ „ v Ð(i.e.,€purchasing€a€program€to€reduce€exposure€risk)€using€the€two€different€informationÏapproaches.€€Education€is€a€highly€significant,€positive€influence€on€the€likelihood€of€supportingÏthe€plan€when€a€more€complex,€compound€lottery€format€(involving€pie€charts)€was€used.€€It€wasÏa€positive,€but€insignificant€influence€when€a€single€risk€measure,€with,€a€risk€ladder,€was€used.€ÏIf€we€interpret€education's€role€as€permitting€increased€understanding€of€the€object€of€choice€inÏthe€complex€case€and€not€necessarily€important€to€the€questions€posed€in€the€other€sub„sample,Ïthen€the€behavioral€link€suggested€by€the€mitigation€results€described€above€would€appear€to€beÏpresent€for€decisions€where€both€prices€and€information€matter.€Ìà  àThe€last€set€of€research€returns€to€actual€behavior€in€a€health€related€context€andÏconsiders€the€effects€of€information€on€continuous€measures€of€activity.€€The€reason€forÏconsidering€it€here€is€to€evaluate€whether€education€would€influence€both€the€decision€toÏundertake€activities€and€the€amount€of€those€activities.× ƒ? ×Ý ƒÿÿÝòòÚ  Ú20Ú  ÚóóÝ  Ý×  ×€€The€results€from€Ippolito€and€MathiosÐ \#N Ð[1990,€1989]€suggest€that€education€may€be€enhancing€the€effect€of€fiber€information€availableÏthrough€cereal€ads€on€the€amount€of€fiber€consumed€as€a€result€of€a€person's€cereal€choice.€€TheirÏanalysis€uses€the€difference€in€these€fiber€consumption€patterns€before€and€during€a€period€whenÏcereal€manufacturers€were€permitted€to€use€fiber€related€health€claims€in€their€advertisements.€ÏEducation€is€a€significant€determinant€of€the€selected€fiber€content€in€cereal€in€each€of€the€twoÐ v,h$( Ðperiods€studied.€€Its€estimated€effect€increased€on€fiber€consumption€for€the€period€with€the€fiber„¼health€claims,€but€these€differences€would€not€be€judged€as€significantly€greater€than€those€forÏthe€period€before€the€advertisements.€€Moreover,€the€results€relate€to€two€òòdifferentóó€samples€ofÐ ² ¤ Ðwomen,€with€somewhat€different€educational€levels€(12.72€years€of€schooling€for€the€sampleÏbefore€the€advertising€and€12.84€after).€€Supporting€evidence€is€also€available€in€their€summaryÏof€the€fiber€consumption€of€women€based€on€the€òòUSDA€Continuing€Survey€of€Food€Intakes€byÐ (  ÐIndividualsóó€in€the€same€years€when€it€is€decomposed€by€education€(Ippolito€and€Mathios€[1989]).€Ð úì  ÐThis€comparison€indicates€significant€differences€in€the€fiber€consumption€between€1985€andÏ1986€for€those€with€some€college€(i.e.,€13„15€years€of€reported€schooling),€but€no€significantÏdifferences€at€other€educational€levels.Ìà  àPerhaps€the€strongest€evidence€of€an€interaction€effect€is€in€Kenkel's€[1991]€study€of€theÏinteraction€effects€of€knowledge€of€the€"harms"€posed€by€smoking€and€drinking€and€the€benefitsÏof€exercise€with€the€level€of€education€as€a€differential€factor€influencing€reported€behavior.€€AsÏin€the€work€associated€radon,€Kenkel€used€respondents'€answers€to€questions€about€each€sourceÏof€health€effect€to€measure€knowledge€and€then€investigates,€using€interaction€terms,€education'sÏadditional€influence€for€each€type€of€behavior€(i.e.,€cigarettes€per€day,€drinks€with€alcoholicÏbeverages€in€the€past€two€weeks,€and€minutes€of€exercise€in€past€two€weeks).€€He€interprets€aÏsignificant€interaction€effect€between€a€knowledge€measure€and€education€as€supporting€aÏdifferential€effect€of€education.€€For€example,€education€enhances€the€impact€of€knowledge€aboutÏthe€ill€effects€of€smoking€and€that€increase€in€impact€is€unlikely€to€be€taste€related.Ìà  àThere€are€two€problems€with€these€results.€€The€first€arises€with€the€interpretation€of€aÐ v,h$( Ðknowledge€variable€as€a€measure€for€what€an€information€program€would€accomplish.€€TheÏsecond€is€recognized€by€Kenkel,€but€not€fully€developed€in€his€results.€€That€is,€it€concerns€theÏseparate€effect€of€education€on€knowledge.€€Education€influences€both€knowledge€and€behaviorÏso€the€interaction€effect€may€simply€reflect€simultaneity€in€Kenkel's€knowledge€and€behaviorÏvariables.€€Because€both€variables€are€based€on€reports€òòafteróó€the€learning€and€the€activity€studiedÐ VH Ðhave€taken€place,€they€may€well€be€jointly€determined.€€à p àÐ (  Ðà  àEqually€important,€Kenkel's€argument€that€an€interaction€effect€of€knowledge€canÏdistinguish€causality€from€taste€assumes€a€fairly€simple€relationship.€€As€the€earlier€workÏsuggests,€there€is€clear€evidence€of€a€link€between€knowledge€and€policies€in€more€controlledÏsettings,€but€it€is€difficult€to€establish€the€"knowledge€equivalent"€of€a€policy.€€Simple€interactionÏeffects€do€not€in€themselves€distinguish€taste€from€the€role€of€education€in€enhancing€theÏeffectiveness€of€an€information€program.€€They€could€reflect€the€need€for€a€more€detailedÏspecification.€€Indeed,€these€interaction€effects€were€significant€and€òòincorrectlyóó€signed€(if€theÐ æØ Ðargument€is€intended€to€suggest€that€they€provide€evidence€of€the€complementary€effects)€withÏtwo€of€the€three€other€activities€using€his€male€sub„sample.€€Simultaneity€between€knowledgeÏlevels€and€the€activities€studied€also€poses€a€potential€problem€in€interpreting€these€interactionÏeffects.€€Unfortunately,€Kenkel€does€not€report€enough€information€to€resolve€matters.€€ModelsÏbased€on€estimates€adjusting€for€simultaneity€are€reported,€but€they€do€not€include€the€interactionÏterms.€€Ìà  àThe€lessons€from€my€selective€review€of€links€between€education€and€people's€behaviorÏin€learning€about€sources€of€environmental€externalities€and€what€the€people€can€do€to€protectÐ v,h$( Ðthemselves€from€these€hazards,€confirms€education's€role€as€enhancing€learning.€€Both€specificÏinformation€(Smith,€et€al.,€[1988])€and€general€reports€(Viscusi,€et€al.,€[1986])€suggest€that€yearsÏof€schooling€improves€the€transmission€of€information€from€brochures€and€labels.Ìà  àEducation€also€seems€to€increase€the€likelihood€of€undertaking€behavior€because€of€thisÏòòenhancementóó€òòinóó€òòknowledgeóó.€€Thus,€the€effect€is€unlikely€to€be€taste€related,€but€it€may€also€beÐ VH Ðpossible€to€accomplish€this€increase€in€the€effectiveness€of€information€programs€in€other€waysÏ(i.e.,€different€information€messages,€as€in€Loomis€and€duVair).€€Moreover,€education's€impactÏon€the€amount€of€effort€devoted€to€mitigation€has€not€been€unambiguously€established€withÏmodels€based€on€actual€behavior.€€At€this€point,€the€empirical€findings€are€suggestive€"hints"Ïabout€an€association.€€To€extend€the€link€beyond€this€point€would€seem€to€require€more€controlÏthan€is€usually€available.€€To€illustrate€whether€such€a€strategy€can€be€successful,€the€next€sectionÏsummarizes€the€findings€from€an€earlier€evaluation€of€a€public€information€program€about€radonÏ(Desvousges,€Smith,€and€Rink€[1992])€and€a€re„analysis€of€those€results€to€determine€whether€theÏcomplementary€effects€of€education€could€be€detected€using€actual€behavior€and€theÏexperimental€controls€that€were€present€as€part€of€an€evaluation€of€that€program.ÌÌò òC.à  àòòSome€New€Evidence€on€Education€and€Averting€Behavioró óóóÐ .%   Ðà  àThis€section€describes€the€results€of€an€extension€to€Desvousges,€et€al.,€[1992]Ïinvestigating€education's€interactive€role€with€information€programs€and€controlling€forÏknowledge€as€well€as€other€economic€and€demographic€effects.€€The€context€of€this€evaluation€isÏa€panel€study€designed€to€evaluate€two€different€risk€communication€programs€intended€toÐ x,j$( Ðincrease€the€likelihood€that€homeowners€would€test€their€residences€for€radon.€€ThreeÏcommunities€in€Maryland€were€selected€for€that€evaluation.€€All€had€high€reported€levels€ofÏradon€and€similar€socio„economic€characteristics.€€Table€2€identifies€the€treatment€and€controlÏcommunities€and€describes€the€characteristics€of€the€two€programs„„an€extensive€program€inÏFrederick€and€a€more€modest€campaign€in€Hagerstown.€€Unfortunately,€the€control€that€can€beÏexercised€in€"real€world"€experiments€is€limited.€€Moreover,€confounding€effects€can€ariseÏindependent€of€experimental€controls.€€Independent€of€our€study,€a€Washington,€D.C.€televisionÏstation€(WJLA)€launched€an€information€campaign€over€the€same€time€period€consisting€ofÏpublic€service€announcements€about€radon€and€a€prime€time€special.€€Because€this€activity€hadÏits€primary€effects€on€Hagerstown€and€Frederick,€òònotóó€our€study's€control€community,€thisÐ pb Ðindependent€effort€confounded€our€ability€to€unambiguously€evaluate€the€impact€of€ourÏcontrolled€program,€but€does€not€affect€the€objective€of€gauging€education's€causal€role€withÏinformation€programs.€€As€Table€2€indicates,€the€WJLA€program€appears€to€have€greatest€effectsÏon€the€community€that€also€received€the€most€extensive€risk€communication€campaign.€€[Table€2Ïhere]Ìà  àBaseline€surveys€were€conducted€in€each€city€prior€to€each€program€(December€1987),€toÏobtain€information€about€attitudes,€knowledge,€prior€radon€testing,€and€demographic€informationÏfor€each€sub„sample.€€A€random€digit€dialed€survey€screened€the€initial€sampled€telephoneÏnumbers€to€be€sure€they€were€associated€with€owner€occupied€dwellings,€and€obtained€a€listingÏof€adult€decision€makers.€€We€randomly€selected€one€of€these€identified€decisions€makers€andÏconducted€a€baseline€and€followup€(after€the€program)€interview€with€that€person.