Influence of Capital, Labor and Wage Levels on Production of Super Tofu Padang

The purpose of this study was to determine the effect of capital, labor, and wage levels on company production. The research method uses quantitative descriptive with 36 samples from January 2016 until December 2018. Based on the partial t test it can be concluded that capital has a negative and significant effect on production. Based on the partial t test it can be concluded that the variable of labor does not affect production. Based on the partial t test it can be concluded that the wage level variable has positive and significant effect on production. Based on the simultaneous F test it can be said that capital, labor, and wage levels together affect production. The ability of the independent variable in explaining the dependent variable was 77.40%, while the remaining 22.60% was influenced by other variables apart from the variables studied.


Background
Indonesia as a developing country is expected to be a developed country. Efforts made to achieve these goals are with national economic development, national economic development has three core objectives including increasing the availability and expansion of distribution as a living necessity and expansion and economic and social choices (Indrajaya 2019). The way to realize success can be done by integrating information systems, increasing the efficiency of information systems to produce more efficient management in business processes. The information system is not only limited to accounting information systems, but has expanded to information systems to manage company resources. Resources consist of human resources (employees), natural resources (material), facility and energy machinery (technology), capital (money), and information (financial data) (Ananda, Putra, and Hedrastyo 2017) Based on economic theory, companies have a goal of achieving the greatest economic profit possible, by definition they are trying to make the biggest difference possible between total demand and total economic costs. In conducting the production process, companies need various input factors which will produce output. The production process can only be carried out if the factors of production are available (Andriani 2018). One of the companies that conducts production is in the Super Indra Tofu industry which was founded in 2002. The Tofu Industry is a medium-scale industry when viewed from the number of workers. The founding of the Super Indra Tofu industry originated from the profession of the business owner as one of the soybean sellers and suppliers for the Tofu industry in Padang City. However, after a few years the customer switched to supplying soybeans to the distributor. Realizing that soybean business is not profitable for entrepreneurs to take the initiative to establish their own tofu business.
The problem that occurs in the Super Indra Tofu industry that is located in the workforce. Manpower is one of the problems faced by Tahu Super Indra. Workers who are not careful during the production process become a source of risk that exists in Tahu Super Indra. The production process is not careful in controlling the time when soaking soybean seeds and also when cooking coconut milk that causes tofu is not ripe. So that the tofu is runny and difficult to be printed and results in the production of Tofu failed and has an impact on the lack of production of Tofu Super Indra. Based on the financial statement data, a very significant increase in production occurred in 2016 in November, amounting to 592, and experienced a very drastic decrease in January 2018 amounting to 313. This was due to an increase in raw material prices. An increase in raw materials forces the business owner to cut the amount of production because the demand and selling price of goods cannot be increased.
Based on previous research a reference from Arsha and Natha (2013) about the Effect of Wage Level, Labor and Working Capital in the study of differences lies in the object. The object of previous research is the textile industry, while in my research that is in the Super Tofu Industry in terms of years in previous studies doing research with annual financial reports, while in my research that is in the monthly financial statements with a sample size of 36 months or 3 years with period January 2016 -December 2018 According to Arsha and Natha (2013) working capital is basically the amount that always exists in sustaining the business that connects the time spent to get materials, tools and services to be used during the production process so as to get sales revenue. Research conducted by Indrajaya (2019), Perdana and Jember (2017), Fachrizal (2016) and Ida Bagus Adi Mahayasa (2013) stated that capital has a positive and significant effect on production. This means that if capital increases, production will increase and vice versa.
According to Amelia (2017) Workers are every person who is able to do work to produce goods or services both to meet their own needs and for the community. Research conducted by Indrajaya (2019), Cahyadinata and Darsana (2018) and Ida Bagus Adi Mahayasa (2013) stated that labor has a positive and significant effect on production. The amount of labor available will affect the amount of production, this research is inversely proportional to the study conducted by Jamli (2012) that labor has a negative effect on production, so the conclusion is labor has a positive effect on production.
According to Cahyadinata and Darsana (2018) wage level is one of the factors that influence the unemployment rate because rising wages will reduce labor demand which will lead to unemployment. Research conducted by Cahyadinata and Darsana (2018), Perdana and Jember (2017), Putri and Kesumajaya (2017) and Arsha and Natha (2013) stated that wage levels had a positive and significant effect on production. Based on the results of research conducted by Cahyadinata and Darsana (2018), Perdana and Jember (2017), Fachrizal (2016) and Arsha and Natha (2013) stated that the level of wages, labor, and working capital had a positive and significant effect on production. An industry needs capital support to be able to increase production so as to impact on increasing business revenue. • H1: capital has a positive effect on production • H2: labor has a positive influence on production • H3: wage level has a positive influence on production • H4: capital, labor and wage levels have a positive effect on production

