Atlantis Press Journal style Copyright © 2018, the Authors. Published by Atlantis Press. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). Research on the Development of County Finance in Guizhou Province in the Promotion of Precise Poverty Alleviation Yu Ding, Mu Zhang School of Finance, Guizhou University of Finance and Economics, Guiyang Guizhou E-mail: 13399855052@163. com, rim_007@163. com Received December 16, 2017 Accepted January 28, 2018 Abstract This paper selects the 50 state-level poverty-stricken counties in Guizhou province as the research object, and uses financial scale, financial efficiency and financial structure to represent the level of financial development in each county, using economic growth and income distribution as controlled variable. The poverty of every county is expressed by poor slow index. Applying the panel data model, the promotion of local financial development to the targeted poverty alleviation is studied. The empirical evidence shows that the financial scale, financial efficiency, financial structure, economic growth and poverty reduction of 50 national poverty-stricken counties in Guizhou province are positively correlated. However the financial scale is more significant to reduce the incidence of poverty than that of financial efficiency. Keywords: Financial development, Precision for poverty alleviation, Panel data model. 1. Introduction As research object, Guizhou is the largest and most impoverished province in China with the largest and the deepest poverty. For years, it follows policy steps. According to the latest statistics from Guizhou Statistics Bureau, there are still as many as 50 poverty-stricken counties in Guizhou province. So there is still a long way to go. In 2014, President Xi Jinping put forward the concept of "Precision for poverty alleviation". This concept has become the basic strategy of the national poverty alleviation work. Finance is a resource which can configuring other resources. Its total allocation and allocation efficiency directly determine the level of economic development. Under the current situation of poverty in Guizhou and the policy of "Precision for poverty alleviation", how to combine financial development to promote poverty alleviation is worth discussing. 2. Literature review Set up fixed effects vector decomposition model with 23 provinces 2011-2008 data. It has shown that the rural financial development can significantly alleviate poverty, directly and indirectly. Instability in the process of the rural financial development has no significant impact on poverty (Tan, 2011) [1] . Select the financial development scale and the financial development efficiency as financial development indicators. Through empirical test for Chinese provincial panel data: financial development can through economic growth, income distribution channels to raise the income level of the poor. But the financial wobbles will offset the effect of financial development of poverty reduction (Cui and Sun, 2012) [2] . 2001 to 2010, China's rural financial development is negatively related to the incidence of poverty (Wu, 2012) [3] . The plight of the rural financial poverty alleviation include: poor credit environment, poor farmers lack of mortgage and financial institutions development is difficult to meet the demand of expanding capital , and regulatory standards hinder financial institutions (Guo, 2013) [4] . Using the credit risk model analyze the problems of rural credit. Credit risk uncertainty even more difficult for the rural registered permanent residence for credit funds, to improve the credit rationing, intensify 52 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 financial poverty alleviation and put forward reasonable Suggestions(Song, Li and Xiao, 2017) [5] . Analyze Sichuan Bazhong 3 counties 1 district nearly four years of panel data. The increase of