The Determinants of Deposit Insurance Coverage: Evidence from 74 Countries

Transcription

1 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 The Deerminans of Deposi Insurance Coverage: Evidence from 74 Counries Hung Ling Chen 1, Edward H. Chow 2 & Wan Yu Liu 3 1 Deparmen of Inernaional Business, Shih Chien Universiy, Taiwan 2 Deparmen of Finance, Naional Chengchi Universiy, Taiwan 3 Enie Bank, Taiwan Correspondence: Hung Ling Chen, Deparmen of Inernaional Business, Shih Chien Universiy, Taiwan. chenhub@gmail.com Received: February 2, 2014 Acceped: February 24, 2014 Online Published: Augus 24, 2014 doi: /ijfr.v5n4p13 URL: hp://dx.doi.org/ /ijfr.v5n4p13 Absrac Using daa from 74 counries beween he years of 1960 o 2008, we performed panel regressions and logi models on poliical seings and he economic siuaion o idenify which facors influence deposi insurance coverage and he probabiliy of changes in coverage. Resuls reveal ha counries wih higher coverage are hose wih less poliical openness, lower ineres rae spreads and deposi ineres raes; conversely, we find ha coverage ends o be higher in counries wih greaer levels of governmen deb. Considering he ime-invarian counry heerogeneiy, counries wih lower ineres rae spreads are likely o increase he coverage and he probabiliy of changes in coverage. The resuls could provide reference poins for policymakers in considering he seing of deposi insurance schemes, wih he aim of ensuring financial sabiliy. Keywords: deposi insurance, coverage, poliical economy 1. Inroducion Given he subsanial differences in he poliical srucure of individual counries, he banking indusry, overall economic environmen, and deposi insurance sysems (DIS) differ enormously from counry o counry. The primary aim of his sudy deermines facors ha affec deposi insurance coverage and he probabiliy of changes in such coverage. By idenifying he facors wih significan influence on coverage, his research provides a new perspecive o consideraions for seing up DIS o ensure financial sabiliy. Afer grea depression banking crisis in 1933, Unied Saes was he firs o esablish DIS in According o Inernaional Associaion of Deposi Insurers (IADI) organizaion s descripion, number of counries eiher esablishing or considering esablishing a deposi insurance sysem has expanded rapidly in recen years. As of 31 January 2014, 113 counries have insiued some form of explici deposi insurance up from 12 in Anoher 40 counries are sudying or considering he implemenaion of an explici deposi insurance sysem. (Noe 1) According o IADI principles, he objecives of a deposi insurance sysem are o proec he majoriy of small and unsophisicaed deposiors and o mainain deposior confidence and enhance he macroeconomic and financial sabiliy of he banking sysem (IADI, 2008). This is paricularly imporan in imes such as he financial crisis of 2008, when counries around he globe suffered from a breakdown in public confidence in he financial sysem as a whole. Deposi insurance acs as a financial safey ne for prevening bank runs and mainaining public confidence. Therefore, i is imporan for each counry o have appropriae coverage according o is own siuaion. If coverage is oo low, i fails o proec small and unsophisicaed deposiors. However, problems relaed o moral hazards are likely o occur if he level of coverage is se oo high. This is because a higher level of coverage provides incenives for banks o ake greaer risks and he poenial o lead o an overall rise in he level of insabiliy wihin he financial sysem. Thus, regulaors mus idenify he mos suiable coverage for heir counry in accordance wih heir own counry-specific characerisics. Several previous sudies have examined he relaionship beween deposi insurance and financial sysems. For Published by Sciedu Press 13 ISSN E-ISSN

2 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 example, explici deposi insurance sysems raised he likelihood of disress (Huchison and McDill, 1999; Demirgüç-Kun and Deragiache, 1998, 2002; Chu, 2003) and had a negaive impac in counries ha lacked well-esablished bank supervision and law enforcemen (Cull, Senbe and Sorge, 2005). The facors direcly influencing coverage levels or changes remain unclear. For example, he resuls from he applicaion of a poliical economy framework underaken by Laeven (2004) revealed ha deposi insurance coverage was higher in counries dominaed by poorly capialized banks and in counries wih poorly educaed deposiors. However, Laeven s analysis did no uncover any significan associaion beween deposi insurance coverage and