Research Methods
This type of research in this study is denag quantitative research methods based on the philosophy of positivity, which uses data in the form of numbers and numerics and illustrates the influence of one variable with another variable. The analysis used to detect the causal relationship between two or more variables is BAARJ | Basic and Applied Accounting Research Journal 29 Volume 1 | Number 1 | June | 2021 causal analysis in which there are independent variables namely the variables that influence and the dependent variable that is the variable affected (Sugiyono 2015). The type of data in this study is time series. Time series (time series) is data consisting of one object covering several time periods obtained from financial statements (Sugiyono 2015). The data source used is secondary data. Secondary data is the source of research data obtained through intermediary media or directly in the form of books, records, evidence that has been available or archives, both published and not publicized.
The object is what is to be investigated in research activities (Sugiyono 2015). In this study the object of my research was the Super Sense Tofu Industry which is located at Jl Koto Kampung No.05 Kelurahanaan Gurun Laweh, Nanggalo District, Padang City. The company is engaged in the production of Tofu and Super Tofu marketing.
The data collection technique used in the study is the Observation technique. Observation is a clear and appropriate technique to study actions and behaviors intentionally, systematically about social phenomena and physical symptoms to then be recorded to find and obtain data related to the focus of research (Now 2017).

Table 1. Variable Operational Definition
The method that will be used in this research is descriptive method. Descriptive research is research conducted to determine the value of each variable, whether one or more variables are independent without making connections or comparisons with other variables. Descriptive analysis technique is to look at the highest, lowest, average, and standard deviation or standard deviation obtained by Sujarweni (2016). The descriptive statistical test was carried out with the SPSS 25 software program.
Normality Test according to (Ghazali 2012) aims to test whether in the regression model the confounding variable or residual has a normal distribution. Like the t test and the F test it assumes that the residual value follows the normal distribution. If this assumption is violated then the statistical test becomes invalid for a small sample size. There are two ways to find out whether it is residual. This study uses the Kolmogorov Smirnov normality test.
Furthermore, to test the normality of data for each variable and each sample size using Production is an activity to create goods and services in order to meet human needs. Production does not only include changes in nature or objects, in the economic sense it is said that production is any action that can create or add utility to economic goods. Labor is every person who can or can do work to produce goods or services both to meet individual needs and for the community.
Total manpower (Amelia 2017) 4 The wage rate is one of the factors that influence the unemployment rate because rising wages will reduce labor demand which will lead to unemployment.  (Mitha, 2014). At Shapiro-Wilk, consideration for choosing analysis is the number of cases or the number of observations. If the number of cases is less than 50, then the Shapiro-Wilk analysis is considered more accurate (10). If the number of cases is the same or 50 cases, then the Kolmogorov-Smirnov analysis is considered more precise. If the Sig. (p value)> 0.05 or 0.01, then the data is declared to be normally distributed, if the Sig. (p value) ≤ 0.05 or 0.01, then the data is declared not normally distributed (Lucy 2016).
Multicollinearity Test according to Ghazali, (2012) aims to test whether the regression model found a relationship between independent variables. A good type of regression should not have a correlation between independent variables. If the independent variables are correlated with each other, then these variables are not orthogonal. Orthogonal variables are independent variables equal to zero. To detect the presence or absence of multicollinearity in the regression model. To get the presence or absence of multicollinearity in the regression model can be seen from: a. Nilai Tolerance b. Variance Inflation Factor (VIF) Both of these measurements indicate which of each independent variable is explained by other independent variables. Tolerance measures the variability of selected independent variables that are not explained by other independent variables. So a low tolerance value equals a high VIF value (because VIF = 1 / Tolerance). The cutoff value that is commonly used to indicate the presence of multicollinearity is if the tolerance value ≤ 0.10 or equal to VIF ≥ 10. If the VIF value is proven to be value ≤ 10, there is no colonierity in the regression variable.
Heterokedastisitas test in this study using glacier test. The Glejser Test is performed by regressing the independent variable with its absolute residual value. If the significance value between the independent variables with absolute residuals is more than 0.05 then there is no heterokedasticity problem (Priyatno 2014).
The autocorrelation test aims to test whether in a linear regression model there is a correlation between the fault error in the t period and the error in the t-1 period (before). If an autocorrelation problem occurs, the probability value above the significance is equal to 0.05. One way to detect the presence or absence of autocorrelation is to use the Run Test.
Run Test as part of non-parametric statistics is used to test whether there is a high correlation between residues. if the residuals are not correlated, it is said that the residual is random or random. The Gun Test is used to see whether the residual data is random or not (systematic). (Sujarweni 2016) Decision making in the run test is as follows: a. if the run test results show a significant value smaller than 0.05, it can be concluded that the residual is not random or there is an autocorrelation between the residual values. b. If the run test results show a significant value greater than 0.05, it can be concluded that a random residual or autocorrelation does not occur between residual values. This T test is an analytical technique for comparing one independent variable. T test shows how far the influence of one independent variable individually in explaining the variation of the dependent variable decision making hypothesis test is also partially based on the probability value obtained from the data processing through the SPSS 25 Parametric Statistics program as follows: a. If the significant value is tested t> 0.05 then H0 is rejected b. If significant is tested t <0.05 then H0 is accepted Testing the compatibility hypothesis (goodness of fit) is testing the hypothesis to determine whether an expected set of frequencies equals the frequency obtained from a distribution, such as a binomial, poisson, normal distribution, or from another comparison. So, the goodness of fit test is a test of the suitability or suitability between certain Volume 1 | Number 1 | June | 2021 observations (frequency of observations) and the frequency obtained based on the expected value (theoretical frequency). F test can be done by comparing F arithmetic with this is also marked significance column value (%) will be greater than alpha. The R2 test is the comparison between the variation of Y explained by X1 and X2 together compared to the total variation of Y. If the analysis used is simple regression, then the R Square value is used. However, if the analysis used is multiple regression, then what is used is Adjusted R Square. Goodness of Fit R-Square (R2) is used to measure the contribution of variables X1 and X2 to variations (ups and downs) of Y. R2 is also called the coefficient of determination, and its value ranges from 0 to 1 (0 <R2 <1) the closer to 1, the more precise the regression line is to predict the value of the Y-related variable (Priyatno 2014) This multiple regression analysis is useful to predict the effect of two or more independent variables on one dependent variable or to prove whether there is a functional relationship between two independent variables (X) or more with a dependent variable (Y) (Priyatno 2014). Multiple regression analysis on This research is used to determine the effect of cash turnover, accounts receivable turnover and inventory turnover towards liquidity. The formulation of the multiple regression equation itself according to Priyatno (2014)