agricultural loans accounted does not improve the farmers' income. In addition, from the perspective of the financial poverty alleviation efficiency index of building, the rural enterprise loan poverty alleviation efficiency during the sample period is on the rise (Deng, 2015) [6] . Financial development on poverty alleviation role of literature research accurately, but as a serious poverty situation of Guizhou province, the current study is seldom the county financial development in Guizhou of precise role in promoting research for poverty alleviation. For this reason, this paper uses panel data model for 2013 to 2015, 50 state-level poverty-stricken counties in Guizhou financial development study of precise role for poverty alleviation. 3. Linear regression model of panel data 3.1. Panel data Panel data refers to the fixed a group of subjects in the interval of time such as continuous observation data, with cross section and time are the two characteristics of the data. Under the panel data in double scalar said, for example , 1, 2, , ; 1, 2, , it Y i N t T  where i corresponding to the panel data in different individuals, N said panel data of individual number. t for panel data in different time, T says the maximum length of time sequence. If the fixed t unchanged, Yi. , i=1, 2, … , N, is in the cross section of N random variables; If i fixed constant, Y. t, t = 1, 2, . . . , T, is the profile of a time series. Panel data is divided into two kinds of characteristics. Is a section on individual number is little, and each individual time span is long. Secondly, the cross section on individual number, and each individual short time span. 3.2. Panel data model Based on panel data regression model is called the panel data model, usually distinguish between linear and nonlinear etc. For convenience, in this paper, the linear panel data regression model is written as: where Yit is interpreted variable for individual values that i at time t, xhit for the hth explained variable for individual values that I at time t; β hit for the hth explained variable be estimated parameters; uit is random error term. Panel model is usually divided into three categories, namely the hybrid model, the fixed effect model and random effect model. Panel model is usually divided into three categories, namely, hybrid model, fixed effect model and random effect model. In one of the fixed effects model parameters  it  are fixed, the random error term said ignored or changes in the individual, at any time and in a given the observation effect and explanatory variables under the condition of the expectation of a random error term is equal to zero, uit homoscedasticity, different individuals and different point corresponding uit are independent of each other. The model forms as follows: There is only individual effect in the simulation, the model in the form of: i  as a stochastic variable to describe the differences between different individuals establish regression function. Because i is invisible, and with the change of the interpretation of the observed variables Xhit associated, so called individual fixed effects model. If time effect exists only in the model, the model in the form of: i  is a random variable, said for N individual item has N different intercept, and its change with Xit, t is a random variable, said to T section (time) of T different intercept, and its change associated with Xit, says this model for individual point double fixed effects model. 4. Empirical analysis 4.1. Model, index and data 4. 1. 