poliical-insiuional variables, nor is here a significan associaion for insiuional developmen variables. Demirgüç-Kun, Kan, and Laeven (2008) found ha counries wih more-democraic environmens and counries wih a larger proporion of risky banks are more likely boh o adop deposi insurance and o design i wih higher insurance coverage. Our findings are disinc from he findings in he exising lieraure: (1) he effecs of economic and poliical facors on he level of coverage and (2) he associaion of he facors and he probabiliy of changes in coverage. In addiion o employing he poliical-insiuional variables from Laeven (2004) and Demirgüç-Kun e al. (2008) in he curren analysis, his work adds oher relevan economic variables (savings raes, ineres rae spreads, deposi ineres rae, and governmen deb) o examine heir effecs on coverage. As opposed o analyzing he design of deposi insurance for only a single year (2000) as in Laeven, his sudy used a comprehensive daabase ha includes boh cross-secional and ime-series daa (74 counries from 1960 o 2008). The findings show ha counries wih lower ineres rae spreads, lower deposi ineres rae, and higher governmen deb levels end o have higher coverage. Democraic counries also end o have lower coverage levels, whereas auocraic counries end o have higher coverage levels. Thus, our more complee daase enabled reaching conclusions ha were unavailable o Laeven. Anoher significan difference exiss beween his paper and ha of Laeven (2004) and Demirgüç-Kun e al. (2008). The curren sudy examined he probabiliy of changes in deposi insurance coverage and found ha counries wih lower poliical openness, lower ineres rae spreads, lower deposi ineres rae, and higher governmen deb have a lower probabiliy of increasing coverage levels. By considering ime-invarian counry heerogeneiy, only lower ineres rae spreads can subsanially increase he changes in coverage and he probabiliy of a coverage increase. The resuls should help regulaors deermine he appropriae level of fuure coverage, which migh help preven bank runs. The remainder of his paper is organized as follows: Secion 2 inroduces definiions of boh he daa and mehodology adoped for his sudy; Secion 3 provides a discussion of he empirical resuls; and lasly, Secion 4 offers a conclusion. 2. Daa and Mehodology 2.1 Daa The empirical analysis in his sudy primarily focuses on facors ha poenially affec deposi insurance coverage. The sample daa were gahered on 74 counries from 1960 o The sample descripions are provided in Table 1. Table 1. Sample counries descripion Panel A: Number of counries High counries Upper-middle counries Lower-middle counries Low counries Norh America 2 2 Lain America & Caribbean Wesern Europe Europe & Cenral Asia Africa 4 4 Middle Eas & Norh Africa Eas Asia & Pacific Souh Asia Toal Published by Sciedu Press 14 ISSN E-ISSN

3 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 Toal Panel B: Raio of deposi coverage changes for every counry High counries Upper-middle counries Lower-middle counries Low counries average Norh America Lain America & Caribbean Wesern Europe Europe & Cenral Asia Africa Middle Eas & Norh Africa Eas Asia & Pacific Souh Asia Average The dependen variable is he exen of deposi insurance coverage, calculaed as he raio of he amoun of deposi insurance coverage o per capia GDP. This raio, which can reduce he economic scale effec for each counry, has also been widely used by he Inernaional Associaion of Deposi Insurers ( IADI ), and by researchers examining relaed deposi insurance coverage issues. To suppor our examinaion of he facors ha are likely o lead o changes in deposi insurance coverage, we defined anoher dependen variable ( CHANGE ), which is equal o 1 if he level of deposi insurance coverage has changed from he level of coverage in he previous year; oherwise, CHANGE is equal o 0. The counries o adop deposi insurance were derived from Demirgüç-Kun e al. (2008). Limied by variables and deleing fixed deposi insurance coverage during he period, he final sample is 74 counries. (Noe 2) The sample consiss of 27 high- counries, 15 upper-middle- counries, 24 lower-middle- counries, and 8 low- counries. The independen variables are divided ino wo differen aspecs, consising of he poliical seing and he overall economic environmen. The economic variables were derived from he World Bank and EIU Counry Daa. The poliical variables are from he Poliy IV daabase. All he variable definiions are provided in Table 2. Some of he measures used in he presen sudy are similar o hose used in Laeven (2004); however, as opposed