€€This€followupÐ v,h$( Ðinterview€took€place€in€April€1988.× ƒ@ ×Ý ƒÿÿÝòòÚ  Ú21Ú  ÚóóÝ  Ý×  ×€€Table€3Ð  Ðincludes€the€response€rates€for€the€two€surveys€and€the€sample€sizes€for€each€community€andÏinterview.€€[Table€3€here]Ìà  àThe€panel€design€of€the€study,€together€with€the€experimental€variation€in€the€riskÏcommunication€programs,€allows€the€separate€effects€of€education€on€learning€and€thenÏmitigation€to€be€evaluated.€€The€outcome€variables€used€in€this€analysis€are€learning€measuresÏand€mitigation€in€the€form€of€conducting€a€home€radon€test.€€The€later€was€a€key€objective€of€theÏcommunication€program„„to€promote€testing€and€self€protection€where€it€was€needed.€€Ìà  àThe€original€analysis€of€this€experiment€focused€on€attitude€change,€learning,€and€testingÏas€measures€of€the€results€for€each€program.€€There€was€no€attempt€made€to€evaluate€whetherÏeducation€enhanced€either€program's€effectiveness.€€Our€earlier€conclusions€on€the€mitigationÏdecisions€were€that€the€concentrated€media€campaign€(in€Frederick)€did€significantly€increaseÏtesting,€but€that€the€modest€media€campaign€was€not€any€different€than€the€control€community€inÏits€radon€testing.€€Moreover,€the€sensitization€associated€with€the€survey€activity€itself€did€notÏalter€these€general€conclusions.Ìà  àMy€re„analysis€of€these€data€focused€on€distinguishing€the€effects€of€education€throughÏlearning€versus€its€direct€impact€on€program€effectiveness€using€the€new€testing€decision€as€theÏoutcome€measure.€€Tables€3€and€4€report€these€results€based€on€probit€and€ordinary€least€squaresÏmodels€with€the€panel€sample€pooled€across€the€three€communities.€€There€is€clear€evidence€ofÏeducation's€complementary€effect€with€risk€communication€programs€on€people's€home€testingÏdecisions.€€Indeed,€the€initial€analysis€overlooked€an€important€component€of€this€effect€byÐ v,h$( Ðfailing€to€allow€for€this€complementarity.Ìà  àTable€3€considers€the€effects€of€selected€attitude€and€demographic€variables€on€eachÏrespondent's€decision€to€test€his€(or€her)€home€for€radon€òòafteróó€the€risk€communication€program€isÐ ² ¤ Ðcompleted.× ƒA ×Ý ƒÿÿÝòòÚ  Ú22Ú  ÚóóÝ  Ý×  ×€€The€first€column€reports€a€somewhat€different€specification€than€originally€reportedÐ „ v Ðin€Desvousges,€Smith,€and€Rink€[1992],€changing€the€treatment€of€information€available€throughÏtelevision€(the€WJLA€effect)€and€introducing€a€specific€learning€variable,€measured€by€theÏpercentage€improvement€on€a€seven€question€quiz€given€during€the€baseline€interview€and€thenÏagain€in€followup€interview.€€Concerns€about€health,€initial€knowledge,€learning,€and€theÏqualitative€variable€indentifying€the€more€extensive€program€were€all€significant€positiveÏinfluences€on€these€people's€decisions€to€test.€€While€the€learning€effect€was€not€recognizedÏbefore,€measures€for€the€remaining€variables€were€reported€in€that€earlier€analysis.€€EducationÏhas€no€apparent€separate€effect€with€this€model.€€However,€we€never€considered€the€possibility€ofÏa€complementary€effect€with€the€program.€Ìà  àThe€next€two€columns€divide€the€sample€into€those€respondents€with€more€than€highÏschool€and€those€with€high€school€education€or€below.€€Here€we€begin€to€see€the€potential€forÏinteraction€effects.€€The€intensive€information€program€has€a€differential€impact€for€òòonlyóó€thoseÐ \#N Ðrespondents€with€more€than€a€high€school€level€of€schooling.€€This€would€suggest€that€Ìthe€complementary€effects€are€selective,€depending€on€the€level€of€education€and€potentially€theÏtype€of€activity€involved.Ìà  àHowever,€this€is€not€the€end€of€the€story.€€Based€on€the€earlier€conceptual€argument,€asÏwell€as€the€Viscusi,€et€al.,€[1986]€and€Smith,€et€al.,€[1988]€results,€we€would€also€expect€to€see€anÐ v,h$( Ðimpact€of€the€risk€communication€programs€on€learning.€€Under€this€hypothesis,€some€of€theÏeffect€captured€in€the€learning€measure€would€be€due€to€any€indirect€effect€of€the€communicationÏprograms.€€Table€4€tests€this€hypothesis€using€a€regression€analysis€of€the€learning€measure,Ï"improve,"€as€a€function€of€demographic€and€program„related€variables.€€The€first€columnÏreports€the€full€sample.€€The€second€and€third€columns€provide€the€same€two€subsamples€(i.e.,Ïgreater€than€high€school€education€and€high€school€and€below)€as€considered€in€describing€radonÏtesting€decisions.€€[Table€4€here]€€Ìà  àThese€findings€are€important€to€the€earlier€conclusions€(i.e.,€Desvousges,€Smith,€and€RinkÏ[1992]).€€By€failing€to€take€explicit€account€of€learning,€our€evaluation€of€each€programÏoverlooked€the€potential€for€differences€in€the€direct€and€indirect€impacts€with€education.€€TheseÏnew€findings€indicate€a€small€effect€of€the€media„only€program€(Hagerstown)€on€testing€throughÏenhancements€in€people's€learning.Ìà  àMore€specifically,€it€appears€that€the€more€extensive€program€has€its€greatest€effect€onÏlearning€for€those€respondents€with€high€school€and€below.€€The€WJLA€television€based€programÏalso€enhanced€learning€for€this€group.€€Within€this€group,€there€is€also€a€separate€effect€of€yearsÏof€schooling.€€By€contrast,€with€those€households€with€greater€than€high€school€education,€thereÏappears€to€have€been€a€learning€effect€due€to€the€media„only€format.ÌThe€primary€effect€of€the€extensive€program€(Frederick)€takes€place€directly€in€inducing€them€toÏundertake€home€testing,€once€they€already€know€about€radon.ò òÐ Ò(Ä $ ÐÌÐ x,j$( ÐD.à  àòòImplicationsó óóóÐ  Ðà  àThree€cases€were€distinguished€in€developing€a€model€that€would€permit€estimatingÏeducation's€social€value:€€(1)€education€as€a€complementary€public€good€with€other€public€goodsÏ(i.e.,€the€preference€equivalent€of€economies€of€scope);€(2)€education€as€a€private€good€thatÏimproves€the€effectiveness€or€reduces€the€cost€of€programs€intended€to€promote€private€actionsÏmitigating€public€bads;€and€(3)€education€as€a€private€good€that€induces€people€to€pursueÏactivities€that€contribute€to€public€goods.Ìà  àThis€section€concerned€the€empirical€evidence€on€these€types€of€relationships€linkingÏeducation€to€environmental€resources.€€To€my€knowledge,€there€has€been€no€research€on€the€firstÏtype€of€relationship.× ƒB ×Ý ƒÿÿÝòòÚ  Ú23Ú  ÚóóÝ  Ý×  ×€€The€last€requires€an€external,€generally€agreed€set€of€private€activities€thatÐ rd Ðactually€contribute€to€an€environmental€public€good.€€Private€efforts€at€recycling€or€in€theÏdefinition€of€eco„labels€illustrate€the€difficulties€in€making€judgments€about€what€is€in€factÏ"environmentally€good,"€private€behavior€(see€Menell€[1993]).€€To€pursue€this€line€of€researchÏfurther€would€also€require€that€there€be€general€recognition€and€acceptance€that€these€activitiesÏwere€consistent€with€producing€improved€environmental€quality.€€There€did€not€appear€to€beÏexamples€that€would€meet€these€criteria.Ìà  àAs€a€result€of€these€constraints,€the€bulk€of€this€section€was€devoted€to€a€selective€reviewÏof€the€literature€on€averting€and€mitigating€behavior€models€to€gauge€whether€a€complementaryÏrole€for€education€(separate€from€taste€effects)€could€be€established.€€The€available€results€seemÏto€suggest€a€complementary€role€in€learning€tasks.€€The€level€of€education€enhances€theÏeffectiveness€of€information€and€risk€communication€programs€in€increasing€people's€ability€toÐ x,j$( Ðlearn.€€No€separate€complementary€effect€of€education€has€been€detected€for€precautionaryÏactivities€with€either€the€environmental€or€health€issues.€€Part€of€the€reason€for€this€conclusion€(inÏthe€environmental€research)€was€that€few€of€the€studies€were€looking€for€these€types€of€effects.€ÏEducation€was€largely€a€control€variable.€€Distinguishing€the€effects€of€education€as€a€proxy€forÏtaste€versus€as€a€causal€factor€was€not€important€to€the€objectives€of€that€research.€€Moreover,€theÏlevel€of€control€necessary€to€do€so€seems€beyond€most€of€this€past€research.Ìà  àThe€last€part€of€this€section€then€asked„„would€it€be€possible€to€evaluate€the€potential€forÏa€complementary€role€when€there€was€sufficient€control€available€in€the€research€design.€€Here,Ïthe€discussion€offers€one€example„„a€re„examination€of€an€early€risk€communication€program€forÏhomeowners'€decisions€to€test€for€radon.€€The€results€were€surprising.€€In€contrast€with€earlierÏconclusions,€not€only€were€the€effects€of€the€risk€communication€programs€more€detailed€andÏcomplex€than€originally€concluded,€but€the€complementary€role€of€education€seems€to€help€inÏexplaining€them.€€For€this€one€example,€then,€there€appears€to€be€an€effect€of€education€inÏenhancing€learning€òòandóó€one€in€promoting€action€(separate€from€the€learning€effect).€€Ð ¸ª Ðò òÌÌÌÌÌIV.€à  àEnvironmental€Policy€Design€and€EvaluationÐ ®* "& ÐÐ ‚,t$( ÐA.à  àòòSpeculations€on€Education's€Impact€on€the€Environmental€Policyó óóóÐ  Ðà  àTo€the€extent€education€plays€a€role€in€influencing€people's€performance€in€market€andÏnon„market€tasks,€then€recognition€of€these€impacts€should€alter€policy.€€This€implication€isÏsimply€an€extension€of€the€Blinder€and€Rosen's€[1985]€recognition€of€the€importance€ofÏheterogeneity€for€the€design€of€tax€policies€(i.e.,€creating€"notches"€or€budget€discontinuities€toÏtarget€the€distortionary€effect€of€these€policies).