1. Model setting The purpose of this paper is to study the financial development on precision of poverty alleviation, choose ithit K h hit ux   1 tt Y  ithit K h hiit ux   1 Y  ithit K h hit ux   1 t Y  T,…1,2,tN;,…,2,1,Y 1     iux ithit K h hititit  53 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 some representative index to indicate the level of financial development in Guizhou, and poverty slow index is used to represent each county poverty. Based on the assumptions about financial development has a positive promoting effect on economic growth, mainly through economic growth and income distribution influence the slow down of poverty. Therefore, this article selects three factors as the empirical equation of the independent variable, the level of financial development, economic rights respectively, the income distribution. Among them, we will measure the financial development level from the financial scale, financial efficiency and financial structure. In addition, there are many factors to reduce poverty, such as government investment, residents' level of education, etc., these factors we unified with the random error term. Get the following basic model: POV said poverty. FS said financial scale. FE said financial efficiency and FSR said financial structure. RGP said the rise of the economy. IG said income distribution. μsaid the other factors that affect poverty to slow, i = 1, 2, 3. . . 50, said 50 key poverty alleviation and development counties in Guizhou; t = 1, 2, 3, according to different years. 4. 1. 2 Selection of indicators Slow poverty indicators (POV): in this paper, the poverty rate is used to measure the degree of poverty. The greater value of the index, the more serious the poverty situation. Scale of financial indicators (FS): the total amount of lending and deposit/GDP as the index of financial scale. The greater the value of the index, it shows that the bigger financial system development and the higher the level of financial development. Financial structure index (FSR): the existence of economies industry capital structure, is one of the important indicators to measure the financial structure, it is equal to direct financing than indirect financing. According to the situation of county of Guizhou province directly, equity financing is the main way of the enterprise financing, bond and other financing way development is not perfect, once again ignored. And since the province a total of 26 listed companies in Guizhou, of which only two located in the state-level counties list, so this article uses the 0-1 programming to deal with variable, with the county value of listed companies to "1", the rest of the values for "zero". Financial efficiency index (FE): use of the whole society fixed assets investment and the banking financial institutions balance to calculate. The greater value of the index, the higher the efficiency of financial institutions to invest in, and the higher the financial efficiency. Economic rights long index (RGP): there are many indicators to measure economic growth. Based on the GDP per capita to measure economic growth, if the indicator into a rising trend, it means economic grow and development well. Income distribution index (IG): considering the comparability and data item, the urban per capita disposable income and rural per capita net income of is adopted to measure the equality of income distribution. The index's rise in value, the income gap of urban and rural residents is bigger and bigger, the income distribution is unfair. The modeling of variable table is shown in Table 1. Table 1. the modeling of variable table types of variables variable name variable definition explained variable POV Slow poverty indicators explanatory variable FS Scale of financial indicators FSR Financial structure index FE Financial efficiency index control variable RGP Economic rights long index IG Income distribution index 4. 1. 3. Data source and description According to the latest statistics yearbook, there are 50 state-level poverty-stricken counties in Guizhou. This paper collected data from 2013 to 2015, 50 state- level counties. It has the same statistical caliber. And it selected data with high reliability. Data mainly comes from from 2013 to 2013. It used excel deal with some simple data. The original data are shown in Table 2. ititititititiit IGRGPFSRFEFSCPOV   54 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 Table 2. 