o using variables o explain he design of deposi insurance for only a single year (2000), we exended he daase o form longiudinal panel daa which is like as Demirgüç-Kun e al.. No all he variables have complee daa for he 49-year period, and hus, applying panel daa when conducing saisical analysis increases he overall sample. Anoher advanage of using panel daa analysis is o observe heerogeneiy across counries. Table 2. Variable definiions Sample Variables Definiion Period Panel A: Dependen variables COV_GDP The raio of he oal amoun of deposi insurance coverage o per-capia GDP CHANGE Equal o 1 if here is a change in deposi insurance coverage from he previous year; oherwise equal o zero if he amoun of coverage remains he same. Panel B: Independen variables POLITY Compued by subracing he AUTOC score from he DEMOC score; he resuling unified poliy scale ranges from +10 (srongly democraic) o (srongly auocraic). DEMOC An addiive eleven-poin scale (from zero o en) derived from coding he compeiiveness of poliical paricipaion, openness of execuive recruimen and consrains on he chief execuive. AUTOC An addiive eleven-poin scale (from zero o en) derived from coding he Published by Sciedu Press 15 ISSN E-ISSN

4 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 compeiiveness of poliical paricipaion, regulaion of paricipaion, compeiiveness of execuive recruimen and consrains on he chief execuive, using differen weighs. SAVINGS Gross savings (% of GDP) INT_SPREAD Ineres rae spread = lending rae deposi rae (%) DEP_INT Deposi rae (% per annum) GOV_DEBT Toal of domesic, exernal and IMF deb owed by general governmen (% of GDP) Mehodology Panel daa were used in he analysis o undersand he facors affecing deposi insurance coverage and he probabiliy of changing deposi insurance coverage. The raionale for using panel daa regression is based on deermining he relaionship beween deposi insurance coverage vis-à-vis poliical and economic facors because he oal amoun of daa available for a single year analysis is exremely limied. Analyzing panel daa can increase he oal number of observaions covering boh he ime-series and cross-secional daa and accoun for all unobserved cross-counry heerogeneiy. Two approaches are commonly used o analyze panel daa: he fixed effecs approach and he random effecs approach. To decide beween he fixed or random effecs approaches, we ran he Hausman (1978) es, where he null hypohesis posis ha he random effecs model is he preferred model, whereas he fixed effecs model is he alernaive model. The resuls of he Hausman es in he presen sudy clearly show he fixed effecs approach o be more appropriae. Following confirmaion ha he fixed effecs approach is more appropriae, panel daa analysis was performed using he following model: Y 0 1X 1, k X k, i where Y is he dependen variable (COV_GDP), X k, refers o he independen variables, i is he counry, and is he year. The independen variables consis of POLITY, DEMOC, AUTOC, SAVINGS, INT_SPREAD, DEP_INT, and GOV_DEBT. β k refers o he coefficien on he independen variables, α i is he counry facors, and ε is he error erm. To suppor he examinaion in his sudy of he facors likely o lead o changes in deposi insurance coverage, we employed a panel logi model, expressed as follows: P L ln 0 1X 1,... k X k, u 1 P where P indicaes ha CHANGE akes he value of 1 if a change has occurred in he level of coverage; oherwise, CHANGE akes he value of 0, L is he log X k, refers o he independen variables, i is he counry, and is he year. If he log L, is posiive, hen he independen variables are considered o have some influence on changes in he level of deposi insurance coverage, hereby indicaing ha he variable is responsible for he change in coverage. 3. Empirical Resuls 3.1 Deerminans of Deposi Insurance Coverage The panel daa observaions include boh cross-secional and ime-series daa aimed a idenifying facors wih significan influences on deposi insurance coverage. The resuls of he fixed effecs panel regression, which includes he explanaory variables POLITY, DEMOC, AUTOC, SAVINGS, INT_SPREAD, DEP_INT, and GOV_DEBT are presened in Table 3; he dependen variable is COV_GDP, which is he raio of coverage o per capia GDP. The poliical-insiuional variable POLITY was found significan wih a negaive coefficien, as are he economic variables INT_SPREAD and DEP_INT. The variable GOV_DEBT, which is he raio of oal governmen deb o GDP, was found o have a posiive correlaion wih deposi insurance coverage, which suggess ha coverage ends be higher for hose counries wih higher governmen deb levels. The coverage ends o be lower in counries wih greaer poliical openness, higher ineres rae spreads, and higher deposi ineres rae. These resuls are consisen (1) (2) Published by Sciedu Press 16 ISSN E-ISSN

