مقالات پژوهشی
Ali Asghar banouei; parisa mohajeri; fatemeh kalhori; zahra abdolmohammadi; zahra zabihi; sahar mohammad karimi; maryam parsa
Abstract
Since the 1950s, many types of non-survey methods have been introduced by regional input-output economists for the estimation of Regional Input-Output Coefficients (RIOCs) and Regional Input-Output Tables (RIOTs). On the one hand, there are different kinds of location quotient methods (for example, , ...
Read More
Since the 1950s, many types of non-survey methods have been introduced by regional input-output economists for the estimation of Regional Input-Output Coefficients (RIOCs) and Regional Input-Output Tables (RIOTs). On the one hand, there are different kinds of location quotient methods (for example, , , , , , , and ) which focus on estimating RIOCs and balancing RIOTs and which require acceptance of 2 types of residuals: 1)exports of a region to other regions and the rest of the world, and 2)regional sectoral value added. On the other hand, there are Commodity Balance (CB) and Cross-Hauling Adjusted Regionalization Method (CHARM) which concentrate on estimating RIOTs and the regional sectoral value added which play a key role in balancing RIOTs. The application of the CHARM method is not appropriate for estimation of RIOTS in Iran. The main reason is that the use of CHARM method needs two residuals for balancing the RIOT. One of these residuals is the trade balance (exports- imports) and the second one is regional sectoral value added and the regional GDP. Using the former as a residual seems to be plausible for row balancing of RIOTS, whereas the latter unnecessarily adjusts the regional sectoral value added as well as regional GDP. This residual is not appropriate for countries like Iran which have regional accounts. Since 2000, the statistical center of Iran (SCI) has been providing regional accounts for 31 provinces. It comprises of 72 regional sectors which are comparable with national account classifications. The regional accounts of SCI provide us with sectoral intermediate inputs, sectoral value added and sectoral output for 72 sectors for these 31 provinces. If we use CHARM method for regionalizing NIOTs, the regional sectoral value added and the regional GDP given by the SCI have to be ignored due to the inevitability of residual regional sectoral value added.
In this paper, we have shown that the sectoral value added in regional accounts of Iran is adjusted involuntarily due to the use of CB and CHARM for estimation of RIOTs. This issue raises an important question: why should theofficial regional sectoral value added be unnecessarily adjusted? To tackle this problem, the new mixed CB-RAS and CHARM-RAS method is proposed. The new proposed mixed method only takes regional trade balance as a residual and uses the official regional sectoral value added of the regional accounts that have been provided by the statistical center of Iran.
For the application of CB, CHARM, CB-RAS and CHARM-RAS methods, the following data were utilized. 1) Based on the survey- based NIOT of 2001, we updated two NIOT for the year 2002. 2) The regional sectoral intermediate inputs, value added and outputs of Gilan provinces were directly taken from the regional accounts of the SCI. 3) The survey- based RIOTS of this province for the year 2002 were obtained. In order to make the applications of the two methods manageable, the data were aggregated into seven sectors: agriculture, mining, agro-based industries, other industries, water, electricity and gas, construction and services. Based on the CHARM and the proposed mixed CHARM-RAS methods, RIOTS for the province of Gilan for 2002 were generated. The results indicate that the official total value added (GDP) of Gilan province in 2002 was 23401590 billion rials, whereas the corresponding estimated figure derived from the CHARM and CB methods was 22847050 billion rials. This suggests that the CHARM and CB methods underestimate the GDP of Gilan province by 2.4 percentage. Moreover, The regional sectoral value added deviations were calculated. The minimum deviation for agro-based industries was -%9 and the maximum deviation was %+54.6 for mining. To fill this gap, we have proposed the new mixed CHARM-RAS and CB-RAS method to generate RIOTs for the this province. Applying these methods no longer requires the unnecessary adjustments , the regional GDP, and the use of official data. Therfore, the new mixed method eliminates the shortcomings of the previous methods. Moreover, five conventional statistical methods, namely MAD, RMSE, TIL, STPE and WAD were utilized for assessing statistical errors between supply multiplier matrixes which are derived from the new mixed CB-RAS and CHARM-RAS method corresponding to the official figures. The results show that the degree of accuracy of the supply multiplier matrix of the mixed CB-RAS and CHARM-RAS method in all of the five statistical methods is much closer to the official figures than the one gained using CB and CHARM methods independently. Hence, the statistical error of CB-RAS and CHARM-RAS method is much less than the statistical error of CB and CHARM method in supply multipliers.