50 state-level poverty-stricken counties in Guizhou in almost three years of relevant indicators data tables. Area Year POV FS FE FSR RGP IG Liuzhi 2013 0. 199712454 1. 561988906 1. 261935037 0 2. 037062147 3. 271872998 2014 0. 1698 1. 360806148 1. 300309598 0 2. 5648 3. 012648345 2015 0. 139 1. 323821759 1. 50255144 0 2. 9673 2. 996302616 Shuicheng 2013 0. 28859804 4. 74945391 0. 514804652 0 1. 861926518 3. 320603539 2014 0. 2481 5. 284909723 2. 036130484 0 2. 4496 3. 06296875 2015 0. 21 5. 232148917 2. 249188423 0 2. 7828 3. 019953808 Pan 2013 0. 2349 0. 984489981 1. 184473418 0 3. 504879738 3. 150674731 2014 0. 1879 0. 909242318 1. 267221972 0 4. 0896 2. 90451464 2015 0. 143 1. 015130525 1. 377510969 0 4. 5397 2. 886136861 Zhengan 2013 0. 21617053 2. 302466037 0. 568922751 0 1. 135984523 3. 457081445 2014 0. 1638 2. 04698211 0. 673866091 0 1. 5592 2. 943560271 2015 0. 151 2. 339418396 1. 056572026 0 1. 847450404 2. 922222222 Daozhen 2013 0. 196069861 2. 469862757 0. 537605587 0 1. 318392271 3. 611835045 2014 0. 1636 2. 370314465 0. 569141755 0 1. 6275 2. 939867354 2015 0. 116 2. 37884185 0. 630391482 0 1. 970250348 2. 921421025 Wuchuan 2013 0. 235257552 2. 573050159 0. 675316875 0 1. 085231143 3. 675471298 2014 0. 1874 2. 461270063 0. 777029961 0 1. 3473 2. 944797987 2015 0. 137 2. 78176 1. 033144446 0 1. 609810851 2. 931643217 Xishui 2013 0. 228370636 1. 597287111 0. 7804341 0 1. 758006561 3. 549972459 2014 0. 1778 0. 905375389 0. 650641026 0 2. 1734 2. 93010449 2015 0. 132 1. 692529881 1. 205118422 0 2. 466186335 2. 914234739 Puding 2013 0. 254689753 1. 55787234 1. 815910393 0 1. 5692 3. 323174482 2014 0. 2077 1. 479775281 1. 932032301 0 1. 876 3. 280739045 2015 0. 151 1. 68896147 2. 166130624 0 2. 1876 3. 207245787 Zhenning 2013 0. 25737951 2. 154028555 2. 12146189 1 1. 8314 3. 371488108 2014 0. 2176 1. 877455566 2. 205821206 1 2. 2544 3. 317306081 2015 0. 187 1. 971832866 2. 514094518 1 2. 631 3. 287148753 Guanling 2013 0. 286480071 1. 595727092 1. 00096707 0 1. 6678 3. 340072553 2014 0. 2416 1. 36863711 1. 073092882 0 2. 0134 3. 275810224 2015 0. 18 1. 52686863 1. 337890201 0 2. 4092 3. 234344272 Ziyun 2013 0. 286838134 1. 802484402 1. 116674748 0 1. 3071 3. 308348413 2014 0. 2427 1. 640382052 1. 24827883 0 1. 6589 3. 2834015 2015 0. 193 1. 809332557 1. 449481813 0 1. 8985 3. 180335085 Dafang 2013 0. 277637848 1. 084636195 0. 800785621 0 1. 667343057 3. 318108889 2014 0. 2351 0. 971021797 0. 801148796 0 2. 0372 3. 274121406 2015 0. 191 1. 079245957 0. 949919431 0 2. 289736958 3. 20050014 Zhijin 2013 0. 319547534 1. 536880811 1. 072905871 0 1. 371192555 3. 459868635 2014 0. 2732 1. 432636597 1. 310437236 0 1. 7361 3. 314683053 2015 0. 225 1. 786448195 1. 679557145 0 1. 957270049 3. 24609209 Nayong 2013 0. 268517252 0. 943353851 0. 991139212 0 1. 837045083 3. 513125884 2014 0. 2239 0. 902212657 1. 141195031 0 2. 2095 3. 472671548 2015 0. 181 0. 997070185 1. 412794795 0 2. 535364984 3. 397934332 Weining 2013 0. 240462341 0. 911448326 0. 939779449 0 0. 976738675 3. 315709469 2014 0. 1863 0. 896862461 1. 021591949 0 1. 1992 3. 277275662 2015 0. 145 0. 965618926 1. 062023815 0 1. 488182555 3. 18981333 55 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 Continued Table 2. Area Year POV FS FE FSR RGP IG Hezhang 2013 0. 260545803 1. 28043682 0. 599641977 0 1. 01379099 3. 501040012 2014 0. 2177 1. 091468101 0. 737974962 0 1. 3937 3. 460645382 2015 0. 179 1. 153913071 0. 85954823 0 1. 711211481 3. 367844364 Jiangkou 2013 0. 235026029 2. 601040087 1. 148558961 0 1. 6017 3. 382943551 2014 0. 1878 2. 463494913 1. 429294756 0 1. 9289 3. 26176566 2015 0. 138 2. 377842143 1. 579488811 0 2. 385833573 3. 285654703 Shiqian 2013 0. 258555943 2. 046813905 0. 