Result and Discussion
The results of this study are based on the production of Indutri Tahu Super Indra Padang production data as much as 36 production data of Tofu, then the researchers conducted a test based on the data obtained and got positive results on the descriptive analysis test. Then the researchers conducted further tests using 36 data on the Shapiro Wilk normality test, but researchers got results that were not normally distributed on the Shapiro Wilk normality test, so the researchers omitted 2 data items, namely data no 11 and 12 with the total data being 34 data and get results that are normally distributed so that researchers conduct further studies using 34 data such as the results belows. These results explain that the average value is greater than the standard deviation so that the data can be said to be positive. The average value for the variable labor is 24.69 or 25 people with the highest value of 30 people, the lowest value of 21 people, and the standard deviation of 2.68 or 3 people. These results explain that the average value is greater than the standard deviation so that the data can be said to be positive.
The average value for the variable wage level is Rp. 72,709,305.55 with the highest value of Rp. 89,750,000, the lowest value of Rp. 60,500,000, and a standard deviation of Rp. 8,180,252,54. These results explain that the average value is greater than the standard deviation so that the data can be said to be positive. The average value for the production variable is 431.69 barrels with the highest value of 592 barrels, the lowest value is 313 barrels, and the standard deviation of 86.72 barrels. These results explain that the average value is greater than the standard deviation so that the data can be said to be positive. Based on the table above, it can be explained that the value is Ashy. sig normality test data in this study amounted to 0.200> 0.05. This means that the pre-test requirements for data analysis have been met where the data has been normally distributed.
The results of tests conducted using the Shapiro Wilk test on the amount of research data 34, can be explained that the Sig value of the Shapiro Wilk normality test in this study is 0.227> 0.05 This means that the pre-test requirements for data analysis have been met where the data has been normally distributed. *. This is a lower bound of the true significance. Based on the above table, it is explained that the capital variable has a VIF value of 1.621 <10 with a tolerance value of 0.617> 0.1, so there is no multicollinearity for the capital variable. On the labor variable has a VIF value of 6.699 <10 and tolerance of 0.149> 0.1 then there is no multicollinearity for the labor variable. The wage level variable has a VIF value of 6.001 <10 and tolerance of 0.167> 0.1, so there is no multicollinearity for the wage level variable. This means that this study does not have symptoms of multicollinearity so it meets the criteria for further statistical tests. Based on the table above for the capital variable has a significance value of 0.218> 0.05 then there is no heteroscedasticity in the capital variable data. For the workforce variable has a significance value of 0.930> 0.05 then there is no heteroscedasticity in the workforce variable data. For the wage level variable has a significance value of 0.506> 0.05 then there is no heteroscedasticity in the wage level variable data. This means that in this study has been freed from the symptoms of heteroscedasticity on all data from the variables studied so that it can be used for further statistical testing. Based on the table above, it can be explained that for the variable capital has a calculated value of t 3.491> t table of 2.034 with the support of a significance value of 0.002 <0.05, then the alternative hypothesis can be accepted. This means that the variable capital has a negative and significant effect on production where each increase in capital will increase production.