مقالات پژوهشی
vahid shaghaghi shahri
Abstract
Among possible measures taken by some countries to safeguard against the phenomenon of globalization is leaning on regional integrations. In fact concurrent with globalization efforts as well as emerging knowledge-based economy, the idea of regional integration has gained momentum in all continents. ...
Read More
Among possible measures taken by some countries to safeguard against the phenomenon of globalization is leaning on regional integrations. In fact concurrent with globalization efforts as well as emerging knowledge-based economy, the idea of regional integration has gained momentum in all continents. By utilizing their comparative advantages in different areas especially by strengthening institution and structures of knowledge based economy, developing countries can ward off against the ill impacts of the globalization era and use these advantages to attract more foreign direct investments and hence achieve higher rates of economic growth.
Methodology
The literature review shows how the classical Gravity Model brings good results to explain the international economic relations. Since the gravity model has physical roots, the trade & foreign investment flows depends on the market size of the origin and the destination countries as well as the distance between them. Recently by the emerging knowledge-intensive economy, several studies have highlighted that an additional effect which should be taken into consideration is the impact of knowledge & new technologies to persuade for attracting of FDI and to improve regional integration & economic relations between countries. For this phenomenon, it is asserted that if some structural changes happen in one country in a way that they affect capital flows, they will affect FDI flows of the neighbors as well, thus, producing positive effects on the volume of bilateral FDI.
In this article an effort is made to evaluate the impact of regional integration & Knowledge based economy on boosting bilateral FDI among the ECO membership countries. We modeled the country-to-country foreign investment flows over the period from 2005 to 2013 for the Economic Cooperation Organization (ECO) by using the Gravity regression model. The sample was restricted to the 10 ECO countries for n=9 years, resulting in a N= (n*T) = 900 observations. In order to choose between random versus fixed effect specification, an Hausmann test was performed. This test highlights the strong preference for the fixed effect model. Also, we modified Wald test for groupwise heteroskedasticity in the fixed effect regression model and Durbin-Watson test for the Residual autocorrelation detection.
Results and discussion
Results of the Gravity model confirm the assumption of the effectiveness of knowledge economy effects on boosting of bilateral FDI for ECO countries. In other words, there is a positive relationship between the observations of FDI & Knowledge economy for ECO member countries.
The results show that the sign in the basic components are similar to the existing literature. GDP of the origin as well as the destination country is significant, as predicted. Furthermore, the inverse distance has the predicted positive impact on the FDI flows.
Conclusion
The results show a good fit of the gravity model to the related data on the bilateral FDI between ECO countries, hence, confirming the importance of the structural variables of the theoretical model and the independent variable. This results is in the confirmation of the theory relating to spillovers and locational factors. The results represent the importance of the GDP origin & destination countries, and the distance between the two countries and adjacency on FDI flow between ECO member countries. Also the results indicate that improving the KEI in host country would increase the bilateral foreign investment between the selected Islamic countries by 41 million dollar.
This fact calls for further cooperation among the ECO member countries and Institutional reforms & infrastructural development for improving knowledge-based economy in attracting of bilateral FDI.
مقالات پژوهشی
Fatemeh Mohammadi Aydoghmish; Mojtaba Rafieian
Abstract
Nowadays, along with globalization which has resulted in the free movement of capital and the releasing of human forces from locative constraints, territories (countries, regions and cities) play a more vital role in attracting capital and mobile human forces, as well as increasing competitiveness. As ...