888954707 0 1. 2592 3. 352767933 2014 0. 212 1. 942521502 1. 046206353 0 1. 5657 3. 190305791 2015 0. 165 2. 116778523 1. 303385894 0 1. 954739259 3. 184238685 Sinan 2013 0. 254316479 2. 028960087 1. 158156172 0 1. 4778 3. 535193549 2014 0. 2132 1. 93537415 1. 168472614 0 1. 7667 3. 279853235 2015 0. 164 2. 042595442 1. 307985486 0 2. 028211285 3. 269407744 Yingjiang 2013 0. 236392835 2. 09458126 1. 074512282 0 1. 7001 3. 448269353 2014 0. 1913 1. 926498048 1. 223401746 0 2. 0722 3. 235050522 2015 0. 143 1. 9746699 1. 521337161 0 2. 606954689 3. 272047349 Dejiang 2013 0. 28802056 1. 438414918 1. 078973324 0 1. 6341 3. 694211828 2014 0. 2579 1. 407503411 1. 255995961 0 1. 9852 3. 329270357 2015 0. 202 1. 507307556 1. 422050628 0 2. 280744337 3. 312446219 Yanhe 2013 0. 271253484 1. 646139756 0. 786537841 0 1. 3106 3. 544692896 2014 0. 2295 1. 50261708 0. 806675939 0 1. 6125 3. 321644295 2015 0. 17 1. 61867534 0. 888931775 0 1. 876595272 3. 276332715 Songtao 2013 0. 233279424 1. 756375498 0. 948684227 0 1. 4971 3. 524577178 2014 0. 1889 1. 529517598 0. 958264326 0 1. 8223 3. 293322063 2015 0. 142 1. 667713871 1. 150993333 0 2. 113165152 3. 299930293 Xingren 2013 0. 207 1. 46647455 1. 57416672 0 1. 8413 3. 352642451 2014 0. 1612 1. 405961057 1. 807336957 0 2. 1951 3. 280693459 2015 0. 106 1. 484808227 2. 164888052 0 2. 700672571 3. 239052745 Puan 2013 0. 196259181 1. 568819818 1. 17560454 0 1. 6741 3. 671681958 2014 0. 1517 1. 499900892 1. 381349206 0 1. 9705 3. 486646884 2015 0. 105 1. 551480865 1. 519460678 0 2. 338521401 3. 4370149 Qinglong 2013 0. 374083742 1. 402714016 0. 889113754 0 1. 4956 3. 80012454 2014 0. 3277 1. 334983127 0. 948440208 0 1. 7919 3. 636463482 2015 0. 257 1. 373372342 1. 225945898 0 2. 223100806 3. 521828906 Zhenfeng 2013 0. 25616614 1. 420402749 0. 83911166 0 2. 0521 3. 416154374 2014 0. 2065 1. 397700549 0. 990508832 0 2. 5059 3. 278937072 2015 0. 161 1. 331254098 1. 184476412 0 2. 981321802 3. 278937072 Wangmo 2013 0. 3302 2. 19442433 1. 196548483 0 1. 0232 4. 049446524 2014 0. 2782 1. 955752212 1. 293567894 0 1. 4544 3. 718786016 2015 0. 212 1. 842509468 1. 4710172 0 1. 90233347 3. 659548717 Ceheng 2013 0. 327299562 2. 392394123 1. 43018208 0 1. 2101 3. 864543206 2014 0. 2917 2. 218577348 1. 276337562 0 1. 5128 3. 677299011 2015 0. 209 1. 92618653 1. 38313203 0 1. 922140992 3. 628398759 Anlong 2013 0. 18765398 1. 399533825 0. 671555203 0 1. 7737 3. 439123678 2014 0. 144 1. 383162218 0. 720708447 0 2. 0314 3. 289023989 2015 0. 094 1. 762609547 1. 121522008 0 2. 450489025 3. 297945393 56 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 Continued Table 2. Area Year POV FS FE FSR RGP IG Huangping 2013 0. 300400356 1. 986044487 0. 551266679 0 1. 18148924 3. 757235792 2014 0. 279 1. 881645739 0. 686358447 0 1. 4245 3. 543672627 2015 0. 23 2. 060513209 0. 80948955 0 1. 6621 3. 496015936 Shibing 2013 0. 315 1. 623027167 0. 901999773 0 1. 776330908 3. 314560106 2014 0. 2681 2. 065495792 0. 571290634 0 2. 091 3. 139756811 2015 0. 219 2. 076375092 0. 755611408 0 2. 3598 3. 12852186 Sansui 2013 0. 337912808 1. 770186807 0. 721300959 0 1. 607073955 3. 540659015 2014 0. 2909 1. 628367413 0. 851790175 0 1. 9833 3. 352302205 2015 0. 238 1. 855583911 0. 958632101 0 2. 2779 3. 307514784 Chenggong 2013 0. 263671106 1. 911758775 0. 901263781 0 1. 661290323 3. 541658895 2014 0. 2539 1. 759073842 0. 946095918 0 1. 9863 3. 347981771 2015 0. 199 2. 246878817 1. 348102786 0 2. 2779 3. 297069916 Tianzhu 2013 0. 317560422 1. 73507733 0. 54547008 0 1. 887096774 3. 398569493 2014 0. 2765 1. 647038917 0. 588824354 0 2. 2678 3. 219562136 2015 0. 225 1. 680926742 0. 697734382 0 2. 5857 3. 18216399 Jinping 2013 0. 327482746 2. 048268621 0. 538821123 0 1. 643229167 3. 826978339 2014 0. 2854 1. 954845089 0. 579051383 0 1. 9746 3. 58334794 2015 0. 237 2. 128669998 0. 670673218 0 2. 2632 3. 531924883 Jianhe 2013 0. 324845698 1. 77445503 0. 