Ìà  àTargeting€policies,€based€on€other€criteria,€are€not€new.€€Differential€policies€for€largeÏand€small€firms,€minority€groups,€by€gender,€age,€etc.€have€been€developed€to€achieve€policyÏobjectives€related€to€these€groups.€€Targeting€based€on€educational€levels€has€generally€beenÏregarded€as€more€difficult€because€of€the€transactions€costs€of€identifying€and€separating€theÏgroups.Ìà  àA€better€understanding€of€how€education€serves€as€a€complementary€resource,€enhancingÏthe€effectiveness€of€environmental€and€other€policies,€may€change€this€preconception.€ÏMoreover,€to€the€extent€that€differentiated€programs€already€exist€(or€would€need€to€beÏdeveloped€to€meet€declining€educational€levels€among€large€segments€of€the€population),€thenÏdirect€efforts€to€promote€enhanced€education€offer€a€substitute€for€these€program€adjustments.Ìà  àThere€is€another€way€in€which€policy€design€is€affected€by€education.€€In€a€democraticÏsociety,€to€the€extent€the€public€has€unwarranted€fears€of€any€phenomenon,€such€as€the€risksÏarising€from€exposures€to€hazardous€wastes,€and€policymakers€direct€resources€to€mitigate€theseÏproblems,€(and€away€from€more€serious€hazards€according€to€"the€experts").€€In€these€cases,€oneÏmight€argue€that€education€could€improve€the€allocation€of€public€resources€by€changing€theÐ x,j$( Ðperceived€fears€driving€the€policy€agenda.Ìà  àThis€suggestion€is€not€a€random€speculation.€€The€disparities€between€expert€and€publicÏevaluations€of€environmental€risks€has€been€argued€to€provide€the€reason€why€EPA's€policyÏinitiatives€appear€inconsistent€with€technical€risk€estimates.€€Indeed,€EPA's€Science€AdvisoryÏBoard€identified€this€disparities€in€beliefs€between€experts€and€the€layperson€as€a€central€problemÏfacing€the€agency€in€its€policymaking.× ƒC ×Ý ƒÿÿÝòòÚ  Ú24Ú  ÚóóÝ  Ý×  ×€€Education€is€related€to€subjective€risk€perceptions€andÐ (  Ðthere€is€now€wide€appreciation€in€economics€of€the€importance€of€the€cognitive€problems€thatÏcan€arise€when€people€consider€environmental€risks.€€However,€to€evaluate€the€full€consequencesÏof€this€misperception€requires€the€definition€of€a€standard.€€Someone€must€specify€the€"right€riskÏperception"€for€each€person.Ìà  àSubjective€risk€perceptions€are€the€result€of€both€an€understanding€of€the€process€givingÏrise€to€risk€and€individuals'€preferences€(e.g.,€judgments€of€the€importance€to€that€individual€ofÏwhat€is€at€risk).€€Thus,€we€cannot€be€sure€that€stated€risk€perceptions€are€"mistakes"€due€toÏcognitive€errors€in€interpreting€risk€information€or€simply€reflections€of€preferences.€€To€alter€theÏlatter€as€a€correction€would€abandon€consumer€sovereignty.€€Nonetheless,€because€there€are€veryÏlarge€discrepancies€between€public€and€expert€perceptions€for€important€sources€of€risk,€there€isÏcertainly€scope€for€improved€efficiency€without€necessarily€defining€a€"best"€subjective€riskÏperception€for€each€person.€€Education's€role€in€reducing€simple€cognitive€errors€and€informationÏlimitations€would€no€doubt€lead€to€substantial€gains€before€entering€into€discussions€that€couldÏbe€argued€as€questioning€consumer€sovereignty.€€Unfortunately,€beyond€identifying€the€issuesÏand€empirical€research€on€risk€perception€processes€(e.g.,€Viscusi€[1993],€Smith,€et€al.,€[1990]),Ð v,h$( Ðthere€has€been€no€research€on€how€enhanced€education€might€influence€the€policy€agenda.ÌÌò òB.à  àòòEducation€and€Policy€Evaluationó óóóÐ ² ¤ Ðà  àInformation€is€essential€for€evaluating€any€policy€activity.€€There€has€been€increasingÏrecognition€of€how€education€is€impacting€the€collection€of€information€from€surveys.€€At€theÏsimplest€level,€reading€skill€among€the€U.S.€population€(currently€at€about€the€seventh€gradeÏlevel)€limits€the€methods€used€to€both€convey€information€and€to€collect€it.× ƒD ×Ý ƒÿÿÝòòÚ  Ú25Ú  ÚóóÝ  Ý×  ×€€Thus,€to€the€extentÐ üî  Ðthat€mailed€surveys€are€used€to€collect€information€for€policy€evaluation,€education€has€a€longÏterm€effect€on€the€types€of€information€that€can€be€collected€using€this€format.Ìà  àTo€my€knowledge,€there€is€little€direct€evidence€on€the€effects€of€education€on€responseÏrates€and€survey€administration.€€There€is€a€developing€set€of€literature€indicating€that€theÏcognitive€effort€devoted€to€answering€attitude€and€related€questions€is€influenced€by€educationalÏattainment.€€Krosnick€[1991],€for€example,€suggests€that€the€likelihood€respondents€will€adoptÏ"satisfying"€approaches€to€answering€questions€(i.e.,€offering€answers€that€appear€acceptable,€butÏdo€not€result€from€a€thoughtful€evaluation€of€the€issues€posed)€will€be€inversely€related€to€theÏrespondents'€cognitive€sophistication.€€This€sophistication€is€described€in€his€and€other's€work€asÏan€ensemble€of€abilities€associated€with€retrieving€information€from€memory€and€integrating€itÏinto€verbally€expressed€summary€judgments.€€These€skills€are€related€to€òòbothóó€innate€factors€andÐ 'ô" Ðlearning€experiences.Ìà  àThus,€education€can,€in€principle,€affect€who€participates€in€surveys€designed€to€collectÏinformation€used€in€policy€evaluation€as€well€as€the€quality€of€the€responses€they€provide.€€InÐ x,j$( Ðshort,€it€contributes€to€the€measurement€error€in€the€economic€models€we€use€to€evaluateÏprograms€and€policy€outcomes.€€In€the€long€term,€these€errors€are€influenced€by€the€educationalÏlevel€(and€associated€cognitive€skills)€of€the€population€providing€that€information.€€Thus,€weÏhave€short€term€effects€on€the€design€of€mechanisms€to€collect€information€comparable€in€its€roleÏto€the€design€of€policies€as€responses€to€existing€educational€levels.Ìà  àThere€is€also€a€longer€term€return€to€education€for€both€policy€and€evaluation.€€TheseÏinfluences€may€be€especially€relevant€to€environmental€policies€where€the€issues€involved€areÏcomplex€and€require€detailed€understanding€of€the€interactions€within€ecosystems€and€betweenÏpeople€and€their€environment.€€Educational€constraints€may€well€limit€the€ability€to€conveyÏinformation€about€such€complex€policy€issues.€€A€few€examples€illustrate€the€dilemma.€ÏDecisions€about€biodiversity€and€species€loss,€climate€change,€and€the€long€term€hazards€posedÏby€alternative€management€policies€for€high€level€nuclear€waste€require€extensive€technicalÏinformation€to€appreciate€their€short€and€long€term€implications.€€Informed€choices€cannot€beÏmade€without€understanding€of€the€technical€issues€involved.€€In€these€areas,€one€might€askÏwhether€education€imposes€a€constraint€on€the€ability€to€rely€on€consumer€sovereignty€in€suchÏlarge€scale€choices€(see€Ecosystem€Valuation€Forum€[1992]€and€Weitzman€[1992,€1993]€forÏmore€technical€arguments).ÌÌÌò òV.à  àThe€Environment's€Role€in€Education's€Social€Benefits:€Next€Stepsó óÐ ¤*–"& ÐÐ x,j$( Ðà  àOver€thirty€years€of€empirical€research€on€the€impact€of€education€has€focused€theÏgreatest€attention€on€its€private€returns.€€While€social€returns€to€education€have€been€informallyÏacknowledged€for€most€of€this€time,€systematic€efforts€to€estimate€these€benefits€have€been€moreÏlimited.€€This€overview€of€the€relationship€between€education€and€efforts€to€enhanceÏenvironmental€quality€has€identified€a€more€general€set€of€questions€associated€with€developingÏmeasures€of€education's€social€benefits.€€Past€efforts€to€explain€the€demand€for€education€haveÏused€two€quite€different€specifications€for€the€education€as€an€object€of€choice€€or€economicÏcommodity.€€One€focuses€on€years€of€school€and€each€individual's€choice€to€pursue€increasedÏeducation.€€Within€this€framework,€social€benefits€have€been€defined€as€a€composite€of€theÏeffects€of€education€on€that€individual's€non„market€activities€and€any€private€decisions€thatÏrelate€to€social€objectives.€€These€linkages€are€best€described€in€a€household€productionÏframework€where€education€òòmustóó€òòbeóó€considered€an€input€that€has€a€causal€role€in€improving€theÐ  Ðeffectiveness€of€these€activities.€€The€classification€of€activities€based€on€their€relationship€toÏprivate€services€or€public€goals€determines€whether€this€causal€link€would€imply€a€public€benefit.Ìà  àThe€second€approach€measured€the€object€of€choice€attributed€to€education€as€some€typeÏof€average€level€of€activity€for€students€within€an€educational€system.€€Usually€these€evaluationsÏfocused€on€primary€and€secondary€education,€but€the€same€logic€could€have€been€applied€toÏpublic€higher€education.€€The€only€attempt€to€distinguish€social€from€private€benefits€in€thisÏcontext€was€based€on€the€attributes€of€households€(e.g.€presence€of€children€and€capitalizationÏeffects)€in€comparison€with€those€without€the€possibility€of€these€sources€of€private€gains.Ìà  àUsing€either€approach€to€construct€separate€measures€for€the€social€benefits€of€educationÐ v,h$( Ðrequires€the€analyst€to€compose€what€is€taken€to€be€the€object€of€choice€motivating€people'sÏdecisions.€€This€strategy€is€òòtheóó€accepted€paradigm€for€revealed€preference€approaches€toÐ à Ò Ðeconomic€analysis.€€Results€are€evaluated,€in€part,€by€how€much€can€be€inferred€from€extremelyÏlimited€information.× ƒE ×Ý ƒÿÿÝòòÚ  Ú26Ú  ÚóóÝ  Ý×  ×€€An€alternative€strategy€would€suggest€that€the€analyst€should€begin€byÐ „ v Ðdetermining€how€people€conceive€of€education€as€an€object€of€choice.