Tests of Normality
For labor variables having a calculated value of 1.472 <t table of 2.034 with the support of a significance value of 0.151> 0.05, the alternative hypothesis can be rejected. This means that the labor variable does not affect production.
For the wage level variable has a t value of 3.750> t table of 2.034 with a support value of significance of 0.001 <0.05, then the alternative hypothesis can be accepted. This means that the wage level variable has a positive and significant effect on production.   Based on the above table, it can be explained that the constant value in this study is -59.404, the value of the capital variable regression coefficient is 0.000000082, the value of the labor variable regression coefficient is -9.4442, and the wage rate variable regression coefficient is 0.0000074. Then the equation obtained is: Y = -59,404 -0,000000082X1 + 9,442X2 + 0,0000074X3 + e The results of the above equation can be interpreted that a constant value of 59.404 which is negative where it means that if the variable capital, labor, and wage levels are not considered or have a value of 0 then production will decrease by 59.404 barrels. The value of the regression coefficient for the capital variable is 0.000000082 which means that every increase of one rupiah of capital will reduce production by 0.000000082 barrels with the relationship formed in the opposite direction. The regression value for the variable labor is 9,442 which means that every increase in one person of labor will increase production by 9,442 barrels with the relationship being unidirectional. The regression value for the wage level variable is briefly 0.0000074 which means that each increase of one rupiah of the wage level will increase production by 0.0000074 barrels with the directional relationship being formed.
The hypothesis in this study explains that alleged capital has a significant effect. Based on multiple linear regression testing for capital variables can be explained that capital has an opposite direction to production where each increase in capital will decrease production. This result is supported by the partial t test which concluded that the capital variable has a negative and significant effect on production.
The results of this study explain that the alternative hypothesis was accepted. These findings are supported by research by Fachrizal (2016), Indrajaya (2019), Ida Bagus Adi Mahayasa (2013), Perdana and Jember (2017). Research conducted by Fachrizal (2016) states that capital has an influence on production, meaning that if capital increases, production will increase and vice versa. Based on research results by Indrajaya (2019) Capital has a significant effect on production, capital has an important role in the production process, so that the more capital used in the production process, the more productive and vice versa. Ida Bagus Adi Mahayasa's research (2013) states that capital has a significant effect on production, meaning that the more available the capital, the more production will be produced. According to Perdana and Jember (2017) that capital influences production, because capital has an important role in increasing product output.
The hypothesis in this study explains that it is suspected that labor has a significant effect on production. Based on multiple linear regression testing for the variable of labor can be explained that labor has a direct relationship with production where every increase in labor will increase production. This result is not supported by the partial t test which concludes that the variable of labor does not affect production. The results of this study explain that the alternative hypothesis is rejected.
These findings do not support the results of previous studies conducted by Cahyadinata and Darsana (2018), Ida Bagus Adi Mahayasa (2013), and Jamli (2012). According to Cahyadinata and Darsana (2018) that labor has a positive and significant effect on the production of the number of existing workers will affect the amount of production, the more a person is employed, the more the production of goods produced and the production obtained will increase. The results of the research by Ida Bagus Adi Mahayasa (2013) stated that the labor force has a positive and significant effect on production, if more and more workers are used, it will be able to influence the increased production produced. According to Jamli (2012) labor has a negative influence on production.
The hypothesis in this study explains that the alleged wage level has a significant effect on production. Based on multiple linear regression testing for variable wage levels, it can be explained that the wage level has a direct relationship with production where each increase in the wage level will increase production. This result is supported by the partial t test which concluded that the wage level variable has a positive and significant effect on production. The results of the study explained that alternative hypotheses were acceptable.
The findings in this study are supported by previous studies conducted by Perdana and Jember (2017), Putri and Kesumajaya (2017), and Arsha and Natha (2013). According to Perdana and Jember (2017) the level of wages is able to directly affect production which means the level of wages has a positive effect on production, the higher the level of production wages will increase as well. Based on research by Putri and Kesumajaya (2017) states that wage levels have a positive and significant effect on production. According to Arsha and Natha (2013) wages have a positive and significant effect on production. If the high wage level will increase productivity or work motivation which will affect the increase in production, conversely a low wage will also affect the motivation of workers. Wages that cannot meet the needs of workers will certainly reduce morale and motivation of workers, the impact of the ability of workers to produce output will be reduced.

Conclusion
The conclusion from the results of the study is the significance value for the variable capital of 0.002 <0.05 meaning that capital has a negative and significant effect on production. The significance value for the variable labor is 0.151> 0.05 meaning that labor does not affect production. The significance value for the variable wage level of 0.001 <0.05 means that the wage level has a positive and significant effect on production. Significance value for the