69124272 0 1. 452532521 3. 734608888 2014 0. 2829 1. 718729955 0. 728682171 0 1. 725 3. 52583691 2015 0. 237 1. 858789305 0. 845340886 0 1. 9953 3. 488044145 Taijiang 2013 0. 320123134 2. 614989224 1. 688534946 0 1. 671319226 3. 90386029 2014 0. 2878 2. 337837838 1. 781118461 0 1. 9982 3. 641577061 2015 0. 199 3. 097484388 2. 298387097 0 2. 2931 3. 602333495 Liping 2013 0. 273840105 2. 167047478 0. 798506466 0 1. 213165338 3. 651010754 2014 0. 2448 1. 984454148 0. 896921017 0 1. 4719 3. 445584525 2015 0. 225 2. 084132192 1. 054982636 0 1. 73 3. 442538333 Rongjiang 2013 0. 362219501 2. 158622636 1. 073681581 0 1. 236649215 3. 765490043 2014 0. 3373 1. 974154589 1. 063465319 0 1. 446 3. 53069719 2015 0. 276 2. 153062949 1. 326029216 0 1. 6868 3. 48019802 Congjiang 2013 0. 316824119 1. 409779646 0. 963978359 0 1. 185019756 3. 532828416 2014 0. 2901 1. 343803263 0. 922477441 0 1. 4128 3. 345739471 2015 0. 239 1. 439978029 0. 993257164 0 1. 6614 3. 307121014 Leishan 2013 0. 279830763 2. 037258157 0. 940628209 0 1. 495079161 3. 563792528 2014 0. 2648 2. 126787417 1. 006111111 0 1. 7947 3. 361312665 2015 0. 208 2. 337487647 1. 121662182 0 2. 0324 3. 313362702 Majiang 2013 0. 376823379 1. 933068536 0. 921480557 0 1. 545126354 3. 713156697 2014 0. 3276 2. 3109319 0. 931235955 0 1. 828 3. 510221465 2015 0. 206 2. 828893229 1. 343370672 0 2. 1728 3. 447252245 Danzhai 2013 0. 337322041 2. 374553522 0. 75112077 0 1. 474846626 3. 728992743 2014 0. 2924 2. 284910965 0. 835390947 0 1. 7435 3. 506482429 2015 0. 2381 2. 410324531 1. 003325346 0 1. 9967 3. 456406368 Libo 2013 0. 309227981 1. 876319487 1. 005916103 0 2. 404984314 3. 238336279 2014 0. 2556 1. 793926247 1. 204545455 0 3. 2619 3. 044679005 2015 0. 2109 1. 970218543 1. 403795148 0 3. 562 3. 030951868 57 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 The basic statistical characteristics of the sample are shown in Table 3. Table 3. The basic statistical characteristics of the sample Variable Mean Value Standard deviation Minimum Value Maximum Value pov 0. 2364972 0. 0614049 0. 094 0. 3768234 fs 1. 840429 0. 6395251 0. 8968624 5. 28491 fe 1. 120168 0. 3989469 0. 5148047 2. 514095 fsr 0. 04 0. 1966157 0 1 rgp 1. 915396 0. 545185 0. 9767387 4. 5397 ig 3. 338166 0. 2438435 2. 847902 4. 049447 4.2. Empirical test process 4. 2. 1. Fixed effect model test Poverty rate as the dependent variable is influenced by independent variables in the model: financial scale, financial efficiency and financial structure, the influence of the economic growth and income distribution, also will be affected by other factors associated with particular observation object or period. Fixed effects model, the comprehensive effects of other factors as fixed, to a certain extent, improved the fitting degree of practical data. Key county of Guizhou province 50 poverty alleviation and development, is the main part of the relatively independent and unified individuals. In order to reflect the differences of the individuals in the whole, this paper first uses the fixed effect variable intercept model, and regression results are shown in Table 4. Table 4. Fixed effect model regression results POV estimate of parameter standard error P>t FS -0. 05774 0. 0110099 0 FE -0. 01377 0. 0074121 0. 06 FSR omitted RGP -0. 11729 0. 0063662 0 IG 0. 02139 0. 0139213 0. 12 C 0. 51337 0. 0616472 0 The result is visible in Table 4 that financial structure is omitted. It means the variables are multicollinearity. Therefore, remove the variable from the model. The main reason for the financial structure appeared multicollinearity is due to the financial structure in calculating the index of listed companies, and listed companies in 50 state-level poverty-stricken counties in Guizhou was small. So data is not ideal, and finally this paper had no way but to remove the index of financial structure. Continued Table 2. Area Year POV FS FE FSR RGP IG Dushan 2013 0. 