€€While€there€are€limits€toÏwhat€can€learned€from€this€approach,€the€literature€seems€to€be€in€little€danger€of€reaching€theÏpoint€of€diminishing€returns.€€No€economic€research€has€considered€this€option.€€Ìà  àWhy€might€it€be€important,€given€the€prospect€of€success€with€the€alternatives?€€TheÏreasons€arise€in€the€role€attributed€to€education€and€measures€of€economic€value.€€TheÏconventional€strategy€relying€on€cost€savings€due€to€education's€role€in€improving€households'Ïabilities€to€accomplish€tasks€that€are€related€to€private€or€public€goals€imputes€values€to€each€ofÏthese€savings€by€considering€a€substitute€for€what€was€accomplished€with€education.€€EducationÏis€a€means€to€some€other€objective.€€It€enhances€private€activities€or€improves€public€mechanismsÏ(i.e.,€public€information€or€risk€communication€policies),€but€òònoóó€òòindependentóó€òòroleóó€òòcanóó€òòeveróó€òòbeóóÐ ¸ª Ðòòdetectedóó.€€This€follows€because€maintained€assumptions€are€required€to€link€education€to€theÐ Š!| Ðactivities€in€the€first€place.€€Even€if€people€appreciate€these€connections,€they€may€have€otherÏobjectives€(unrelated€to€private€returns)€for€supporting€public€education.€€Ìà  àThese€more€general€efforts€should€be€regarded€as€requiring€longer€term€researchÏprograms.€€In€the€short€term,€there€does€appear€to€be€a€significant€role€for€education€as€aÏcomplementary€factor€in€improving€the€performance€of€public€information€and€riskÏcommunication€related€to€environmental€resources.€€With€the€increasing€complexity€ofÐ v,h$( Ðenvironmental€resource€problems,€programs€to€communicate€information€about€environmentalÏresources€seem€likely€to€play€a€growing€role€in€future€environmental€policy€design€andÏevaluation.€€Because€of€these€trends,€it€would€be€desirable€to€consider€opportunities€for€shortÏterm€research€insights€from€investigating€links€between€education€and€these€past€policyÏinitiatives.Ìà  àOver€the€past€decade,€the€Environmental€Protection€Agency€sponsored€a€substantialÏprogram€of€research€on€public€information€and€risk€communication.€€These€efforts€involved€job„¼related€risk,€product€labeling,€indoor€radon,€food€safety€(i.e.,€pesticide€residues),€siting€decisionsÏinvolving€undesirable€land€uses,€recycling,€and€others.€€In€some€cases,€the€studies€had€a€specificÏcomponent€designed€to€evaluate€the€effectiveness€of€the€policy€interventions€being€considered.€ÏTo€my€knowledge,€none€of€these€efforts€attempted€to€evaluate€how€education€might€haveÏenhanced€programmatic€objectives€or€if€the€policies€could€be€designed€to€reflect€the€effects€ofÏeducation€on€people's€responses.€€My€re„analysis€of€one€of€these€earlier€research€efforts€indicatesÏevidence€of€a€complementary€role€for€education€"waiting€in€the€responses"€that€was€neverÏobserved€because€the€researchers€did€not€look.€€Equally€important,€there€are€new€initiatives,€suchÏas€the€EPA's€Toxic€Release€Inventory,€that€make€emission€information€generally€available€to€theÏpublic.€€These€programs€could€be€leading€to€changes€in€community€response€activities€that€differÏby€education€levels.€€Such€differences€in€education€may€offer€a€natural€experiment€to€furtherÏinvestigate€these€complementarities,€provided€causal€and€taste€related€effects€can€be€separated.Ìà  àIn€all€of€these€efforts,€it€is€important€to€distinguish€social€from€public€values,€where€theÏlatter€are€assumed€to€arise€from€public€goods.€€There€is€also€a€need€to€distinguish€value€fromÐ v,h$( Ðmonetary€benefit.€€Reconstructions€of€imputed€cost€savings€due€to€measures€of€education'sÏeffects€on€household€or€policy€performance€do€not€mean€that€people€would,€in€fact,€pay€theseÏamounts€for€the€education.€€They€are€an€analyst's€imputation.€€Measuring€willingness€to€pay€fromÏthese€implicit€decisions€requires€that€people€recognize€the€objects€of€choice€in€ways€that€areÏconsistent€with€what€analysts€assume.Ì€à  àÐ (  ÐÐ  úì  ÐÑ7€ <NNdÄXXdé7ÑØJ9ØÓB.°°` ¸ hÀpÈ xÐ (#ÿÿŒXBÓò òà@ìì*ìàˆà0  àà À àReferencesó óÐ(#(# ÐÌà0  àà ° àAndreoni,€James,€1989.€"Giving€with€Impure€Altruism:€Applications€to€Charity€and€RicardianÏEquivalence,"€òòJournal€of€Political€Economyóó,€Vol.€97,€December,€1447„1458.І x(#(# Ðà0  àà ° àBaumol,€William€J.,€John€C.€Panzar,€and€Robert€D.€Willig,€1982.€òòContestable€Markets€and€theÐ XJ ÐTheory€of€Industry€Structureóó€(New€York,€N.Y.:€Harcourt€Brace€Jovanovich,€Inc.).Ð* (#(# ÐBecker,€Gary€S.,€1964.€€òòHuman€Capitalóó€(New€York,€N.Y.:€Columbia€University€Press).Ð üî  Ðà0  àà ° àBehrman,€Jere€R.,€1994.€€"Social€Benefits€of€Education:€Introduction,€Conceptual€andÏMeasurement€Issues€and€Policies,"€€Department€of€Economics,€University€ofÏPennsylvania,€draft€paper,€June€26.Ð (#(# Ðà0  àà ° àBehrman,€Jere€R.€and€Nancy€Birdsall,€1983.€€"The€Quality€of€Schooling:€Quantity€Alone€isÏMisleading,"€òòAmerican€Economic€Reviewóó,€Vol.€73,€December,€928„946.Ð(#(# Ðà0  àà ° àBergstrom,€Theodore,€Lawrence€Blume,€and€Hal€Varian,€1986.€€"On€the€Private€Provision€ofÏPublic€Goods,"€€òòJournal€of€Public€Economicsóó.€€Vol.€29,€25„49.к¬(#(# Ðà0  àà ° àBergstrom,€Theodore€C.€and€R.€Goodman,€1973.€€"Private€Demand€for€Public€Goods,"€òòAmericanÐ Œ!~ ÐEconomic€Reviewóó,€Vol.€63,€June,€286„296.Ð^#P(#(# Ðà0  àà ° àBergstrom,€Theodore€C.,€Daniel€L.€Rubinfeld,€and€Perry€Shapiro,€1982.€€"Micro„BasedÏEstimates€of€Demand€Functions€for€Local€School€Expenditures,"€òòEconometricaóó,€Vol.€50,Ð 'ô" ÐSeptember,€1183„1206.Ð (#(# Ðà0  àà ° àBetts,€Julian€R.,€1993.€"Does€School€Quality€Matter?€Evidence€from€the€National€LongitudinalÏSurvey€of€Youth"€unpublished€paper,€Department€of€Economics,€University€ofÐ x,j$( ÐCalifornia,€San€Diego,€October.Ð (#(# Ðà0  àà ° àBlinder,€Alan€S.€and€Harvey€S.€Rosen,€1985.€"Notches"€òòAmerican€Economic€Reviewóó,€Vol.€75,Ð à Ò ÐSeptember,€736„747.Ð (#(# Ðà0  àà ° àBockstael,€Nancy€E.€and€Catherine€L.€Kling,€1988.€"Valuing€Environmental€Quality:€WeakÏComplementarity€with€Sets€of€Goods,"€òòAmerican€Journal€of€Agricultural€Economicsóó,Ð VH ÐVol.€70,€654„662.Ð (#(# Ðà0  àà ° àBorcherding,€Thomas€E.€and€Robert€T.€Deacon,€1972.€"The€Demand€for€the€Services€of€Non„¼Federal€Governments,"€òòAmerican€Economic€Reviewóó,€Vol.€62,€December,€891„901.Ð̾ (#(# Ðà0  àà ° àBradford,€David€and€George€Hilderbrandt,€1977.€€"Observable€Preferences€for€Public€Goods,ÏòòJournal€of€Public€Economicsóó,€Vol.€8,€111„131.Ðpb(#(# Ðà0  àà ° àCard,€David€S.€and€Alan€B.€Krueger,€1992.€€"Does€Schooling€Quality€Matter?€€Returns€toÏEducation€and€the€Characteristics€of€Public€Schools€in€the€United€States,"€òòJournal€ofÐ  ÐPolitical€Economyóó,€Vol.€100,€February,€1„40.ÐæØ(#(# Ðà0  àà ° àCarson,€Richard€T.,€W.€Michael€Hanemann,€Nicholas€E.€Flores,€1994.€"On€the€Creation€andÏDestruction€of€Public€Goods:€The€Matter€of€Sequencing,"€Department€of€Economics,ÏUniversity€of€California,€San€Diego,€unpublished€paper,€March.Ð (#(# Ðà0  àà ° àCarson,€Richard€T.,€and€Robert€Cameron€Mitchell,€forthcoming,€"Sequencing€and€Nesting€inÏContingent€Valuation€Surveys"€òòJournal€of€Environmental€Economics€and€ManagementÐ 'ò" Ðóó(in€press).ÐÒ(Ä $(#(# Ðà0  àà ° àCornes,€Richard,€1993.€€"Dyke€Maintenance€and€Other€Stories:€Some€Neglected€Types€of€PublicÏGoods,"€òòQuarterly€Journal€of€Economicsóó,€Vol.€108,€February,€259„272.Ðv,h$((#(# Ðà0  àà ° àCornes,€Richard€and€Todd€Sandler,€1986.€€òòThe€Theory€of€Externalities,€Public€Goods,€and€ClubÐ  ÐGoods,€óó(Cambridge,€England:€Cambridge€University€Press).Ðà Ò(#(# Ðà0  àà ° àà0(#(#àà ° àCornes,€Richard€and€Todd€Sandler,€1994.€€"The€Comparative€Static€Properties€of€the€ImpureÏPublic€Good€Model"€€òòJournal€of€Public€Economicsóó,€Vol.€54,€July,€403„422.Є v(#(# Ðà0  àà ° àCropper,€Maureen€L.,€1994.€"Comments€on€Estimating€the€Demand€for€Public€Goods:€TheÏCollective€Choice€and€Contingent€Valuation€Approaches,"€unpublished€paper,€The€WorldÏBank,€Washington,€D.C.Ð (#(# Ðà0  àà ° àDesvousges,€William€H.,€V.€Kerry€Smith,€and€Hillary€H.€Rink€III,€1989.€€òòCommunicating€RadonР̾  ÐRisk€Effectively:€Radon€Testing€in€Marylandóó,€final€report€to€U.S.€EnvironmentalÐ ž ÐProtection€Agency,€EPA„230„03„89„048,€March.Ð (#(# Ðà0  àà ° àDesvousges,€William€H.,€V.€Kerry€Smith,€and€Hillary€H.€Rink€III,€1992.€€"CommunicatingÏRadon€Risks€Effectively:€The€Maryland€Experience"€òòJournal€of€Public€Policy€andÐ  ÐMarketingóó,€Vol.€11,€Fall,€68„78.ÐæØ(#(# Ðà0  àà ° àà0(#(#àà ° àDiamond,€Peter€A.,€Jerry€A.€Hausman,€Gregory€K.€Leonard,€and€Mike€A.€Denning,€1993.€"DoesÏContingent€Valuation€Measure€Preferences:€Experimental€Evidence"€in€òòContingentÐ Š!| ÐValuation:€A€Critical€Assessmentóó€(Amsterdam:€North€Holland).Ð\#N(#(# Ðà0  àà ° àEcosystem€Valuation€Forum,€1992.€Report€to€U.S.€Environmental€Protection€AgencyÏ(Washington,€D.C.€Conservation€Foundation).Ð (#(# Ðà0  àà ° àErlich,€I.,€and€G.€Becker,€1972.€€Market€insurance,€self„insurance€and€self„protection.€€òòJournal€Ð Ò(Ä $ Ðof€Political€Economy,€óóVol.€80,€623„648.Ф*–"&(#(# Ðà0  àà ° àFreeman,€A.€Myrick€III,€1993.€€"Nonuse€Values€in€Natural€Resource€Damage€Assessment,"€inÐ v,h$( ÐòòValuing€Natural€Assets:€The€Economics€of€Natural€Resource€Damage€Assessmentóó,€editedÐ  Ðby€R.€J.€Kopp€and€V.K.€Smith€(Washington,€D.C.:€Resources€for€the€Future).Ð (#(# Ðà0  àà ° àFuchs,€Victor€R.,€1982.