5 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 wih he discussions in mos prior lieraure. Angkinand (2005) found ha poliical corrupion was more likely o increase he probabiliy of a banking crisis. Such counries also end o have lower levels of poliical openness (Sandholz and Kozle, 2000), which leads o he finding ha counries wih lower levels of poliical openness are associaed wih higher levels of coverage. Alhough Laeven (2004) also found ha less-democraic counries have higher coverage raio bu have lile power in explaining variaion in COV_GDP, he use of longiudinal panel daa found ha poliical openness could affec he coverage level. On he oher side, Demirgüç-Kun e al. (2008) found ha counries wih more-democraic environmens are more likely boh o adop deposi insurance and o design i wih higher coverage raio. Bu he effec of poliical openness on coverage raio is weak significanly afer adoping a sysem of explici deposi insurance. (Noe 3) Table 3. Panel regression resuls of he deerminans of deposi insurance coverage Variable Models (1) (2) (3) (4) (5) (6) POLITY (0.000) *** (0.000) *** DEMOC (0.022) ** (0.035) ** AUTOC (0.000) *** (0.000) *** SAVINGS (0.582) (0.379) (0.893) (0.719) (0.516) (0.802) INT_SPREAD (0.000) *** (0.000) *** (0.000) *** DEP_INT (0.003) *** (0.002) *** (0.005) *** GOV_DEBT (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** Inercep (0.469) (0.063) * (0.332) (0.277) (0.027) ** (0.543) Hausman Tes (0.003) *** (0.001) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** No. of Obs Noes: The dependen variable GOV_GDP is he raio of coverage o per-capia GDP (refer o Table 2 for he definiions of he independen variables); figures in parenheses are p-values. * indicaes significance a he 10% level; ** indicaes significance a he 5% level; and *** indicaes significance a he 1% level. The null hypohesis of he Hausman es is for random effecs. In addiion o he discussion on he relaionship beween poliical openness and coverage, Laeven (2002) indicaed ha higher deposi insurance coverage would also resul in higher coss (premiums). Zarruk and Madura (1992) found ha counries wih higher deposi insurance coss would invariably have lower ineres rae spreads, hereby supporing our finding ha counries wih lower ineres rae spreads generally have higher levels of coverage. The panel regression furher indicaes a negaive correlaion beween he deposi ineres rae and coverage. The finding is inconsisen wih Demirgüç-Kun and Deragiache (1998) because hey suggesed ha nominal ineres rae is highly significan in increasing banking secor fragiliy. To preven insabiliy and panic, mos counries end o have higher coverage o ensure ha heir deposiors are relaively safe, and he coefficiens of governmen deb are significanly posiive. This means ha counries wih higher governmen deb levels end o have lower credibiliy and are associaed wih relaively greaer risk. Thus, coverage ends o be higher for counries wih higher governmen deb levels. In summary, we found ha counries wih less poliical openness, lower ineres rae spreads, lower deposi ineres rae, and higher governmen deb end o have higher coverage. This is consisen wih he findings in prior lieraure ha counries wih deposi insurance sysems end o suffer from financial insabiliy. (Noe 4) Published by Sciedu Press 17 ISSN E-ISSN

6 Inernaional Journal of Financial Research Vol. 5, No. 4; Probabiliy Deerminans of Changing Deposi Insurance Coverage This work adoped he panel logi model o deermine facors ha affec he probabiliy of changes in coverage. As shown in Table 1 Panel B, counries wih boh upper-middle- levels (39%) and lower-middle- levels (34%) end o change heir coverage more ofen han hose wih eiher high- levels (13%) or low- levels (19%). The panel logi model resuls presened in Table 4 reveal ha he GOV_DEBT variable shows a significan decrease in he probabiliy of a change in coverage. The INT_SPREAD and DEP_INT variables have a significan posiive effec, which means ha counries wih higher ineres rae spread or higher deposi ineres rae have a higher probabiliy of changing heir coverage level. We ake he resul of Model (1) in Table 4 as an example for our examinaion of he marginal effec on change in coverage. By applying he coefficiens ino he panel logi model, we obain he following equaion: L P ln POLITYi, SAVING 1 P INT _ SPREAD GOV _ DEBT u Thereafer, we