273012006 1. 97870742 0. 790070241 0 1. 621393159 3. 201440771 2014 0. 2273 1. 766224323 0. 826587038 0 2. 0422 2. 997296913 2015 0. 1759 2. 087012568 1. 127719347 0 2. 3035 2. 962288054 Pingtang 2013 0. 315138198 1. 642500785 0. 996803011 0 1. 357098566 3. 260047272 2014 0. 2598 1. 514265503 0. 988072336 0 1. 776 3. 049054905 2015 0. 1955 2. 024569926 1. 483892488 0 1. 9885 3. 02705992 Luodian 2013 0. 3551 1. 673145121 1. 121704211 1 1. 594785095 3. 051587356 2014 0. 3027 1. 508854782 1. 214974913 1 1. 9817 2. 863407821 2015 0. 2416 1. 675535871 1. 4811593 1 2. 2175 2. 847902003 changshun 2013 0. 318250439 1. 460728353 0. 976326357 0 1. 716578884 3. 180008698 2014 0. 2656 1. 303118202 1. 270614278 0 2. 2211 2. 979166667 2015 0. 2052 1. 736274723 1. 513155201 0 2. 4685 2. 930855856 Sandu 2013 0. 356204934 1. 881020408 1. 053421689 0 1. 199058579 3. 203770577 2014 0. 323 1. 64083219 1. 074776386 0 1. 6361 2. 984166913 2015 0. 2526 1. 923600327 1. 260590897 0 1. 8281 2. 959970965 58 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 4. 2. 2. Delete the common linear variables (1) Hausman test Remove the financial structure variables from original data. Use Stata11. 0 software to do Hausman test . The results are shown in Table 5. Table 5. Hausman test result Hausman test Chi(0)=0. 00 Prob>chi2=0. 00 Due to the P value is 0. 00, so strongly rejected the null hypothesis, it should use a fixed effect model to estimate more appropriate. (2) Fixed effect model regression results Deleted the financial structure from the original data , parameter estimation using Stata11. 0 software, the results as shown in Table 6. Table 6. The regression results of the fixed effect model after the total linear variables were deleted POV estimate of parameter standard error P>t FS -0. 05830 0. 01101 0 FE -0. 01388 0. 00742 0. 064 FSR -0. 11753 0. 00637 0 RGP 0. 02141 0. 01394 0. 128 C 0. 51502 0. 06168 0 From Table 6 shows that financial scale, financial efficiency and economic growth indicators of poverty- stricken counties in Guizhou's poverty rate is negatively related. It illustrate the three indicators can slow the poor development of state-level poverty-stricken counties in Guizhou. It accords with the real economic significance. The greater the income distribution gap, the higher the poverty rate in poverty-stricken counties in Guizhou. Both of them were positively correlated. Financial scale, financial efficiency and economic growth indicators of P values are less than 0. 05, which is under the 95% confidence level were significantly. P value of income distribution was 0.128, and it shown that in 12. 8% of cases of income distribution have no effect on state-level poverty-stricken counties in Guizhou's slow. It failed to pass the test of significance. So, not significant items will be deleted from the model. Income distribution is too poor to slow. The reason mainly is income distribution index using urban per capita disposable income/income. Per capita net income of rural income distribution should not only consider between urban income and rural income distribution, industrial management should also be considered. 4. 2. 3. Delete distinctive item (1) Hausman test Remove the income distribution (IG) from the original data. Using Stata11. 0 software to do Hausman test , the results are shown in Table 7. Table 7. The Hausman test results were deleted without significant items Hausman test Chi(0)=0. 00 Prob>chi2=0. 00 (2) Fixed effect model regression results According to 2013-2015, 50 state-level poverty-stricken counties in Guizhou raw data, to delete the income distribution, parameter estimation using Stata11. 0 software, as shown in Table 8 results are obtained. Table 8. The regression results of the fixed effect model after the total linear variables were deleted POV estimate of parameter standard error P>t FS -0. 05769 0. 