€ð ðTime€Preference€and€Health:€€An€Exploratory€Study,ðð€in€òòEconomicÐ ² ¤ ÐAspects€of€Healthóó,€edited€by€V.R.€Fuchs€(Chicago,€Ill.:€€University€of€Chicago€Press).Є v(#(# Ðà0  àà ° àà0(#(#àà ° àGertler,€Paul€and€Paul€Glewwe,€1990.€€"The€Willingness€to€Pay€for€Education€in€DevelopingÏCountries,"€òòJournal€of€Public€Economicsóó,€Vol.€42,€251„275.Ð( (#(# Ðà0  àà ° àGramlich,€Edward€M.€and€Daniel€L.€Rubinfeld,€1982.€€"Micro€Estimates€of€Public€SpendingÏDemand€Functions€and€Tests€of€the€Tiebout€and€Median€Voter€Hypotheses,"€òòJournal€ofР̾  ÐPolitical€Economyóó,€Vol.€90,€June,€536„561.О(#(# Ðà0  àà ° àGrossman,€Gene€M.€and€Alan€B.€Krueger,€1991.€€"Environmental€Impacts€of€a€North€AmericanÏFree€Trade€Agreement."€€NBER€Working€Paper€No.€3914,€National€Bureau€of€EconomicÏResearch,€November.Ð (#(# Ðà0  àà ° àHarrison,€Glenn€W.€and€Jack€Hirshleifer,€1989.€€"An€Experimental€Evaluation€of€WeakestÏLink/Best€Shot€Models€of€Public€Goods"€€òòJournal€of€Political€Economyóó,€Vol.€97,Ð ¸ª ÐFebruary,€201„226.Ð (#(# Ðà0  àà ° àHausman,€J.A.,€editor,€1993.€€òòContingent€Valuation:€A€Critical€Assessmentóó,€(Amsterdam:€NorthÐ \#N ÐHolland).Ð (#(# Ðà0  àà ° àHaveman,€Robert€H.€and€Barbara€L.€Wolfe,€1984.€€"Schooling€and€Economic€Well„Being:€TheÏRole€of€Non„Market€Effects,"€òòJournal€of€Human€Resourcesóó,€Vol.€19,€No.€3.,€408„429.ÐÒ(Ä $(#(# Ðà0  àà ° àHirshleifer,€Jack.€1983.€"From€Weakest„Link€to€Best€Shot:€The€Voluntary€Provision€of€PublicÏGoods."€òòPublic€Choice,€óóVol.€41€(3):371„86€and€correction€in€òòPublic€Choiceóó€(1985).Ðv,h$((#(# Ðà0  àà ° àHoehn,€John,€1991.€"Valuing€the€Multidimensional€Impacts€of€Environmental€Policy:€TheoryÏand€Methods,"€òòAmerican€Journal€of€Agricultural€Economicsóó,€Vol.€73,€289„299.Ðà Ò(#(# Ðà0  àà ° àHoehn,€John€P.€and€John€B.€Loomis,€1993.€€"Substitution€Effects€in€the€Valuation€of€MultipleÏEnvironmenal€Programs"€òòJournal€of€Environmental€Economics€and€Managementóó,€Vol.Ð „ v Ð25,€January,€56„75.Ð (#(# Ðà0  àà ° àHuang,€Ju€Chin€and€V.€Kerry€Smith,€1995.€€"Weak€Complementarity€and€Quasi„Rents,"Ïunpublished€paper,€Center€for€Environmental€and€Resource€Economics,€Duke€University,ÏOctober.Ð (#(# Ðà0  àà ° àIppolito,€Pauline€M.€and€Alan€D.€Mathios,€1989.€òòHealth€Claims€in€Advertising€and€Labeling:€AÐ ž ÐStudy€of€the€Cereal€Market€óó(Washington,€D.C.:€€Federal€Trade€Commission),€August.Ðpb(#(# Ðà0  àà ° àIppolito,€Pauline€M.€and€Alan€D.€Mathios,€1990.€"Information,€Advertising,€and€Health€Choices:ÏA€Study€of€the€Cereal€Market,"€òòRand€Journal€of€Economicsóó,€Vol.€21,€Autumn,€459„480.Ð(#(# Ðà0  àà ° àJimenez,€Emmanual,€1994.€€"Human€and€Physical€Infrastructure:€Public€Investment€and€PricingÏPolicies€in€Developing€Countries"€in€òòHandbook€of€Development€Economicsóó,€Vol.€3,Ð ¸ª Ðedited€by€J.€Behrman€and€T.N.€Srinivasan€(Amsterdam:€North€Holland,€in€press).Ð (#(# Ðà0  àà ° àKahneman,€David€and€Jack€Knetsch,€1992.€"Valuing€Public€Goods:€The€Purchase€of€MoralÏSatisfaction"€òòJournal€of€Environmental€Economics€and€Managementóó,€Vol.€22,€January,Ð .%   Ð55„70.Ð (#(# Ðà0  àà ° àKemp,€Michael€A.€and€Christopher€Maxwell,€1993.€€"Exploring€A€Budget€Constraint€forÏContingent€Valuation€Estimates"€in€J.A.€Hausman,€editor,€òòContingent€Valuation:€AÐ ¤*–"& ÐCritical€Assessmentóó,€(Amsterdam:€North€Holland).Ðv,h$((#(# Ðà0  àà ° àKenkel,€Donald€S.,€1991.€€"Health€Behavior,€Health€Knowledge,€and€Schooling"€€òòJournal€ofÐ  ÐPolitical€Economyóó,€Vol.€99,€April,€287„305.Ðà Ò(#(# Ðà0  àà ° àKirsch,€Irwin,€Ann€Jungeblut,€Lynn€Jenkins,€and€Andrew€Kolstad,€1993.€€òòAdult€Literacy€inÐ ² ¤ ÐAmericaóó,€(Washington,€D.C.:€Office€of€Educational€Research€and€Improvement,Ð „ v ÐDepartment€of€Education,€September.Ð (#(# Ðà0  àà ° àKolhase,€Janet.,€1991.€€"The€Impact€of€Toxic€Waste€Sites€on€Housing€Values,"€òòJournal€of€UrbanÐ (  ÐEconomicsóó,€Vol.€30,€January,€1„26.Ðúì (#(# Ðà0  àà ° àKrosnick,€Jon€A.,€1991.€"Response€Strategies€for€Coping€with€the€Cognitive€Demands€ofÏAttitude€Measures€in€Surveys,"€òòApplied€Cognitive€Psychology,óó€Vol.€3,€213„236.О(#(# Ðà0  àà ° àLoomis,€John€B.€and€Pierre€J.€DuVair,€1993.€"Evaluating€the€Effect€of€Alternative€RiskÏCommunication€in€Devices€of€Willingness€to€Pay:€Results€from€a€Dichotomous€ChoiceÏContingent€Valuation€Experiment,"€òòLand€Economicsóó,€Vol.€89,€August,€289„298.Ð(#(# Ðà0  àà ° àLutter,€Randall€and€John€F.€Morrall€III,€1994.€€"Health„Health€Analysis:€A€New€Way€to€EvaluateÏHealth€and€Safety€Regulation,"€€òòJournal€of€Risk€and€Uncertaintyóó,€Vol.€8,€€€January,€43„Ð ¸ª Ð66.Ð (#(# Ðà0  àà ° àMachina,€Mark,€1990.€€"Choice€Under€Uncertainty:€Problems€Solved€and€Unsolved"€in€R.B.ÏHammond€and€Rob€Coppock,€eds.,€òòValuing€Health€Risks:€Costs€and€Benefits€forÐ .%   ÐEnvironmental€Decision€Makingóó€(Washington,€D.C.:€National€Academy€Press).Ð'ò"(#(# Ðà0  àà ° àMadden,€Paul,€1991.€€"A€Generalization€of€Hicksian€Substitutes€and€Complements€withÏApplication€to€Demand€Rationing,"€òòEconometricaóó,€Vol.€59,€November,€1497„1508.Ф*–"&(#(# Ðà0  àà ° àMagat,€Wesley€A.€and€W.€Kip€Viscusi,€1992.€€òòInformational€Approaches€to€RegulationóóÐ v,h$( Ð(Cambridge,€Mass:€MIT€Press).Ð (#(# Ðà0  àà ° àMler,€Karl€Gran,€1974.€€òòEnvironmental€Economics:€A€Theoretical€Inquiryóó€(Baltimore,€Md.:Ð à Ò ÐJohns€Hopkins€University).Ð (#(# Ðà0  àà ° àMartinez„Vazquex,€Jorge,€1981,€"Selfishness€versus€Public€'Regardiness'€in€Voting€Behavior"€inÏòòJournal€of€Public€Economicsóó,€Vol.€15,€June,€349„362.ÐVH(#(# Ðà0  àà ° àMcConnell,€K.E.,€1983.€"Existence€and€Bequest€Value,"€in€òòManaging€Air€Quality€and€Scenic€Ð (  ÐResources€at€National€Parks€and€Wilderness€Areasóó,€edited€by€R.€D.€Rowe€and€L.G.Ð úì  ÐChestnut€(Boulder,€Co.:€Westview€Press).Ð (#(# Ðà0  àà ° àMenell,€Peter€S.,€1993.€"Eco„Information€Policy:€A€Comparative€Institutional€Perspective"ÏWorking€Paper€No.€104,€Stanford€Law€School,€John€M.€Olin€Program€in€Law€andÏEconomics,€April.Ð (#(# Ðà0  àà ° àMichaels,€R.€Gregory€and€V.€Kerry€Smith,€1990.€"Market€Segmentation€and€Valuing€AmenitiesÏwith€Hedonic€Models:€The€Case€of€Hazardous€Waste€Sites,"€òòJournal€of€UrbanÐ æØ ÐEconomicsóó,€Vol.€28,€September,€223„242.иª(#(# Ðà0  àà ° àMincer,€Jacob,€1974.€òòSchooling,€Experience,€and€Earningsóó€(New€York:€Columbia€UniversityÐ Š!| ÐPress).Ð (#(# Ðà0  àà ° àMitchell,€Robert€Cameron,€and€Richard€T.€Carson,€1989.€€òòUsing€Surveys€to€Value€Public€Goods:Ð .%   ÐThe€Contingent€Valuation€Methodóó€(Washington,€D.C.:€Resources€for€the€Future).Ð'ò"(#(# ÐNational€Research€Council,€1989,€òòImproving€Risk€Communicationóó€(Washington,€D.C.:€National€Ð Ò(Ä $ Ðà  àAcademy€Press).Ìà0  àà ° àOates,€Wallace€E.,€1994,€"Estimating€the€Demand€for€Public€Goods:€The€Collective€Choice€andÐ v,h$( ÐContingent€Valuation€Approaches,"€unpublished€paper,€Department€of€Economics,ÏUniversity€of€Maryland,€January.Ð (#(# Ðà0  àà ° àPlourde,€Charles,€1975.€"Conservation€of€Extinguishable€Species,"€òòNatural€Resources€Journalóó,Ð ² ¤ ÐVol.€15,€October,€791„797.Ð (#(# Ðà0  àà ° àPoe,€Gregory€L.€and€Richard€C.€Bishop,€1993.€"Prior€Information,€General€Information,€andÏSpecific€Information€in€the€Contingent€Valuation€of€Environmental€Risks:€The€Case€ofÏNitrates€in€Groundwater,"€Working€Paper€SP93„11,€Department€of€AgriculturalÏEconomics,€Cornell€University,€July.Ð (#(# ÐRidker,€Ronald€G.,€1967.€€òòEconomic€Costs€of€Air€Pollutionóó€(New€York:€Praeger€Publishers).Ð ž Ðà0  àà ° àSandler,€Todd,€1992.€€òòCollective€Action:€Theory€and€Applicationóó€(Ann€Arbor:€University€ofÐ pb ÐMichigan€Press).Ð (#(# Ðà0  àà ° àSlovic,€Paul,€Baruch€Fischoff,€and€Sarah€Lichtenstein,€1985.€"Regulation€of€Risk:€AÏPsychological€Perspective"€in€òòRegulatory€Policy€and€the€Social€Sciencesóó,€edited€by€R.€G.Ð æØ ÐNoel€(Berkeley,€Calif.:€University€of€California€Press).Ð (#(# Ðà0  àà ° àSmith,€V.€Kerry,€1991a.€"Household€Production€Functions€and€Environmental€BenefitÏEstimation"€in€J.B.€Braden€and€C.D.€Kolstad,€eds.,€òòMeasuring€the€Demand€forÐ \#N ÐEnvironmental€Qualityóó€(Amsterdam:€North€Holland).Ð.%  (#(# Ðà0  àà ° àSmith,€V.€Kerry,€1991b.€"Environmental€Risk€Perception€and€Valuation:€Conventional€versusÏProspective€Reference€Theory,"€in€D.W.€Bromley€and€K.€Segerson,€eds.,€òòThe€SocialÐ Ò(Ä $ ÐResponse€to€Environmental€Riskóó€(Boston:€Kluwer€Academic€Publishers).Ф*–"&(#(# Ðà0  àà ° àSmith,€V.€Kerry,€1992.€€"Arbitrary€Values,€Good€Causes,€and€Premature€Verdicts,"€òòJournal€ofÐ v,h$( ÐEnvironmental€Economics€and€Managementóó,€Vol.€22,€January,€71„89.Ð(#(# Ðà0  àà ° àSmith,€V.€Kerry,€1993a.€€"Nonmarket€Valuation€of€Environmental€Resources:€An€InterpretiveÏAppraisal,"€òòLand€Economicsóó,€February,€1„26.в ¤(#(# Ðà0  àà ° àSmith,€V.€Kerry€and€William€H.€Desvousges,€1986a.€"Averting€Behavior:€Does€it€Exist?"ÏòòEconomic€Lettersóó,€Vol.€20,€291„296.ÐVH(#(# Ðà0  àà ° àSmith,€V.€Kerry€and€William€H.€Desvousges,€1986b.€"The€Value€of€Avoiding€a€LULU:ÏHazardous€Waste€Disposal€Sites,"€òòReview€of€Economics€and€Statisticsóó,€Vol.€68,€May,Ð úì  Ð293„299.Ð (#(# Ðà0  àà ° àSmith,€V.€Kerry€and€William€H.€Desvousges,€1987.€"An€Empirical€Analysis€of€the€EconomicÏValue€of€Risk€Changes,"€òòJournal€of€Political€Economyóó,€Vol.95,€February,€89„114.Ðpb(#(# Ðà0  àà ° àSmith,€V.€Kerry,€William€H.€Desvousges,€Ann€Fisher,€and€F.€Reed€Johnson,€1988.€"LearningÏAbout€Radon's€Risk,"€òòJournal€of€Risk€and€Uncertaintyóó,€Vol.