can also calculae he incremen in he probabiliy of change (P ) for a single uni increase in he GOV_DEBT variable. The calculaion is expressed as follows: P 1 P e 0.027GOV _ DEBT e P The marginal effec is deermined as 1 P i, ( 1 Pi, ) Table 4. Panel logi model resuls of he CHANGE variable Variable Models (1) (2) (3) (4) (5) (6) POLITY (0.053) * (0.227) DEMOC (0.147) (0.550) AUTOC (0.021) ** (0.055) * SAVINGS (0.281) (0.249) (0.266) (0.471) (0.416) (0.485) INT_SPREAD (0.000) *** (0.000) *** (0.000) *** DEP_INT (0.000) *** (0.000) *** (0.000) *** GOV_DEBT (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** Inercep (0.006) *** (0.009) *** (0.094) * (0.008) *** (0.018) ** (0.047) ** Wald Chi-Square (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** No. of Obs Noes: The dependen variable CHANGE is a dummy variable which is equal o 1 if here has been a change in coverage as compared o he previous year; oherwise zero (refer o Table 2 for he definiions of he independen variables); figures in parenheses are p-values. * indicaes significance a he 10% level; ** indicaes significance a he 5% level; and *** indicaes significance a he 1% level. (3) (4) This can be inerpreed as a single uni increase in governmen deb, leading o a decrease in he probabiliy of a change in coverage. Based on our daa, we observe ha a change in coverage ypically involves an increase in he Published by Sciedu Press 18 ISSN E-ISSN

7 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 level of coverage. From all he observaions of changes in he level of deposi insurance coverage examined in his sudy, only 18 resuled in any reducion in insurance coverage. We inerpre his o mean ha an increase in governmen deb reduces he probabiliy of any increase in coverage. According o he IADI, (Noe 5) relying on he governmen o suppor a higher coverage level creaes a fiscal burden for he governmen. Because Table 3 shows ha counries wih higher governmen deb levels already have higher coverage raios, i is no surprising ha Table 4 shows hese counries o have a lower probabiliy of raising heir coverage levels because heir fiscal burdens are already oo high. 3.3 Deerminans of Change in Deposi Insurance Coverage Table 5 shows he logi model of he probabiliy of changes in coverage on he changes in independen variables. An examinaion of facors likely o affec he probabiliy of changes in coverage shows ha an increase in changes of ineres rae spread (d_int_spread), deposi ineres rae (d_dep_int), and governmen deb (d_gov_debt) ends o reduce he probabiliy of any increase in coverage. Table 5. Logi model of changes in deposi insurance coverage Variable Models (1) (2) (3) (4) (5) (6) d_polity (0.935) (0.756) d_democ (0.995) (0.612) d_autoc (0.853) (0.974) d_savings (0.588) (0.693) (0.529) (0.592) (0.699) (0.517) d_int_spread (0.024) ** (0.024) ** (0.024) ** d_dep_int (0.027) ** (0.026) ** (0.028) ** d_gov_debt (0.357) (0.342) (0.384) (0.002) *** (0.002) *** (0.002) *** Inercep (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** (0.000) *** Wald Chi-Square (0.019) ** (0.019) ** (0.019) ** (0.008) *** (0.008) *** (0.008) *** No. of Obs Noes: The dependen variable CHANGE is a dummy variable which is equal o 1 if here has been a change in coverage as compared o he previous year; oherwise zero (refer o Table 2 for he definiions of he independen variables); figures in parenheses are p-values. * indicaes significance a he 10% level; ** indicaes significance a he 5% level; and *** indicaes significance a he 1% level. This sudy also conduced a random effecs panel regression o examine changes in independen variables arising from changes in he level of deposi insurance coverage (d_cov_gdp). Table 6 indicaes ha only changes in ineres spreads (d_int_spread) have negaive effecs on changes in coverage once ime-invarian counry heerogeneiy is accouned for, which is consisen wih he resuls repored in Table 3 and Table 5. This indicaes he srongly negaive significance of he relaionship beween changes of ineres rae spreads and changes of coverage over per capia GDP, which is consisen wih he repor by Zarruk and Madura (1992). Counries wih lower ineres rae spread (INT_SPREAD) have higher levels of coverage and a lower probabiliy of changing coverage. Counries increase changes of deposi insurance coverage and he probabiliy of changes in coverage when hey decrease ineres rae spreads (d_int_spread). Published by Sciedu Press 19 ISSN E-ISSN