Read More
Nowadays, along with globalization which has resulted in the free movement of capital and the releasing of human forces from locative constraints, territories (countries, regions and cities) play a more vital role in attracting capital and mobile human forces, as well as increasing competitiveness. As a concept which was until recently solely used for business corporations, competitiveness has entered the literature on places due to the heightened role of territories in attracting capital and mobile human forces. Therefore, the policy approaches of countries and regions have changed from relative advantages to competitive advantages.
When discussed in the context of territories and locations, competitiveness became more complex and its conceptualization and application were the focus of many pundits. Territorial competitiveness is a multi-dimensional concept which is generally defined as efforts to enhance location advantages through enhancing the value of the location for different activities. This concept has been approached from two different viewpoints. The first approach considers the internal factors of business corporations (relative advantage element), while the second approach captures soft dimensions such as human, social, knowledge and environmental resources in addition to economic and efficiency dimensions. The main techniques for evaluating territorial competitiveness are benchmarking and use of spatial instruments.
The current study has adopted the second approach and the benchmarking technique. However, negligence of the non-economic dimensions of territorial competitiveness and the identical application of the factors affecting this type of competitiveness does not seem to suit the Iranian context. Accordingly, the aim of this study is to identify and evaluate the components of territorial competitiveness in the provinces of Iran. Hence, the study consists of two major steps. In the first step, the components of territorial competitiveness in Iran are extracted and identified. As for the next step, the competitive rank of each province is determined. Therefore, the study is driven by the following research questions:
What are the main factors affecting territorial competitiveness in Iran?
What is the rank of each province in Iran in terms of territorial competitiveness?
In order to address the research questions, a number of indices were investigated to operationalize the analysis of territorial competitiveness. This was done by reviewing the dominant viewpoints and considering the major indices in theoretical and experimental research studies through the documentary method and reliance upon secondary data. In order to collect the necessary information about the investigated indices, except for some indices which had to be extracted from reports and maps, the researchers used statistical yearbooks and selected indices of each province which were available on the website of the Statistical Centre of Iran. Then, the data were entered into SPSS Software in order for factor analysis to be run. By using factor analysis, some of the major shortcomings of evaluating competitiveness through the benchmarking technique (e.g. overlap and possible correlation among the measures, identical effect coefficients for each of the factors) could be overcome. Since the benchmarking of regions and cities in terms of their competitive status may guide policy-making and decision-making procedures, the current study aimed to determine the competitive rank of each province after identifying the factors affecting territorial competitiveness. To this end, the results of factor analysis were used to design a network of goals (territorial competitiveness), criteria (factors extracted through factor analysis) and sub-criteria (indices related to each factor) which were analyzed using Super Decisions Software. Simultaneous use of factor analysis and analytic network process (ANP) makes it possible to overcome the subjectivity of expert judgments in analytic network process and diminish the possibility for any incompatibility in the judgments.
As a result of this procedure, 7 factors were extracted from factor analysis with the cumulative variance of 79.9%. The extracted factors included (1) knowledge, technology, cultural and human resources; (2) national and international relations; (3) economic performance; (4) potential for economic partnership; (5) physical and infrastructural resources; (6) per capita of higher education and healthcare; (7) environmental conditions. The results of ranking the provinces using ANP indicated that Tehran is the most competitive province of Iran, followed by Yazd and Bushehr provinces at the 2nd and 3rd ranks. The significant difference between the competitive score of Tehran and the subsequent province (Yazd) is a clear sign of the accumulation and concentration of resources and competitive grants as well as the highlighted national and international role of this province. On the other hand, Lorestan, North Khorasan and Sistan and Baluchestan Provinces stood at the lowest ranks (29th, 30th, and 31st) among the provinces of Iran in terms of competitiveness levels. The undesirable conditions in these provinces may have been caused by the lack of sufficient infrastructures in addition to economic, social and knowledge deprivation. Also, investigation of the radar charts indicated that although Tehran is the most competitive province in Iran, its environmental conditions are below the national average. On the contrary, Sistan and Baluchestan enjoys a higher-than-average status in terms of environmental conditions. Therefore, provinces have to be considered in terms of each of the factors, since improvement in some dimensions must not lead to negligence of the others. By adopting suitable strategies, lower-score dimensions have to be strengthened in order to reach a higher level of competitiveness.