€1,€June,€233„258.Ð(#(# Ðà0  àà ° àSmith,€V.€Kerry,€William€H.€Desvousges,€F.€Reed€Johnson,€and€Ann€Fisher,€1990.€"Can€PublicÏInformation€Programs€Affect€Risk€Perception"€òòJournal€of€Policy€Analysis€andÐ ¸ª ÐManagementóó,€Vol.€9,€January,€41„59.Њ!|(#(# Ðà0  àà ° àSmith,€V.€Kerry,€William€H.€Desvousges,€and€John€W.€Payne,€1995.€€"Do€Risk€InformationÏPrograms€Promote€Mitigating€Behavior?"€€òòJournal€of€Risk€and€Uncertaintyóó,€Vol.€10,Ð .%   ÐMay,€203„221.Ð (#(# Ðà0  àà ° àSmith,€V.€Kerry,€Donald€J.€Epp,€and€Kurt€A.€Schwabe,€1994.€€"Cross€Country€Analyses€Don'tÏEstimate€Health„Health€Responses,"€òòJournal€of€Risk€and€Uncertaintyóó,€Vol.€8,€January,Ð ¤*–"& Ð67„84.Ðv,h$((#(# Ðà0  àà ° àSmith,€V.€Kerry€and€Ju€Chin€Huang,€1995.€"Can€Markets€Value€Air€Quality?€A€Meta€Analysis€ofÏHedonic€Property€Value€Models"€òòJournal€of€Political€Economyóó,€Vol.€103,€February,€209„Ð à Ò Ð227.Ð (#(# Ðà0  àà ° àThe€World€Bank,€1992,€òòWorld€Development€Report€1992:€€Development€and€the€EnvironmentóóÐ „ v Ð(New€York:€€Oxford€University€Press,€1992).Ð (#(# Ðà0  àà ° àTiebout,€Charles,€1956,€ð ðA€Pure€Theory€of€Local€Expenditures,ðð€òòJournal€of€Political€Economyóó,Ð (  ÐVol.€64,€October,€416„424.Ð (#(# Ðà0  àà ° àUnited€Nations€Environment€Programme,€1992,òò€Urban€Air€Pollution€in€the€Megacities€of€theР̾  ÐWorld€óó(Oxford,€U.K.:€Blackwell€Publishers,€1992).О(#(# Ðà0  àà ° àViscusi,€W.€Kip,€1993.€€òòFatal€Tradeoffs:€Public€and€Private€Responsibilities€for€Risk€óó(NewÐ pb ÐYork,€N.Y.:€€Oxford€University€Press).Ð (#(# Ðà0  àà ° àViscusi,€W.€Kip€and€Wesley€A.€Magat,€1987.€òòLearning€About€Risk:€Consumer€and€WorkerÐ  ÐResponses€to€Hazard€Warningsóó€(Cambridge,€Mass.:€Harvard€University€Press).ÐæØ(#(# Ðà0  àà ° àViscusi,€W.€Kip,€Wesley€A.€Magat,€and€Joel€Huber,€1986.€€"Informational€Regulation€ofÏConsumer€Health€Risks:€An€Empirical€Evaluation€of€Hazard€Warnings,"€òòRand€Journal€ofÐ Š!| ÐEconomicsóó,€Vol.€17,€Autumn,€351„365.Ð\#N(#(# Ðà0  àà ° àWeitzman,€Martin€L.,€1993,€"What€to€Preserve?€€An€Application€of€Diversity€Theory€to€CraneÏConservation,"€òòQuarterly€Journal€of€Economicsóó,€Vol.€108,€February,€157„184.Ð'ò"(#(# Ðà0  àà ° àWeitzman,€Martin€L.,€1992.€"On€Diversity",€òòQuarterly€Journal€of€Economicsóó,€Vol.€107,€363„Ð Ò(Ä $ Ð406.Ð (#(# Ðà0  àà ° àWestat,€Inc.,€1989.€òòInvestigation€of€Possible€Recall€Reference€Period€Bias€in€National€Surveys€ofÐ v,h$( ÐFishing,€Hunting€and€Wildlife€Associated€Recreationóó,€Final€Report€to€U.S.€Department€ofÐ  ÐInterior,€December.€Ð (#(# ÐWildavsky,€Aaron,€1988.€€òòSearching€for€Safetyóó.€(New€Brunswick:€Transaction€Books.)Ð ² ¤ Ðà0  àà ° àWolfe,€Barbara€L.€with€Sam€Zuvekas,€undated,€"NonMarket€Outcomes€of€Schooling,"Ïunpublished€paper,€University€of€Wisconsin,€Madison.Ð (#(# Ðà0  àà ° àWyckoff,€James€H.,€1984.€€"The€Nonexcludable€Publicness€of€Primary€and€Secondary€PublicÏEducation,"€òòJournal€of€Public€Economicsóó,€Vol.€24,€August,€331„352.ÓÓÐúì (#(# ÐÐ  ãÕ  ÐÑÑÑTROèØ'33Ø'Ø'3$Letter (Landscape)ÿ3Ø'TÑÑ7€ <NNdÄNNdÄ7ÑØDUØÒ„°ÒÒ„°ÒÓT@„„Ü4Œ ä <”ìDœôL¤ü!T$¬&)\+´- 0°°XTÓà@ÔÔ"ÈàÔ‡Nô‰NXX-³ÔÔ‡iNµƒNNNô‰ÔTable€1.€€Overview€of€Modeling€ImplicationsˆÐ  ÐÌÓ zSÓÌÔ*qƒjk ddd Xdd Xdd X„ 0„ 0qÔÔ,dd@ÔÔ,± dd@ÔÔ,Y dd@ÔÔ,E dd@ÔÔ,2 dd@ÔÔ+  ÔÐ 3€gg"  ö €J€‚ˆ‚ƒƒ„C„3ÐÓ$„,„Ü4„„“X$ÓÓ zSÓÌÌFrameworkÓ ‘…ÓÐ ;€gg*T FCgg „C„;ÐÓ'„,„Ü4Œ „,ƒX'ÓÓ ‘…ÓÌPotential€TheoreticalÌRestrictionsÓ ‘…ÓÐ ;€gg*T F Cgg „C„;ÐEducation€Causal€orÓ ‘…ÓÌSource€of€UnobservedÌHeterogeneityÓ ‘…ÓÐ ;€gg*T F Cgg „C„;ÐÓ ‘…ÓÌMeasurement€IssuesÓ ‘…ÓÐ ;€gg*p bCgg „C„;ÐTypes€ofÓ ‘…ÓÌEnvironmental€ApplicationÓ ‘…ÓÐ LB*p bC ˆ J gg €¨€‚x‚ƒƒ„C„LÐÓ$„,„Ü4„,„X$ÓModel€IÓ ‘…ÓÌ(Simple)Ó ‘…ÓÐ '¤–C „C„'ÐÓ'„,„Ü4Œ „,ƒX'ÓWeak€Complementarity;Ó ‘…ÓÌPotential€for€cascading€Ïchain€of€linkages;€GroupÏlinkage€across€public€goods€Ó ‘…ÓÐ 'l^ C „C„'ÐDifficult€to€establishÓ ‘…ÓÌcausality€withÌtheoretical€restrictionsÌin€this€format€and€onlyÌRevealed€PreferenceÌData€(RPD)Ó ‘…ÓÐ '4& C „C„'ÐMeasuring€amount€ofÏeducation„Years€ofÏschoolings€versus€averageÏlevel€ofÓ ‘…ÓÌoverall€education;ÌAllocation€of€"Joint"ÌbenefitsÌÓ ‘…ÓÐ 'üî $C „C„'ÐUnique€rareÓ ‘…ÓÌenvironmental€resources€inÏgroups€withÏEducation;examplesÏinclude€Bidiversity,Ïhistoric€and€Ìcultural€resourcesÓ ‘…Óñ ñÐ?5 +C x ¨ €Z €‚x‚ƒƒ„C„?Ðñ ññ ñÐ ?5 +C x ¨ €Z €‚x‚ƒƒ„C„?Ðñ ñÓ$„,„Ü4„,„X$ÓModel€IIÓ ‘…ÓÌHouseholdÌProductionÌFunctionsÌ(HPF)Ó ‘…ÓÐ ' C „C„'ÐÓ'„,„Ü4Œ „,ƒX'ÓGroup€privateÓ ‘…ÓÌcommodities€based€onÌòòa€priorióó€restrictions€due€toÐ N @ Ðdefined€activities;Ìassumes€knowledge€ofÌa€process;€Links€educationÏto€activities€based€onÏassumed€òòa€Priorióó€knowledgeÓ ‘…ÓÐ '´ C „C„'ÐLikely€to€defineÓ ‘…ÓÌrationale€for€causality€withÏprior€HPFÌrestrictions€and€RPDÓ ‘…ÓÐ '2 $C „C„'ÐEstimating€HPF€functions;ÏTimeÓ ‘…ÓÌallocation€measures€ofÏpollution;€Jointness€inÏeducation's€role€across€HPFÏas€a€type€of€public€input;Ïand€may€be€Jointness€inÏother€inputs€confoundingÏrole€of€educationÌÓ ‘…ÓÐ 'Š|C „C„'ÐMitigating€exposure€toÓ ‘…ÓÌpollution;€recreation;Ìself€protection€andÌrisk€perceptionÓ ‘…ÓÐ ?52 $C x Z €€‚x‚ƒƒ„C„?ÐÓ$„,„Ü4„,„X$ÓModel€IIIÓ ‘…ÓÌPublic€GoodÌProductionÌby€Individualsñ ññ ñÓzaÓñ ññ ñÓ ‘…ÓÐ '’„ #C „C„'ÐÓ'„,„Ü4Œ „,ƒX'ÓStructure€of€Hirshleifer'sÏSocial€CompositionÏFunction€(SCF)Ð '®  &C „C„'ÐComparable€to€simpleÏmodelÐ 'ʼ (C „C„'ÐSubstantial€data€issues;ñ ññ ñÓzaÓñ ññ ñÓ ‘…ÓÌMeasuring€behavioralÌActivities;€EstimatingÌSCFñ ññ ñÓzaÓñ ññ ñÓ ‘…ÓÐ '’„ ,C „C„'ÐGreen€labeling;ñ ññ ñÓzaÓñ ññ ñÓ ‘…ÓÌEnvironmental€altruismñ ññ ñÓzaÓñ ññ ñÓ 2ÓÐ-#!ʼ .C x   -ÐÓT@„„Ü4Œ ä <”ìDœôL¤ü!T$¬&)\+´- 0„,„XTÓÔ#†iXa©!XNiNµƒKò#ÔÔ‡iNµƒN!XiXa©ÔÔ#†Nô‰NNiNµƒjò#Ôñ ññ ñÔ‡N <NNNô‰ÔÔ†ÿÿÔñ ññ ñÔ#†X-³XNNô‰Å#Ôñ ññ ñÔ  ÔÔ  Ôñ ññ ññ ñÌÐ  Œ~/ Ðñ ññ ñÑ7€ <NNdÄNNdÄ7ÑÑX°ÑÑX°ÑØEFØÓT@„„Ü4Œ ä <”ìDœôL¤ü!T$¬&)\+´- 0„„“XTÓà@;;ÈàTable€2.€€Design€Features€of€Maryland€Public€Information€ProgramˆÐ ¶ ÐÌÓ gR2ÓÌÔ*“ƒmn dddd@± dd@Y dd@E dd@2 dd@jk„ 0„ 0“ÔÔ,Œdd@ÔÔ,3dd@ÔÔ,°dd@ÔÔ,Tdd@ÔÔ,Tdd@ÔÔ,(dd@ÔÔ,·dd@ÔÔ,dd@ÔÔ+  ÔÐ 3€gg"  _ €<€‚ˆ‚ƒƒ„C„3ÐÓ gRÓÌÌÌCommunityÓ ؃ÓÐ ;€gg*X ¢Cgg „C„;ÐÓ! „,„Ü„„“X!ÓÓ ؃ÓÌRole€inÌEPAÌTestÓ ؃ÓÐ ?€gg.X ¢ Cgg „C„……?ÐÓ9%„ ,„Ü4Œ ä <”ìDœ„,‚X9ÓÓ ؃ÓÌòò€€€€€€€€€€€€€€€€€€€€€€Elements€of€Treatmentòòaóó€€€€€€€€€€€€€€€€€€€€€€€€€€€€óóÐ † Ð  ÐÌà0 ÿ àà0Wÿ7%ÿ7%àà0¯W7%W7%àà0¯7%¯7%àà0_7%7%àà0·_7%_7%àà _ àà  àà ¯ àà W àà ÿ àà § àMediaà ÿ àà W àà ¯ à€€€€€ÔQQQ¨¨ÔMailingsà _ à€€€€€€€Ô¬¬¬ïïÔOutreachÓ-„,„Ü4Œ ä <„,ŠX-ÓÓ ؃ÓÐ H€gg7,X ¢·7%·7%Cgg „C„HÐImpact€ofÓ ؃ÓÌIndependentÌWJLAÌProgramòòbóóÓ ؃ÓÐ ?€gg.X ¢Cgg „C„……?ÐÓ ؃ÓÌ€€€SampleÌ€€€€SizeÌBà0 - à€ÔÕ-Õ-Õ-$$ÔFòòdóóÓ „,„„,†XÓÓ ؃ÓÐ YO7X ¢-£/-£/C ˆ < gg €J€‚x‚ƒƒ„C„YÐRandallstownÐ 'É C „C„'ÐÓ! „,„Ü„,X!ÓControlÐ 'É C „C„'ÐÓ'„,„Ü4Œ „,‚X'ÓNoneòòcóóÐ 'É C „C„'ÐÓ$„,„Ü4„,„X$ÓNoneòòcóóÐ 'É C „C„'ÐNoneòòcóóÐ 'É C „C„'ÐÓ! „,„Ü„,ƒX!Ó10€kitsÐ 'É C „C„'ÐÓ „,„„,‚XÓ495Ó ؃ÓÌ(48)Ó ؃ÓÐ '²üC „C„'Ð307Ó ؃ÓÌ(78)Ó ؃ÓÐ ?5²ü!C x J €©€‚x‚ƒƒ„C„?ÐHagerstownÐ '] "C „C„'ÐÓ! „,„Ü„,X!ÓModestÓ ؃ÓÌMediaÌCampaignÓ ؃ÓÐ 'å/ %C „C„'ÐÓ'„,„Ü4Œ „,‚X'Ó5€different€PSAÓ ؃ÓÌto€3€radio€stationsÌevery€two€weeks;Ìproject„designedÌposters€ran€inÌnewspaper€oneÌweek€in€February;Ìfour€additionalÌarticles€in€paperÓ ؃ÓÐ '[¥.C „C„'ÐÓ$„,„Ü4„,„X$ÓPamphletÓ ؃ÓÌinserted€inÌutility€billsÌduringÏFebruary/MarchÏbilling€cycleÓ ؃ÓÐ ' ê4C „C„'ÐNoneÐ '] 5C „C„'ÐÓ! „,„Ü„,ƒX!Ó93€kitsÐ '] 6C „C„'ÐÓ „,„„,‚XÓ529Ó ؃ÓÌ(64)Ó ؃ÓÐ 'üF 8C „C„'Ð432Ó ؃ÓÌ(82)Ó ؃ÓÐ F€g5üF :C x © €t €‚x‚ƒƒ„C„FÐFrederickÐ ;€g*¼;Cg „C„;ÐÓ! „,„Ü„,X!ÓExtensiveÓ ؃ÓÌMediaÌCampaignÓ ؃ÓÐ ;€g*ŽØ>Cg „C„;ÐÓ'„,„Ü4Œ „,‚X'Ó5€different€PSAÓ ؃ÓÌto€2€radio€stationsÌevery€two€weeks;Ìfive€of€eightÌarticles€in€localÌnewspaperÌproject„relatedÓ ؃ÓÐ ;€g*2|ECg „C„;ÐÓ$„,„Ü4„,„X$ÓPamphletÓ ؃ÓÌinserted€inÌutility€billsÌduringÌFebruary/MarchÌbilling€cycleÓ ؃ÓÐ ;€g*I“KCg „C„;ÐPosters€(4)Ó ؃ÓÌin€locationsÌaround€town;€nineÏpresentation€toÏcommunity€groupsÌbetween€JanuaryÏand€March€1988;Ïorganized€RadonÏAwareness€Week;Ìarranged€for€MayorÏand€Alderman€to€Ìmonitor€their€homesÓ ؃ÓÐ ;€g*¿" WCg „C„;ÐÓ! „,„Ü„,ƒX!Ó669€kitsÐ ;€g*¼XCg „C„;ÐÓ „,„„,‚XÓ523Ó ؃ÓÌ(56)Ó ؃ÓÐ ;€g*¥ïZCg „C„;Ð432Ó ؃ÓÌ(83)Ó q1ÓÐ:0.¥ï\C x t  g :ÐÐ  ¸#\ ÐÓWC„,„Ü4Œ ä <”ìDœôL¤ü!T$¬&)\+´- 0„,XWÓTable€2€(continued)Ó q1ÓÌÌòòaóóà0 Ü àThe€primary€messages€emphasized€in€the€public€service€announcements€(PSA),€posters,€public€meetings,€etc.€were:ЈÒÜ 0Ü 0 ÐÌà Ü àà 4 à„€Radon€is€a€serious€health€risk.€€òòYouóó€may€be€at€risk.€€The€only€way€to€find€out€is€to€test.Ð Z ¤ Ðà Ü àà 4 à„€Testing€is€easy€and€inexpensive.Ìà Ü àà 4 à„€Radon€problems€can€be€fixed.ÌÌòòbóóà0 Ü àIndependent€of€the€EPA€Project,€a€Washington,€D.C.€television€station,€WJLA,€conducted€a€month„long€campaign€to€encourageÐ þ H Ðpeople€to€test€their€homes€for€radon€during€January/February€of€our€study€period.€€The€campaign€had€multiple€components,€withÏcoordination€by€television€and€newspapers,€and€availability€of€radon€test€kits€at€a€reduced€price€at€Safeway€Supermarkets.€€TheÏcampaign€began€on€January€12,€1988,€included€a€three„part€news€series€during€the€week€of€January€18,€and€ended€with€another€three„¼part€series€during€the€week€of€February€15„19.€€Advertisements€for€the€effort€appeared€in€the€òòWashington€Postóó€and€on€television.€Ð ¢ì  Ð100,000€radon€tests€kits€were€purchased,€with€70,000€returned€for€analysis.€€Television€ratings€indicated€an€audience€of€76,000Ïviewers.€€The€television€programs€targeted€the€Washington,€D.C.€viewing€area.