8 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 Table 6. Panel regression of changes in influenial facors Variable Models (1) (2) (3) (4) (5) (6) d_polity (0.807) (0.813) d_democ (0.595) (0.657) d_autoc (0.862) (0.949) d_savings (0.341) (0.323) (0.359) (0.792) (0.775) (0.809) d_int_spread (0.000) *** (0.000) *** (0.000) *** d_dep_int (0.510) (0.512) (0.518) d_gov_debt (0.681) (0.665) (0.693) (0.639) (0.627) (0.646) Inercep (0.385) (0.389) (0.368) (0.488) (0.491) (0.481) Hausman Tes (0.463) (0.354) (0.547) (0.702) (0.630) (0.761) No. of Obs Noes: The dependen variable d_gov_gdp is he changes raio of coverage o per-capia GDP (refer o Table 2 for he definiions of he independen variables); figures in parenheses are p-values. * indicaes significance a he 10% level; ** indicaes significance a he 5% level; and *** indicaes significance a he 1% level. The Hausman Tes is for random effecs. The null hypohesis of F Tes is no fixed effecs 4. Conclusion This sudy focuses on finding he relaionship beween economic and poliical facors and he level of coverage, and deermines facors ha affec he probabiliy of changes in coverage. Using panel daa, we found ha counries wih lower ineres rae spreads, lower deposi ineres rae, and higher governmen deb levels end o have higher coverage. Democraic (auocraic) counries also end o have lower (higher) coverage levels. Our more complee daase enabled us o reach findings ha were unavailable o Laeven. From he panel logi analysis, we found ha counries wih lower poliical openness, lower ineres rae spreads and deposi ineres rae, and higher governmen deb reduce he probabiliy of changes in coverage. A decrease in ineres rae spreads, deposi ineres rae, and governmen deb increases he probabiliy of changes in coverage. Finally, only changes in ineres rae spreads can explain changes in coverage afer accouning for ime-invarian counry heerogeneiy, which is anoher significan difference beween our paper and he Laeven sudy. In summary, our findings provide a useful reference for regulaors in seing coverage levels. References Angkinand, A. (2005). Deposi insurance and financial crises: Invesigaion of he cos benefi radeoff. Working Paper, Claremon Cenre for Economic Policy Sudies, Claremon Graduae Universiy. Chu, K. H. (2003). Deposi insurance and banking crises in he shor and long run. Cao Journal, 32, Cull, R., Senbe, L. W., & Sorge, M. (2005). Deposi insurance and financial developmen. Journal of Money Cred and Banking, 37, hp://dx.doi.org/ /mcb Demirgüç-Kun, A., & Deragiache, E. (1998). The deerminans of banking crises in developing and developed counries. Inernaional Moneary Fund Saff Papers, 45, hp://dx.doi.org/ / Demirgüç-Kun, A., & Deragiache, E. (2002). Does deposi insurance increase banking sysem sabiliy? An empirical invesigaion. Journal of Moneary Economics, 49, hp://dx.doi.org/ /s (02)00171-x Demirgüç-Kun, A., Kane, E. J., & Laeven, L. (2008). Deerminans of deposi-insurance adopion and design, Published by Sciedu Press 20 ISSN E-ISSN

9 Inernaional Journal of Financial Research Vol. 5, No. 4; 2014 Journal of Financial Inermediaion, 17, hp://dx.doi.org/ /j.jfi Hausman, J. A. (1978). Specificaion ess in economerics. Economerica, 46, hp://dx.doi.org/ / Huchison, M., & McDill, K. (1999). Are all banking crises alike? The Japanese experience in inernaional comparison. Journal of he Japanese and Inernaional Economies, 13, hp://dx.doi.org/ /jjie Laeven, L. (2002). Pricing of deposi insurance. Working paper, World Bank Policy Research Working Paper No hp://dx.doi.org/ / Laeven, L. (2004). The poliical economy of deposi insurance. Journal of Financial Services Research, 26, hp://dx.doi.org/ /b:fina e1 Sandholz, W. & Koezle, W. (2000). Accouning for corrupion: Economic srucure, democracy, and rade. Inernaional Sudies Quarerly, 44, hp://dx.doi.org/ / Zarruk, E. R., & Madura, J. (1992). Opimal bank ineres margin under capial regulaion and deposi insurance. Journal of Financial and Quaniaive Analysis, 27, hp://dx.doi.org/ / Noes Noe 1. IADI websie hp:// Noe 2. Demirgüç-Kun e al. (2008) used 181 counries during he period o idenify facors ha influence decisions abou a counry s adoping deposi insurance. Of hese counries, 87 have adoped explici deposi insurance. We used hose 87 counries daa and exended he sample period hrough 2008 o incorporae daa on recen adopers. Noe 3. In page 433 of Demirgüç-Kun e al. (2008), he coefficien of Poliy score in model (1) is and significan a 10%. Noe 4. Huchison and McDill (1999), Demirgüç-Kun and Deragiache (2002), Chu (2003), and Cull e al. (2005). Noe 5. Inernaional Associaion of Deposi Insurers (IADI) websie hp:// Published by Sciedu Press 21 ISSN E-ISSN