Even though the spatial distribution of competitiveness in Iran does not follow a specific pattern, it is clear that the obvious difference between Tehran and other provinces has placed it at a unique level (completely competitive). At the same time, the competitiveness levels of many provinces in Iran are lower than the national average. These provinces must be prioritized in any future planning, since the continuation of such flow may significantly reduce competitive values.
مقالات پژوهشی
Habib Ansari Samani; Razieh Davoodi
Abstract
Value Added Taxes, as a suitable substitute for all types of sales taxes, despite the benefits, lead to disadvantages such as the pressure on the general level of prices (Ikpe & Nteegah, 2013). As inflation has always been one of the most important problems in the Iranian economy, it was a concern that ...
Read More
Value Added Taxes, as a suitable substitute for all types of sales taxes, despite the benefits, lead to disadvantages such as the pressure on the general level of prices (Ikpe & Nteegah, 2013). As inflation has always been one of the most important problems in the Iranian economy, it was a concern that in a situation where the country faces a severe structural inflation and inflationary sanctions, will the implementation of the value added tax program increase inflation?
Taking into account the predictions made by the researchers and the executives of the project, we aim to study the issue. 7 years have passed since the VAT was implemented, therefore, it will be necessary to survey the effects of this tax. This paper will survey the inflationary effects of VAT in Iran provinces.
Theoretical Framework
Implementation of value added tax (VAT) from two direct and indirect channels affects the general level of prices. Since the value added tax is an indirect tax based on consumption, a portion of this tax burden is imposed on the final consumer. As a result, implementation of VAT, at least in the early stages, directly leads to a leap in the price of goods and services that are taxed.
Also, the implementation of the value added tax affects the general level of prices indirectly by affecting determining factors of inflation, such as liquidity, expectations and production costs. Experiences of different countries in controlling inflationary effects of VAT show that there are two issues in this regard. The first is which tax does VAT replace? (Which taxes have been eliminated) and the second is the country’s monetary situation (Tait, 1988). If the value added tax replaces one or more direct and indirect taxes, it does not impose a severe pressure on prices because the amount of tax received from the goods does not change much, and only the types of taxes and taxation methods change.
The study of other countries’ experiences shows that if inflation control is one of the main goals of monetary and credit policies and that the government in monetary and financial policies takes into account the issue of inflation, then the implementation of the VAT system along with the inflation controlling short-run policies can be made without a sharp increase in prices (Naderan & Ranjabarkey, 2008).
Methodology
In order to test the hypothesis that VAT has a positive and significant effect on inflation in Iran's provinces, we use provinces variables such as inflation, effective rate of VAT, unemployment rate, current and construction, local government size, oil and without oil production per capita, and the ratio of deposits to production from 2008 to 2013. The Econometric hypothesis testing method is a multivariate regression analysis with panel data.
Findings
The findings show that there is a positive and significant relationship between unemployment rate and provinces’ inflation, which shows the stagflation in Iran's economy Moreover, the current size of the government has a positive effect on inflation, since an increase in government spending will lead to an increase in the budget deficit, public sector debt will increase, and will lead to an increase in the supply of money. Given the positive relationship between the general level of prices and liquidity, raising the money supply will lead to an increase in the general level of prices. Whatever the state budget allocates to development expenditures and government development expenditures by improving production and strengthening the supply side will cause macroeconomic surpluses and will reduce inflation. Since VAT affects the sale price of final goods, it is likely that sellers will conceive of increasing production costs and transfer tax burden to consumers. On the other hand, some manufacturers of goods that are not subject to VAT imagine that the tax on their goods and services will also apply and increase the price of their goods. Therefore, the findings show that there is a positive and significant relationship between the value added tax and inflation in the provinces of the country. The high ratio of deposits to production in the in the ceteris paribus situation indicates that people prefer to hold liquidity in banks. This liquidity will be driven by investment in the supply of various sectors of the economy and will reduce inflation. An increase in domestic production of oil will lead to an increase in the supply of foreign currency and, subsequently, a decline in the exchange rate. Due to the negative relationship between the exchange rate and the rate of inflation, the exchange rate cuts have led to an increase in inflation, and the negative relationship between oil production and inflation is justified. Also, the findings show that there is a positive and significant relationship between the oil-free domestic production and inflation in different provinces.