€€Residents€of€Frederick€were€most€likely€to€watch€theÏstation€airing€the€programs.€€Hagerstown€residents€could€view€WJLA,€but€reception€is€poor€without€cable.€€Randallstown€is€outside€theÏviewing€area,€but€reception€is€good.€€Numbers€refer€to€the€radon€test€kits€returned€from€each€community.Ð Ü 0Ü 0 ÐÌòòcóóà0 Ü àNone€refers€to€the€fact€that€there€were€no€project„initiated€information€materials€presented€in€public€media€mailings€or€meetings.€Ð b ÐSome€independent€newspaper€articles€did€appear€in€a€local€paper.Ð Ü 0Ü 0 ÐÌòòdóóà0 Ü àThe€numbers€in€parentheses€below€the€sample€sizes€were€the€response€rates.€€For€the€details€of€how€they€were€computed€seeÐ Ó ÐDesvousges,€Smith€and€Rink€[1989].€B€designates€the€baseline€survey€and€F€the€followup€survey€with€the€same€individuals.€€TheÏnumbers€in€parentheses€below€the€sample€sizes€are€the€response€rates€in€percentage€terms.Ð Ü 0Ü 0 Ðñ ññ ñÌÐ  wÁ Ðñ ñññÑ7€ ôôdÁNNdÄ7ÑÓE1„„Ü4Œ ä <”ìDœôL¤ü!T$„,”XEÓà@s s ÈàTable€3.€€Testing€Education's€Role€With€Maryland€Risk€Communication€ExperimentòòaóóˆÐ ¶ ÐÓ gRÓÌÔ*ƃpq ddŒdd@3dd@°dd@Tdd@Tdd@(dd@·dd@dd@mn„ 0„ 0ÆÔÔ,4dd@ÔÔ,4dd@ÔÔ,4dd@ÔÔ,4dd@ÔÔ+  ÔÐ 3€gg" Ò €j€‚ˆ‚ƒƒ„C„3ÐÓ'„,„Ü4Œ „„ŽX'ÓIndependentÓ gRÓÌVariablesÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐÓ ؃ÓÌFull€SampleÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐGreater€thanÓ ؃ÓÌHigh€SchoolÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐHigh€SchoolÓ ؃ÓÌor€BelowÓ ؃ÓÐ LB*C  C ˆ j gg €J€‚x‚ƒƒ„C„LÐInterceptÐ '´ þ C „C„'Є3.072òò*óóÓ ؃ÓÐ ´ þ  Ð(„7.50)Ó ؃ÓÐ ' ç C „C„'Є4.113òò*óóÓ ؃ÓÐ ´ þ Ð(„4.99)Ó ؃ÓÐ ' çC „C„'Є3.422òò*óóÓ ؃ÓÐ ´ þ Ð(„2.80)Ó ؃ÓÐ ?5 çC x J €J€‚x‚ƒƒ„C„?ÐYears€of€SchoolingÐ 'þ HC „C„'Ð.007Ó ؃ÓÌ(0.25)Ó ؃ÓÐ 'ç 1C „C„'Ð.066Ó ؃ÓÌ(1.32)Ó ؃ÓÐ 'ç 1C „C„'Ð.063Ó ؃ÓÌ(0.60)Ó ؃ÓÐ ?5ç 1C x J €J€‚x‚ƒƒ„C„?ÐFamily€IncomeÓ ؃ÓÌ(in€Thousands)Ó ؃ÓÐ '1{ C „C„'Ð.050Ó ؃ÓÌ(1.45)Ó ؃ÓÐ '1{ C „C„'Ð.065Ó ؃ÓÌ(1.50)Ó ؃ÓÐ '1{ C „C„'Ð.028Ó ؃ÓÌ(0.46)Ó ؃ÓÐ ?51{ C x J €J€‚x‚ƒƒ„C„?ÐHealth€AttitudeòòbóóÐ '’Ü !C „C„'Ð.195òò+óóÓ ؃ÓÐ ’Ü " Ð(1.62)Ó ؃ÓÐ '{Å #C „C„'Ð.078Ó ؃ÓÌ(0.51)Ó ؃ÓÐ '{Å %C „C„'Ð.361òò+óóÓ ؃ÓÐ ’Ü & Ð(1.80)Ó ؃ÓÐ ?5{Å 'C x J €J€‚x‚ƒƒ„C„?ÐIdentified€TVòòcóóÐ 'Ü&(C „C„'Ð.156Ó ؃ÓÌ(1.28)Ó ؃ÓÐ 'Å*C „C„'Ð.127Ó ؃ÓÌ(0.79)Ó ؃ÓÐ 'Å,C „C„'Ð.167Ó ؃ÓÌ(0.85)Ó ؃ÓÐ ?5Å.C x J €J€‚x‚ƒƒ„C„?ÐKnowòòdóóÐ '&p/C „C„'Ð.239òò*óóÓ ؃ÓÐ &p0 Ð(6.17)Ó ؃ÓÐ 'Y1C „C„'Ð.229òò*óóÓ ؃ÓÐ &p2 Ð(4.48)Ó ؃ÓÐ 'Y3C „C„'Ð.244òò*óóÓ ؃ÓÐ &p4 Ð(3.86)Ó ؃ÓÐ ?5Y5C x J €J€‚x‚ƒƒ„C„?ÐImproveòòeóóÐ 'pº6C „C„'Ð.230òò*óóÓ ؃ÓÐ pº7 Ð(3.61)Ó ؃ÓÐ 'Y£8C „C„'Ð.183òò+óóÓ ؃ÓÐ pº9 Ð(1.81)Ó ؃ÓÐ 'Y£:C „C„'Ð.253òò*óóÓ ؃ÓÐ pº; Ð(2.97)Ó ؃ÓÐ ?5Y£<C x J €J€‚x‚ƒƒ„C„?ÐFrederick€(=1)Ð 'º=C „C„'Ð.418Ó ؃Ó*Ì(2.61)Ó ؃ÓÐ '£í?C „C„'Ð.659òò*óóÓ ؃ÓÐ º@ Ð(3.28)Ó ؃ÓÐ '£íAC „C„'Є.025Ó ؃ÓÌ(„0.09)Ó ؃ÓÐ ?5£íCC x J €J€‚x‚ƒƒ„C„?ÐHagerstown€(=1)Ð 'NDC „C„'Ð.082Ó ؃ÓÌ(0.46)Ó ؃ÓÐ 'í7FC „C„'Ð.163Ó ؃ÓÌ(0.71)Ó ؃ÓÐ 'í7HC „C„'Є.175Ó ؃ÓÌ(„0.61)Ó ؃ÓÐ ?5í7JC x J €a€‚x‚ƒƒ„C„?ÐPseudo€Ròò2óóÐ 'N˜KC „C„'Ð.148Ð 'N˜LC „C„'Ð.160Ð 'N˜MC „C„'Ð.152Ð F€g5N˜NC x a €q€‚x‚ƒƒ„C„FÐnÐ ;€g*¯ ùOCg „C„;Ð1132Ð ;€g*¯ ùPCg „C„;Ð633Ð ;€g*¯ ùQCg „C„;Ð499Ó q1؃ÓÐ:0.¯ ùRC x q g :ÐÓH4„,„Ü4Œ ä <”ìDœôL¤ü!T$„,„XHÓÓ q1ÓÌà@ÈÈ7ÈàˆÌÌòòaóóà0 Ü àThe€numbers€in€parentheses€below€the€estimated€coefficients€are€Z€statistics€based€on€the€estimated€asymptotic€standard€errors€for€theÐ c$­U Ðnull€hypothesis€of€no€association.Ð Ü 0Ü 0 ÐÌòòbóóà0 Ü àThis€is€a€qualitative€variable€(0,1)€for€a€health€attitude€question€asked€during€the€baseline€overview.€€€It€asks€each€respondent€toÐ ˆÒ Ðdescribe€themselves€as:€"I€ask€my€physician€a€lot€of€questions€about€my€health."€€The€question€asks€to€indicate€if€these€things€are€trueÏabout€the€respondent€always,€often,€sometimes,€or€never.€€Those€answering€always€or€often€were€coded€as€health€concerned€(1)Ïotherwise€as€less€concerned€(0).ññññÐ Ü 0Ü 0 ÐññññÐ  , v ÐññññÑ7€ ôôdÁôôdÁ7ÑÓE1„„Ü4Œ ä <”ìDœôL¤ü!T$„,XEÓà@$ $ ÈàTable€4.€€Education's€Impact€on€Learning:€Maryland€Risk€Communication€ExperimentòòaóóˆÐ ¶ ÐÓ gRÓÌÔ*‚ƒst d d4dd@4dd@4dd@4dd@pq„ 0„ 0‚ÔÔ,4dd@ÔÔ,4dd@ÔÔ,4dd@ÔÔ,4dd@ÔÔ+  ÔÐ 3€gg" Ò €j€‚ˆ‚ƒƒ„C„3ÐÓ'„,„Ü4Œ „„ŽX'ÓIndependentÓ gRÓÌVariablesÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐÓ ؃ÓÌFull€SampleÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐGreater€thanÓ ؃ÓÌHigh€SchoolÓ ؃ÓÐ ;€gg*C Cgg „C„;ÐHigh€SchoolÓ ؃ÓÌor€BelowÓ ؃ÓÐ LB*C  C ˆ j gg €J€‚x‚ƒƒ„C„LÐInterceptÐ '´ þ C „C„'Ð.448òò*óóÓ ؃ÓÐ ´ þ  Ð(2.09)Ó ؃ÓÐ ' ç C „C„'Ð1.192òò*óóÓ ؃ÓÐ ´ þ Ð(„3.67)Ó ؃ÓÐ ' çC „C„'Є.939òò+óóÓ ؃ÓÐ ´ þ Ð(„1.91)Ó ؃ÓÐ ?5 çC x J €J€‚x‚ƒƒ„C„?ÐYears€of€SchoolingÐ 'þ HC „C„'Ð.008Ó ؃ÓÌ(0.74)Ó ؃ÓÐ 'ç 1C „C„'Є.034Ó ؃ÓÌ(„1.46)Ó ؃ÓÐ 'ç 1C „C„'Ð.107òò*óóÓ ؃ÓÐ þ H Ð(3.00)Ó ؃ÓÐ ?5ç 1C x J €J€‚x‚ƒƒ„C„?ÐHealth€AttitudeÐ 'H’ C „C„'Ð.061Ó ؃ÓÌ(1.07)Ó ؃ÓÐ '1{ C „C„'Ð.024Ó ؃ÓÌ(0.33)Ó ؃ÓÐ '1{ C „C„'Ð.112Ó ؃ÓÌ(1.27)Ó ؃ÓÐ ?51{ C x J €J€‚x‚ƒƒ„C„?ÐAge€(in€years)Ð '’Ü C „C„'Є.004òò*óóÓ ؃ÓÐ ’Ü ! Ð(2.14)Ó ؃ÓÐ '{Å "C „C„'Є.004Ó ؃ÓÌ(„1.59)Ó ؃ÓÐ '{Å $C „C„'Є.001Ó ؃ÓÌ(„0.44)Ó ؃ÓÐ ?5{Å &C x J €J€‚x‚ƒƒ„C„?ÐMale€(=1)Ð 'Ü&'C „C„'Є.153òò*óóÓ ؃ÓÐ Ü&( Ð(2.62)Ó ؃ÓÐ 'Å)C „C„'Є.142òò+óóÓ ؃ÓÐ Ü&* Ð(„1.89)Ó ؃ÓÐ 'Å+C „C„'Є.138Ó ؃ÓÌ(„1.51)Ó ؃ÓÐ ?5Å-C x J €J€‚x‚ƒƒ„C„?ÐIdentified€TVÐ '&p.C „C„'Ð.090Ó ؃ÓÌ(1.55)Ó ؃ÓÐ 'Y0C „C„'Ð.016Ó ؃ÓÌ(0.22)Ó ؃ÓÐ 'Y2C „C„'Ð.159òò+óóÓ ؃ÓÐ &p3 Ð(1.76)Ó ؃ÓÐ ?5Y4C x J €J€‚x‚ƒƒ„C„?ÐFrederickÐ 'pº5C „C„'Ð.118òò+óóÓ ؃ÓÐ pº6 Ð(1.64)Ó ؃ÓÐ 'Y£7C „C„'Є.004Ó ؃ÓÌ(„0.05)Ó ؃ÓÐ 'Y£9C „C„'Ð.317òò*óóÓ ؃ÓÐ pº: Ð(2.59)Ó ؃ÓÐ ?5Y£;C x J €J€‚x‚ƒƒ„C„?ÐHagerstownÐ 'º<C „C„'Ð.122òò+óóÓ ؃ÓÐ º= Ð(1.67)Ó ؃ÓÐ '£í>C „C„'Ð.163òò+óóÓ ؃ÓÐ º? Ð(1.72)Ó ؃ÓÐ '£í@C „C„'Ð.141Ó ؃ÓÌ(1.21)Ó ؃ÓÐ ?5£íBC x J €a€‚x‚ƒƒ„C„?ÐPseudo€Ròò2óóÐ 'NCC „C„'Ð.015Ð 'NDC „C„'Ð.018Ð 'NEC „C„'Ð.046Ð F€g5NFC x a €q€‚x‚ƒƒ„C„FÐnÐ ;€g*e¯GCg „C„;Ð1315Ð ;€g*e¯HCg „C„;Ð721Ð ;€g*e¯ICg „C„;Ð594Ó q1؃ÓÐ:0.e¯JC x q g :ÐÓE1„„Ü4Œ ä <”ìDœôL¤ü!T$„,„XEÓà@ÈÈ7ÈàÓ q1ÓˆÌòòaóóà0 Ü àThese€estimates€are€based€on€ordinary€least€squares€with€the€dependent€variable€defined€as€the€proportionate€increase€in€the€count€ofÐ G ‘K Ðcorrect€answers€to€a€seven€question€multiple€choice€quiz,€asking€the€same€respondent€about€radon's€effects,€home€testing€andÏmitigation.€€The€numbers€in€parenthesis€below€the€coefficients€are€t„ratios€for€the€null€hypothesis€of€no€association.€€The€variables€areÏdefined€in€Table€3.Ð Ü 0Ü 0 ÐÐ ë#5O Ðòò*óóà0 Ü àSignificantly€different€from€zero€at€least€at€the€five€percent€level.Ð¶Ü 0Ü 0 ÐÌòò+óóà0 Ü àSignificantly€different€from€zero€at€least€at€the€one€percent€level.ЈÒÜ 0Ü 0 ÐÌññññÐ  Z ¤ ÐññññÑ7€ <NNdÄôôdÁ7ÑØ:KØÓB.°°` ¸ hÀpÈ xÐ (#„„ŽXBÓÓÓà@883Èàò òEndnotesó óˆÐ ¶ ÐÌ*à Ü àArts€and€Sciences€Professor,€Duke€University€and€University€Fellow€Resources€for€the€Future.€€Thanks€are€due€Ray€Kopp€forÐ \ ¦ Ðcollaborating€in€the€development€of€the€general€conceptual€issues€associated€with€valuing€nonmarket€resources,€to€Jere€Behrman,€WinstonÏHarrington,€Nevzer€Stacey,€Albert€Tuijnam,€and€Dennis€Yang€for€constructive€comments€on€earlier€drafts,€and€Paula€Rubio,€Terrie€Rouse,Ïand€Tricia€Shore€for€assuring€a€continuously€changing€manuscript€was€completed.€€Partial€support€for€this€research€was€provided€by€the€U.S.ÏDepartment€of€Education€and€the€UNC€Sea€Grant€Program€under€grant€R/MRD25.ÌÌà0 Ü àà „ à1.à Ü àThis€observation€was€first€made€by€Wildavsky€[1988]€in€considering€the€apparent€contradiction€between€concerns€about€risk€amongÐ H’ Ðpeople€in€developed€economics€and€the€actual€risks€they€experience.Ð Ü 0Ü 0 Ðà0 Ü àà „ à2.à Ü àGiven€the€concern€over€environmental€quality€in€the€U.S.€and€most€developed€economics,€this€comparison€may€seem€surprising,€but€itÐ ì6 Ðis€readily€confirmed€for€air€and€water€quality€where€there€are€reasonable€monitoring€data€and€may€well€be€an€adequate€summary€forÏhazardous€substances€as€well.€€See€the€World€Bank's€1992€report€òòDevelopment€and€the€Environment€óóor€the€United€Nation's€òòUrban€AirÐ Ú ÐPollution€in€Megacities€of€the€World.óóÐb¬Ü 0Ü 0 Ðà0 Ü àà „ à3.à Ü àIndeed,€the€National€Environmental€Education€Act€of€1990€sought€to€develop€educational€programs€and€curricula€that€respond€to€theÐ 0z Ðcomplex€nature€of€environmental€problems.€€The€legislation€notes€that:Ð Ü 0Ü 0 ÐÐ Ô"  ÐÒ„ÒÒ„Òà ` àà0 ¸ àà ` àà ¸ à".€.€.€effective€response€to€complex€environmental€problems€requires€understanding€of€the€natural€andÏbuilt€environment,€awareness€of€environmental€problems€and€their€origins,€and€the€skills€to€solve€theseÏproblems."Ð ¸ ˆ,¸ ˆ, Ðà0 ` àà  àà ` àà ¸ àMore€recently,€the€National€and€Community€Service€Trust€Act€of€1993€sought€to€expand€communityÏservice€and€student€service€activities€as€part€of€social€learning€that€it€argued€provided€visible€benefits€to€theÏparticipants€in€such€programs€and€to€their€communities.Ð ` ˆ,` ˆ, Ðà Ì àà  à4.à0 ` àà  àà ` àThese€models€were€proposed€by€Borcherding€and€Deacon€[1972]€and€Bergstrom€and€Goodman€[1973]€asÐ ¢ì  Ðtheoretically€consistent€alternatives€to€a€earlier€literature€that€used€€regression€models€that€sought€to€describeÏpublic€expenditures€per€capita€in€a€variety€of€differentcategories€as€a€function€of€state€and€local€areaÏcharacteristics.€€The€standard€models€usually€maintain€a€constant€elasticity€demand€function.à €% àÐb` ˆ,` ˆ, Ðà0 ` àà  à5.à ` àSee€Huang€and€Smith€[1995]€for€further€discussion€of€the€production€interpretation€of€weak€complementarity€asÐ ê4 Ða€basis€for€measuring€the€economic€value€of€non„marketed€resources€that€affect€production€activities.