Trust and socio-economic indicators: model including religion factor

The article is concerned with the following issues: definitions, indicators of trust were reviewed; the working hypotheses of the research were formed; the choice of factors related to the trust indices was made; cluster analysis of the relationship between individual trust indices and economic indicators was carried out; a correlation analysis of the relationship between individual trust indices and socio-cultural indicators was conducted; a neural network for modeling the general index of trust based on a well-founded set of economic and socio-cultural indicators was developed. The hypothesis about the influence of socio-cultural factors on trust and out of which there was distinguished a relation to a specific religion. By means of correlation analysis and neural networks, it was shown that Protestantism and Catholicism are the most significant religions that affect the general index of interpersonal trust. However, atheism has a more significant impact. Following 198 observations, each of which represented the country for a given year in the period from 1995 to 2014, the neural network produced satisfactory results in forecasting the total trust index on the basis of the following factors: GDP per capita, GINI coefficient, atheism (percentage of population, support such an attitude to religion). The neural network recognized 89.9% of the data and 90% of the test set indicating that the network got adjusted and could be used for modeling. The scatter diagram for a 5% error indicates that most of the data is within the required value. But it should be noted, that the model overestimates trust in Ukraine at the end of the analyzed period. This gives grounds for the assumption that in Ukraine there are additional factors that negatively affect interpersonal trust.


Relevance of the purpose and hypothesis of the study
The subject of modern studies are various aspects of the socio-economic nature of trust as a mechanism of interpersonal, intergroup, interinstitutional relations. Despite the substantial number of foreign and domestic publications that deal with the problems of trust, the theoretical understanding of this complex and multi-faceted phenomenon only begins , (Merkulova, 2014), , . Measurement of personal and institutional trust is the subject of discussions, and empirical studies on this topic are limited.
Two conceptual approaches can be distinguished in the theoretical analysis of trust: individualistic approach and economic and sociological approach .
The first assumes that trust is the product of individual (psychological, physical, religious, etc.) personality characteristics. The second approach implies that trust is formed under the influence of economic and social factors, including income, education, etc. The working hypothesis of our study is the assumption of a mixed set of factors that influence the formation of trust.
This set is the goal of the studyidentifying the links between social and economic indicators and trust indicators. The following questions were considered in the work: review of definitions, indicators of trust; the working hypotheses of the research are formulated; The choice of factors related to the trust indexes was made; Cluster analysis of the relationship between individual trust indices and economic indicators was implemented; a correlation analysis of the relationship between individual trust indices and socio-cultural indicators was conducted; A neural network was developed for modeling the general index of trust based on a well-founded set of economic and socio-cultural indicators.
Definitive aspect. In modern socio-economic studies, the division of trust in the interpersonal, which, in turn, is subdivided into trust to strangers and acquaintances (relatives, friends etc.) and institutional, which is also analyzed in positionstrust in various state and public institutions, etc. These types of trust have their own distinctions that determine the measurement capabilities (Merkulova, 2014).
In general, trust is a measure of confidence that the subject will behave as expected, given the lack of control over the environment in which it operates. "Trust is the expectation of behavior of other individuals that influence the decision of a person in a situation when he must begin to act without knowing if these acts are performed" (quoted in "This trust belongs to interpersonal trust" (Institutional economics, 2005)).
There are separate indices for estimating distinct types of trust, which are calculated based on the respondents' survey results according to a certain scheme with 4 possible answers. The most well-known sociological surveys of the level of trust are General Social Survey (GSS), World Values Survey (WVS), European Social Survey (ESS), Latinobarómetro, and others (Merkulova, 2014). In our study, we used mainly the WVS database   A comparative analysis showed that Ukraine relates to the countries with strong family and foreign credibility, but very weak in government and police trust.

Choice of factors that affect trust
Economic factors. High confidence in the country is strongly associated with high household incomes. A higher level of trust is closely linked to a lower income inequality. The explanatory reasons remain unclear, but there are several theories. Inequality of income can make it more difficult for people in distinct levels to have a common sense of purpose and trust each other. Another reason is that low level of trust can hinder the development of positive social relationships, which, in turn, contributes to a high inequality of income (OECD, 2011).
In this paper, instead of household income, GDP per capita was selected to analyze how this macroeconomic indicator relates to the level of trust. The GINI index was also taken as an indicator of income inequality.
Socio-cultural factors. Trust in others may be caused by moral or cultural reasons. In addition, religious beliefs may be important since different religions have disparate attitudes to social interactions and state structures.
In , the influence of religion on trust was insignificant, but this work was conducted only for the United States. Other researchers working on interstate examples have found that religious affiliation sometimes affects the quality of government. One of the explanations for this conclusion is based on the attitude of different religions in relation to the common good and social interaction with others, especially to another religion . In this paper, an analysis of the impact of major world religions and the availability of higher education on the level of trust in the country was implemented.

Neural network for the general index of trust
In order to construct the neural network model, the following steps were undertaken: For the selection of the most noteworthy features, a correlation analysis was carried out ( Figure  1). For analysis, Deductor software was used.  (Table 1). With the growth of GDP per capita, the percentage of people who have received university education and the reduction of inequality (GINI), the trust index is increasing.

Source: author's development
The influence of religions on the trust index is ambiguous. It is important to point out that only Buddhism, Hinduism and Atheism, have a positive effect on trust.
Indicators that have the highest significance on the trust index: GDP per capita, GINI, religion: atheism -the percentage of people who support this attitude towards religion were selected. The starting point is the general trust index. A neural network was built on a maximum sample of available information. 198 observations were used, where each observation is a country for a specific year in the period 1995-2014. Data from Ukraine were excluded from the sample for a test case for recognition.
Neural network has the following parameters: the number of layers -1; number of neurons in the layer -7; activation function -sigmoid; steepness activation function -1; algorithm -online training; learning speed 0.1; moment -0.9; Error -0.05.
The neural network recognized 89.89% of the data and 90% of the test set, indicating that the network is adjusted and can be used for simulation. The maximum error for the data is 0.189, and the average 0.0193, for the test set the maximum error is 0.0538 and the average is 0.014. The scatter plot, where the red lines show a 5% error, indicates that most of the data is in an acceptable region (Fig. 2).
The most accurate estimates of trust model are provided for the countries represented in Table. 2 To assess the quality of the model, it was tested on the example of Ukrainea country that was not part of the working set in the development of the neutron network ( Table 3). The deviation of the predicted values (Tdata) from the model (Testimated) shows the error .

Source: author's development
The result shows that the error does not exceed 20%; it should also be noted that the model gives an overestimate of confidence at the end of the analyzed period. This gives the ground for the assumption that in Ukraine there are additional factors that negatively affect interpersonal trust.

Conclusions
The hypothesis about the influence on the trust of socio-cultural factors, among which the relation to a specific religion was tested. Using correlation analysis and neural network simulations, it has been shown that Protestantism and Catholicism are the most significant religions that affect the general index of interpersonal trust. However, atheism has higher significance level. The constructed neural network model provides satisfactory results in forecasting the general index of trust based on factors: GDP per capita, GINI coefficient, atheism (percentage of population supporting such attitude to religion). The neural network recognized 89.89% of the data and 90% of the test set, indicating that the network is adjusted and can be used for simulation. The scatter plot for a 5% error indicates that most of the data is within an acceptable region.
But it should be noted that the model gives an overestimated assessment of confidence in Ukraine at the end of the analyzed period. This gives the ground for the assumption that in Ukraine there are additional factors that negatively affect interpersonal trust.