Conclusions and Suggestions
In this study, the inflationary effects of VAT were investigated using multivariate regression in panel data. For this purpose, after collecting provincial data, the inflation effect of VAT in a model with dependent variable of price level growth rate along with the major variables affecting inflation was measured. Given the positive impact of VAT on inflation, it is recommended that the government initially takes caution in increasing the VAT rate, and secondly, pay closer attention to direct taxes and alternative VAT taxes that have high incomes and low inflationary effect.
مقالات پژوهشی
zahra nasrollahi; Mehran Zarei
Abstract
Introduction
Today, the importance of economic planning and predictions in the development of different countries, especially developing countries, is obvious to anyone. Due to the application of input-output tables in cases such as identification and forecasting the needs of the labor and energy market, ...
Read More
Introduction
Today, the importance of economic planning and predictions in the development of different countries, especially developing countries, is obvious to anyone. Due to the application of input-output tables in cases such as identification and forecasting the needs of the labor and energy market, analysis of economic structure and its changes, identifying the leading (key) sectors and analyzing intersectoral dependencies, the role of these models in development planning has been highlighted.
In the past decades, introducing the concept of regional planning and applications of the input-output model in this context have led researchers and policy-makers to consider regional input-output tables more seriously. Due to the fact that survey-based method for providing input-output tables is expensive and time-consuming and with regards to the weakness of regional data in Iran, non-survey based methods, especially location quotient (LQ), are the most common methods for providing regional input-output tables in Iran. In recent years, LQ method and its various functions has significantly evolved and SFLQ function has recently been introduced. The aim of this study is introducing the most popular LQ functions (with emphasis on SFLQ) and providing a method for evaluation of their performance in regionalization of national input-output tables (the case study of Yazd province). For this purpose, the national input-output table of parliament research center in 2011 was utilized. Based on empirical studies in this field, flegg location quotient functions, which include FLQ, AFLQ and SFLQ have a significant advantage when estimating regional coefficients compared to other LQ methods. The current study evaluates flegg location quotient functions.
Methodology
Compiling regional input-output tables with the use of LQ methods is done based on two basic principles. Firstly, it is generally true that the smaller the economic area, the more dependent that area’s economy on trade with “outside” areas – transactions that cross the region’s borders – both for sales of regional outputs and purchases of inputs needed for production (Miller & Blair, 2009: 69). Secondly, all LQ methods are based on this assumption that regional and national technologies are identical and that regional trade coefficients differ from national input coefficients to the extent to which goods and services are imported from other regions (Kowalewski, 2015: 242). Given these two points, the role of LQ adjustment coefficient is to take account of a region’s purchases of inputs from other regions in the nation.
An important issue that is challenging to the use of flegg LQ, is the selection of optimal values for the unknown parameter δ in this formulas. In the absence of regional survey-based table, choosing this value is very challenging. In this study, to determine the optimal values of δ and δj in flegg LQ methods and evaluating their performance, the proposed method by Bazzazan et al (2007) has been modified. In this method for each δ value, one regional input-output table will be calculated. Then, the supply-driven Ghosh coefficients will be calculated and using the value added data of the province economic sectors, which is available in the regional accounts of Statistical Center of Iran, the output of each sector will be estimated using the related formulas. In the next step, a value of δ, which minimizes the statistical errors caused by the difference between real and estimated outputs, will be selected as the optimal value.