Ð ` ˆ,` ˆ, Ðà0 ` àà  à6.à ` àBecker's€[1964]€discussion€of€differences€between€the€private€and€social€rates€of€return€to€education€focusesÐ ŽØ Ðprimarily€on€the€subsidies€and€indirect€costs€not€reflected€in€his€estimates.€€No€consideration€was€given€in€toÏnon„market€effects.Ð ` ˆ,` ˆ, Ðà0 ` àà  à7.à ` àThe€HPF€framework€allows€the€analyst€to€introduce€structure€into€the€way€goods,€services,€and€time€combine€toÐ !N Ðyield€increases€in€well„being.€€See€Smith€[1991a]€for€a€discussion€of€other€restrictions€commonly€used€toÐ Ö"   Ðestimate€non„market€values€within€this€framework.Ð ` ˆ,` ˆ, Ðà0 ` àà  à8.à ` àThese€models€sometimes€include€constructed€price€measures.€€See€Gramlich€and€Rubinfeld€[1982]€an€example.ЈÒ` ˆ,` ˆ, Ðà0 ` àà  à9.à ` àMartinez„Vazquez€[1981]€considered€local€bond€referenda€to€evaluate€whether€public€goods€effects€(beyondÐ Z ¤ Ðdirect€private€use€benefits)€appear€to€influence€the€voting€behavior.€€He€concluded€that€the€private€net€benefitsÏwas€a€weak€explanatory€factor€because€high€income€groups€voted€in€favor€of€proposals€that€would€haveÏbenefited€lower€income€groups.Ð ` ˆ,` ˆ, Ðà0 ` àà  à10.à ` àThis€estimate€assumes€a€locally€constant€elasticity€demand€function€and€computes€the€consumer€surplus€per€unitÐ ¢ì  Ðof€education€based€on€the€elasticity€from€the€total€demand€(including€private€and€public€components€inÏWyckoff's€framework)€and€the€public€component€elasticity.€€See€Smith€and€Desvousges€[1986b]€for€theÏderivation.Ð ` ˆ,` ˆ, Ðà0 ` àà  à11.à ` àSee€Freeman€[1993]€for€an€overview€of€the€issues€associated€with€defining€and€measuring€existence€(or€nonuse)Ð ê4 Ðuses.Ð ` ˆ,` ˆ, Ðà0 ` àà  à12.à ` àSee€Wolfe€with€Zuvek€as€(undated)€for€an€update€to€the€Haveman„Wolfe€analysis.ÐŽØ` ˆ,` ˆ, Ðà0 ` àà  à13.à ` àThe€embedding€issue€was€first€identified€by€Kahneman€and€Knetsch€[1992]€as€a€fundamental€problem€with€theÐ `ª Ðuse€of€survey€techniques€in€non„market€valuation.€€Support€for€this€view€as€offered€in€some€of€the€studiesÏreported€in€Hausman€[1993],€notably€the€Kemp„Maxwell€[1993]€and€Diamond,€Hausman,€Leonard€and€DenningÏ[1993]€papers.€€By€contrast,€an€explanation€that€recognizes€embedding€as€an€aggregation€issue€can€be€found€inÐ Ö"   ÐSmith€[1992]€and€Carson€and€Mitchell€[forthcoming].Ð ` ˆ,` ˆ, Ðà0 ` àà  à14.à ` àSee€Ridker€[1967]€for€the€first€example€using€this€basic€idea€for€measuring€the€value€of€reducing€air€pollution.ЈÒ` ˆ,` ˆ, Ðà0 ` àà  à15.à ` àSee€Slovic€et.€al.€[1985]€and€the€National€Research€Council€[1989]€for€discussions€of€these€discrepancies€andÐ Z ¤ Ðtheir€policy€implications.€€Viscusi's€[1993]€prospective€reference€theory€offers€a€different€interpretation€for€thisÏresearch,€suggesting€it€is€a€reflection€of€informal€Bayesian€formulation€of€subjective€risk€perceptions.Ð ` ˆ,` ˆ, Ðà0 ` àà  à16.à ` àBehrman's€[1994]€discussion€of€this€issue€in€a€general€context,€reflects€the€longstanding€examples.€€One€of€theseÐ Ð  Ðrelevant€to€some€of€my€discussion€by€Fuchs€[1982]€argues€that€unobservables,€such€as€differential€rates€of€timeÏpreferences,€could€have€been€responsible€for€more€educated€individuals€making€greater€health€investments,Ïrather€than€a€causal€link€between€education€and€health.Ð ` ˆ,` ˆ, Ðà0 ` àà  à17.à ` àSeparate€examples€for€this€type€of€relationship€using€information€from€actual€housing€sales€can€be€found€inÐ b ÐMichaels€and€Smith€[1990]€and€Kolhase€[1991].Ð ` ˆ,` ˆ, Ðà0 ` àà  à18.à ` àThere€was€subsequent€evidence€of€systematic€recall€errors€as€the€time€horizon€for€the€summary€judgement€ofÐ ¼ Ðactivity€is€increased€see€Westat€[1989].Ð ` ˆ,` ˆ, Ð19.à0 ` àà  àà ` àMachina€[1990]€has€argued€that€in€evaluating€the€economic€efficiency€of€allocation€decisions,€analysts€shouldÐ `ª Ðbe€prepared€to€intervene€and€"correct"€risk€misperceptions€people€form€in€valuing€policy€intervention€if€they€areÏbased€on€incorrect€interpretations€of€or€incorrect€use€of€logical€methods€(i.e.€probability€theory)€with€theÏavailable€information€on€the€risk.€€His€arguments€raise€important€questions€about€how€consumer€sovereigntyÐ Ö"   Ðaffects€the€analyst's€interpretation€of€"mistakes."€€Education€presumably€enters€this€brew€by€influencing€the€rateÏof€mistakes.€€This€would€seen€especially€relevant€to€decisions€involving€risk,€where€people€seen€to€have€a€veryÏdifficult€time€understanding€probability€concepts.€€I€return€to€this€point€as€a€more€general€issue€in€the€sectionÏIV.Ð ` ˆ,` ˆ, Ðà0 ` àà  à20.à ` àSeparate€analysis€of€these€mitigation€decisions€of€household€in€the€New€York€panel€study,€considering€theirÐ þ H Ðdollar€expenditures€have€a€number€of€problems.€€Because€the€activities€do€not€have€well€defined€prices€(i.e.€theÏradon€mitigation€must€be€adjusted€to€the€characteristics€of€each€house)€it€was€not€possible€to€obtain€meaningfulÏprice€measures.€Equally€important€the€actions€involved€a€mix€of€private€homeowners€activities€and€monetaryÏexpenditures.€€The€later€can€include€equipment€that€serves€multiple€objectives€(e.g.€a€new€furnace€ventilationÏsystem).€€Finally,€the€mitigation€reports€included€a€limited€number€of€observations€reporting€expenditureÏexclusively€for€radon€reduction.€€Thus,€there€are€reasons€to€question€what€can€be€learned€from€the€monetaryÏexpenditures€data.€€A€tobit€analysis€was€consistent€with€these€reservations.€€€The€overall€sign€of€effectsÏidentified€with€the€discrete€classification€of€activities€were€consistent,€but€few€of€the€variables€would€have€beenÏjudged€statistically€significant€by€conventional€criteria.Ð ` ˆ,` ˆ, Ðà0 ` àà  à21.à ` àThere€was€also€an€independent€sample€collected€after€the€public€information€programs€in€each€town€to€evaluateÐ 2| Ðthe€potential€for€sensitization€of€respondents€in€the€panel€sample€due€to€the€baseline€survey.€€These€results€didÏnot€appear€to€be€important€(see€Desvousges,€Smith€and€Rink€[1989]).ÐÖ"  ` ˆ,` ˆ, Ðà0 ` àà  à22.à ` àThis€outcome€measure€is€a€òònewóó€decision€to€test€for€radon.€€The€panel€survey€was€evaluated€three€ways€„€withoutÐ ¶ Ðthe€homeowners'€reporting€that€they€had€undertaken€tests€in€the€baseline€survey,€adjusting€for€prior€tests€as€aÏselection€effect,€and€the€current€method€that€treats€them€as€unaffected€by€the€program.€€All€three€approachesÏyield€comparable€conclusions€and€the€complementary€role€of€education.€€The€current€method€was€selectedÏbecause€it€makes€the€impact€"most€difficult"€to€isolate€and€offers€the€largest€sample€for€disaggregated€analysis.Ð ` ˆ,` ˆ, Ðà0 ` àà  à23.à ` àThree€studies,€Hoehn€[1991],€and€Hoehn€and€Loomis€[1993]€have€considered€how€people's€responses€toÐ Ð  Ðvaluation€questions€are€influenced€by€the€sequencing€of€miles€of€closely€related€environmental€public€goods.€ÏThis€research€seeks€to€evaluate€the€effects€of€contemporaneously€provided€public€goods€due€to€substitution€andÏcomplementarity€relationships€that€may€exist€between€the€goods.€€See€Madden€[1991]€and€Carson,€HanemannÏand€Flores€[1994]€for€a€discussion€of€the€conceptual€rationale€for€these€types€of€tests.€Because€education'sÏpublic€good€effects€are€presumably€long€term€impacts,€this€type€of€contemporaneous€effect€seems€unlikely.Ð ` ˆ,` ˆ, Ðà0 ` àà  à24.à ` àThe€EPA€Science€Advisory€Board€report€that€started€these€activities;€òòUnfinished€Business:€A€ComparativeÐ ¼ ÐAssessment€of€Environmental€Problemsóó€[1987]€contributed€to€changing€the€agency's€approach€to€evaluating€itsÐ ŽØ Ðpolicy€goals.Ð ` ˆ,` ˆ, Ðà0 ` àà  à25.à ` àA€recent€study€of€adult€literacy€in€the€U.S.€sketches€an€even€more€discouraging€picture,€suggesting€thatÐ 2| Ðlanguage€barriers€and€insufficient€education€were€important€barriers€to€common€tasks€like€"writing€a€letterÏabout€a€billing€error€or€calculating€the€length€of€a€bus€trip€from€a€schedule."€€see€Kirsch€et€al.,€[1993].à @) àÐÖ"  ` ˆ,` ˆ, Ðà0 ` àà  à26.à ` àA€vivid€example€of€this€tendency€can€be€found€in€Fuchs'€Foreward€to€Mincer's€classic€study€of€experience€andÐ ¶ Ðearnings€where€he€observesÐ ` ˆ,` ˆ, ÐÒ°Òà0  àà0` ˆ,ˆ,àà  àà ° àà  àà ` à"In€the€pages€that€follow€Mincer€demonstrates€his€skill€as€a€wielder€of€Occam's€€€razor.€His€objective€is€toÏexplain€a€great€deal€with€a€little"€(P€xiii)à p àÐ, v` ˆ,` ˆ, Ðà0  àà ° àà  àWhat€he€does€not€say€is€whether€the€explanations€are€consistent€with€what€people€believed€they€were€doing€when€theyÏmade€those€decisions.ÓÓññññÐ Ð ˆ,ˆ, ÐÌòòcóóà0  àThis€variable€attempts€to€take€account€of€the€effects€of€the€WJLA€radon€information€program.€€Because€these€effortsÐ Ÿé Ðwere€largely€related€to€public€service€announcements€on€TV,€a€qualitative€variable€coded€as€one€for€those€individualsÏidentifying€TV€as€the€source€of€their€information€and€zero€otherwise.Ð ˆ,ˆ, ÐÌòòdóóà0  àKnow€is€a€count€of€the€correct€answers€for€the€seven€questions€during€the€baseline€interview.ÐC ˆ,ˆ, ÐÌòòeóóà0  àImprove€is€the€proportionate€change€in€the€number€of€correct€answers€between€the€baseline€and€followup€interviews.€Ð  _ ÐThis€is€interpreted€as€the€effect€of€the€program€on€learning.Ð ˆ,ˆ, ÐÌòò*óóà  àSignificantly€different€from€zero€at€least€at€the€five€percent€level.Ð Ð  ÐÌòò+óóà0  àSignificantly€different€from€zero€at€least€at€the€ten€percent€level.Ð¢ì ˆ,ˆ, ÐÌññññññÔ†ÿÿÔÔ†ÿÿÔÔ  ÔÔ  ÔññññÐ  t¾  ÐÔ€Nô‰NXX-³ÔÔ€N <NNNô‰ÔòòÌÌ