011084 0 FE -0. 01582 0. 007365 0. 03 RGP -0. 12475 0. 004328 0 C 0. 60159 0. 025240 0 From Table 8, the regression results are significant, and the scale of financial development most important state-level poverty-stricken counties in guizhou's slow, and the regression equation is obtained: 4.3. The empirical result analysis From the point of empirical results, state-level poverty-stricken counties in Guizhou about financial scale, financial efficiency and economic growth slow positively to poverty. Expand the scale of financial and improve financial efficiency, promote economic growth will reduce the incidence of poverty, and it also can alleviate poverty effectively. ititititit RGPFEFSPOV  12475.001582.005769.060159.0 59 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60 From the regression equation of financial scale and financial efficiency of regression coefficient : financial scale relative to financial efficiency, it plays a more important role in slowing poverty. Improving Financial efficiency can give residents more store credit support services. And the expansion of financial development means that the expanding of the total number of financial institutions and financial, thus it is effective to relief poverty. State-level poverty-stricken counties in Guizhou financial development on poverty relief play a positive and effective role. Financial development in addition to the direct action slow poverty, And also alleviate poverty through economic growth. 5. Conclusions Although many literature research in the current financial relationship with poverty alleviation, but study of Guizhou county to originality in this paper. In this paper, the results show that 50 state-level poverty- stricken counties in Guizhou about financial scale, financial efficiency and economic growth slow positively to poverty. Financial scale relative to financial efficiency, it plays a more important role in slowing poverty. . In addition, economic development is an important role in poverty reduction. The local government should develop the local economy, and poverty reduction effect will increase greatly. In this paper, the original selection of the income distribution is urban per capita disposable income/income and rural per capita net income. Eventually the result was not significant. If this paper can use Gini coefficient to show income distribution index of each county may be better. But with data is not available, so this paper cannot calculate the Gini coefficient. It is a big regret. Hope that in later research we can continue to improve and perfect the defects. References Tan L L. Rural Financial And Poor. Jilin university, 2011. Cui Y J, Sun G. The reason why financial development is slow poverty?- evidence from China. Journal of financial research, 2012, (11) : 116-127. Wu Y. Poverty reduction effect of China's rural financial development research - based on the analysis of national and subregional. Journal of southwest university for nationalities (humanities and social science edition), 2012 (7) : 109-113. Guo W. The experience of the rural financial poverty relief, the difficulties and countermeasures, in fuchuan county as an example of guangxi. Journal of theoretical exploration, 2013, (5) : 98-102. Song X, Li L, Xiao L. Review of research on credit risk management for rural credit cooperatives. Journal of Risk Analysis & Crisis Response, 2017, 7(1):21. Deng K. Financial poverty alleviation and efficiency evaluation - in qinba mountains bazhong, for example. Journal of rural economy, 2015, (5) : 86- 91. Cui Y J. China's Financial Development Impact on Poverty Slow: Theory and Empirical. The Northeast University of Finance and Economics, 2012. 60 ___________________________________________________________________________________________________________ Journal of Risk Analysis and Crisis Response, Vol. 8, No. 1 (March 2018) 52-60