Results and Discussion
The results of the current study show that based on different statistics, FLQ and AFLQ Methods used to estimate the Yazd province table have similar performance. However, the SFLQ partially improved the performance. Moreover, SFLQ has a different performance when used to calculate forward and backward linkages compared to other methods. In FLQ and AFLQ, the only sector with forward and backward linkages larger than one is "manufacture of chemicals, coke and refined petroleum products". Nevertheless, this sector with an SLQ of .11, is one of the smallest non- specialized sectors in Yazd province. The table derived from the SFLQ method, demonstrates that sectors of "manufacture of rubber and plastics products" and "manufacture of electrical and office machinery", which are specialized sectors in Yazd province (SLQ> 1), have forward and backward linkages larger than one.
The optimal values of δ and δj as expected are very high. Therefore, the optimal value of δ based on all statistics is .999, the maximum value possible. The optimal value of δj is very high for most sectors as well.
Conclusion
It is expensive and time-consuming to produce a table based on a survey of establishments in a particular economy, both at national and regional level. This problem, in addition to the insufficiency of data in Iran has caused the use of location quotient method as a non-survey based method in Iranian Studies. However, few studies have evaluated the performance of these method formulas. The aim of this study was introducing the most popular LQ functions (with emphasis on SFLQ) and evaluation of their performance in regionalization of national input-output tables (the case study of Yazd province). For evaluation of these functions performance, a method based on modification of Bazzazan et al’s (2007) method was presented. The results show that FLQ and AFLQ Methods have a similar performance both in term of estimation of output and calculation of forward and backward linkages, when preparing the input-output table for Yazd province. However, the SFLQ has partially improved their performance.
مقالات پژوهشی
Samad Hekmati Farid; fahmideh fattahi
Abstract
The aim of this study is estimating the effect of technology indices on poverty probability and poverty intensity by using rural and urban Household data from Statistics Center of Iran in 2014. For this purpose, first, we estimated the relative poverty line then the impact of ICT indices and household ...
Read More
The aim of this study is estimating the effect of technology indices on poverty probability and poverty intensity by using rural and urban Household data from Statistics Center of Iran in 2014. For this purpose, first, we estimated the relative poverty line then the impact of ICT indices and household characteristics on poverty probability and poverty intensity were analyzed by two-step Heckman Method. In this method, the first step estimates the probit model to investigate the effect of variables on poverty. Moreover, in this step we calculate poverty probability. In the second step, we regress explanatory variables on poverty intensity variable after calculating poverty intensity.
The results indicate that an increase in ICT indices has a negative and significant impact on the probability of being poor and poverty intensity. Therefore, if the percentage of urban households with mobile phones, computers and internet, increases, the probability of being poor decreases respectively to 21.42, 11.92, 10.8 percent for urban households and 22.73, 15.44 and 10.7 percent for rural households. Furthermore, an increase in the percentage of urban households with mobile phones, computers and internet, decreases poverty intensity respectively 35.84, 31.98, 31.87 percent for urban households and 26.17, 34.56, 17.03 percent for rural households.
Moreover, the results show that an increase in the education level and marriage of heads of households have a positive and significant impact on poverty reduction probability and poverty intensity. In addition, the results indicate that the relationship between age and poverty may not be linear, Therefore, Increasing the age before the aging stage reduces poverty probability but after this stage, Increasing the age lead to increasing poverty probability. In addition, increasing household size and employment of the head of household have a negative and significant impact on the probability of being poor and poverty intensity.
Extended Abstract
Poverty is wide-spread and it is a global phenomenon that cuts across all countries of the world. No nation, not even the most technically and economically advanced economy, could boastfully assert the absence of at least a single dimension of poverty within her economy. However, poverty seems to be predominantly a fundamental trait among developing and the less developed countries alike. (Anigbogu, Onwuteaka, Anyanwu, & Okoli, 2014). There are many factors that can affect poverty alleviation. Information and communication technology (ICT) is one of the most important factors that affects poverty. Throughout history technology has been a powerful instrument for economic and social development. Technology played a critical role in reducing poverty in vast areas of the world in the past and can play a crucial role in the battle against poverty today. It can be employed in a variety of fields, from increasing agricultural productivity to the generation of cheap energy, from providing clean water to improving health. In particular, information and communication technology (ICT) can address the problem of poverty by increasing people’s access to education, health, and financial services. Strikingly, even simple technologies might make a difference in poverty reduction. The case of cellular phones in Africa is a well-known example. Likewise, small businesses and social enterprises creating access to primary goods are greatly helped by new technologies (Popoli, 2015). Given the importance of this issue the aim of this study is surveying the effect of ICT indices on poverty in Iran.
Methodology
In this paper, we have used rural and urban household data from Statistics Center of Iran in 2014 that contain information about 18885 urban households and 19390 rural households. For surveying the effect of ICT indices on poverty, we investigate the impact of information and communication technology indices on poverty probability and poverty intensity. For this purpose, first, we estimate relative poverty line as half of the average household expenditure in Iran. It is an indicator of inequality at the bottom of the income distribution, which acts as a cause of social exclusion and undermines equality of opportunity.
After determining relative poverty line we can identify poor households and then the impact of ICT indices and household characteristics on poverty probability and poverty intensity by two-step Heckman Method. In this Method, the first step is estimates probit model to investigate the impact of variables on poverty. Also in this step we calculate poverty probability. In second step, calculating poverty intensity, we regress explanatory variables on poverty intensity variable. In the first step the estimated coefficients are used to estimate the inverse Mills ratio for each individual. In Step 2, the estimated Mills ratio is used as an instrument or regressor in the logit model.
Results and Discussion
The results show that relative poverty line for urban and rural households in 2014 was 94023127 and 104338980 Rials respectively which implies that households by lower expenditure than this line are poor households. The estimation of the first step of Heckman model showed that all of the coefficients are significant. Marginal effects after probit coefficients indicate that an increase in ICT indices have a negative and significant impact on the probability of being poor and poverty intensity. Therefore, an increase in the percentage of urban households with mobile phone, computer and internet, decreases the probability of being poor 21.42, 11.92, and 10.8 percent for urban households respectively and 22.73, 15.44 and 10.7 percent for rural households. It should be noted that lack of infrastructure, especially in developing countries, results in the poor being deprived of the information and knowledge that would help them to live healthier lives, improve their educational standards, and gain employment and business opportunities. ICTs have the potential to process and disseminate vast amounts of information and can therefore have a far greater impact on the lives of the poor than informal information networks. ICTs enable households, that works in firms, to improve productivity and income generation by allowing them to access the market information faster and cheaper. This may strengthen forward linkages to the market (Mbuyisa & Leonard, 2015)
In addition, the results indicate that the relationship between age and poverty may not be linear, So that Increasing the age before the aging stage reduces poverty probability but after this stage, Increasing the age lead to increasing poverty probability. Furthermore, an increase in human capital index (education level) and marriage of heads of households has a positive and significant impact on poverty reduction probability. In addition, employment of the head of the household has a significant and negative impact on the probability of being poor.
To estimate the second step of the Heckman method first we calculated poverty intensity. Therefore, purpose, we should square the poverty gap, which measures the extent to which households fall below the poverty line, for each household. This index puts more emphasis on observations that fall far short of the poverty line rather than those that are closer.
Poverty intensity will be a dependent variable in the second step of Heckman method and its value for non-poor households will be zero. Estimating poverty intensity equation in the second step of Heckman show that an increase in the percentage of urban households with mobile phone, computer and internet, decreases poverty intensity for 35.84, 31.98, and 31.87 percent for urban households respectively and 26.17, 34.56, and 17.03 percent for rural households that indicate that ICT has an important role in decreasing poverty intensity.
Moreover, the results show that an increase in education level and household size, together with being married and of the employment of the head of household have a negative and significant impact on poverty intensity.
Conclusion
The results of this paper show that improvement in ICT indices can reduce probability of being poor and poverty intensity and provides evidence for benefits of ICT and the role that it plays in poverty alleviation. As ICT contributes to poverty reduction, there are good reasons for governments to promote the use of ICTs in the business sector and households for poverty alleviation.
Furthermore, results indicate that household characteristics such as education level, household size, being married, age and employment of the head of the household have a significant impact on poverty. Hence, governments should pay attention to this characteristic in poverty alleviation programs.