LUMS study determines actual liability: 'cement sector involved in Rs 10 billion sales tax evasion'

The cement sector is involved in sales tax evasion to the tune of Rs 10 billion, as the sales tax paid by the sector stood at Rs 19.42 billion in 2013-14, which is only 66 percent of the actual sales tax liability. This has been mentioned in a recent study on cement sector, conducted by LUMS Lahore experts and handed over to the Federal Board of Revenue (FBR) for necessary action.
Sources told Business Recorder here on Wednesday that the study's estimates show that in 2013-14, exemptions of input tax on input value of fuel, chemicals, packing material, spare parts and lubricants come to Rs 21.02 billion while net sales tax liability after adjusting for cement exports and input tax exemptions @17 percent comes to Rs 29.425 billion. Average sales tax liability on each firm comes to Rs 0.981 billion. The sales tax paid by the cement sector in 2013-14 is Rs 19.42 billion, which is only 66 percent of the sales tax liability calculated by this study. In sum, we conclude that there is sales tax evasion in the cement sector to the tune of Rs 10 billion.
According to the report, tax evasion can take the form of misreporting of inputs and outputs by the manufacturing firms. However, the nature and extent of misreporting of firms cannot be detected easily as the undocumented part of the economy is quite large. Against these odds, experts adopted an econometric measure known in economics literature as the 'stochastic frontier approach' to benchmark best practice firms who are on the frontier. This benchmarking is done on the basis of firm performance in converting inputs into output. In technical jargon, this is termed as technical inefficiency of firms. Firms who operate on the best practice frontier are regarded as most efficient since they convert inputs into outputs most efficiently, while others are termed as inefficient. It is assumed that deviations from the best practice frontier are explained by managerial inefficiency or misreporting of data by the firms to evade taxes.
The results suggested that most of the variation in cement production across firms is explained by technical inefficiency effects across firms. In other words large variation in reporting of inputs and outputs has been detected by cement factories. Mean technical efficiency in the cement sector is 69 percent, which indicates that most cement factories are away from the frontier and that misreporting by cement firms is quite large.
In contrast to the evidence of effective input-output ratios, the benchmarking exercise suggests that most cement factories tend to misreport their factor inputs. If these firms are to be believed then lubricants am spare parts are the two most important inputs of cement production while fuel has insignificant share. We also find over-reporting of packing material by the firms.
To investigate the matter further, the data of best performing firms has been separated from the total sample and re-ran regressions to find that input shares closely match the reported figures. It needs to be clarified the stochastic frontier approach measures' relative inefficiency or misreporting. If cement firms are involved in misreporting as a group, the best practice frontier will be set at a lower level. In such cases, the deviation from the frontier would be very little. However, there is no basis for us to conclude that organized misreporting is taking place in the cement sector of Pakistan.
As per report, the evidence shows that mean technical inefficiency is 69 percent. Had the cement firms been fully technically efficient, they may have increased their production potential by 31 percent. It also estimated the value of cement production to calculate sales tax liability of the cement sector. Using the proportion of production in 2005-06 at the firm level, the quantum index for 2013-14 and adjusting this production for technical inefficiency or misreporting of production, we estimate that the cement sector produced Rs 340.24 billion worth of cement in 2013-14, which is 29 percent higher than the output value reported by the cement industry in 2013-14.
The analysis has also predicted the effects of changes in output tax rates on sales tax revenue in the cement sector. The experts used the partial equilibrium analysis to market demand and market supply elasticity of cement and measure the relationship between tax rate and tax evasion. This information has been used to predict tax revenue changes that may result by changing the tax rate. The tax revenue was calculated at different tax rates using the implied quantities and prices from the demand and supply curves and the tax evasion estimated, it said.
The report findings suggest that the current rate of sales tax at 17 percent falls to the left of the optimal tax rate that comes out to be 18.3 percent. The simulations suggest that increasing tax rate would maximise tax revenue. However, the revenues at the optimal tax rate come out to be marginally better than the current tax collection. It has been estimated that the increase in tax revenue will be only 0.57 percent. These results suggest that the current tax rate policy for the cement sector is very close to the efficient policy. It needs to be cautioned that this result changes if we change the assumption on tax evasion. Assuming tax evasion is 3 percent for every percentage change in the tax rate, the optimal tax rate comes to 23.4 percent and increase in tax revenue is estimated to be 9 percent. However, the baseline result that the current tax rate is to the left of the optimal rate is robust to alternative assumptions. It is concluded that simulations on the relationship between tax evasion and tax rate can be improved if the FBR shares with us the real time data on sales tax collection under different sales tax regimes since 1991. This exercise would result in more precise estimates for the optimal tax rate.
The report said that the estimates of the potential income tax collection for cement sector show that total profit is Rs 87.42 billion and the average profit made by each firm is Rs 2.91 billion. However, there is considerable variation in the profits across firms measured by the standard deviations. The highest estimated profit by a firm comes out to be Rs 10 billion whereas the lowest profit is only Rs 1.4 million. Based on an income tax rate of 34 percent, the industry as a whole should pay around Rs 29.72 billion as income tax. This implies that each firm, on average should by paying Rs 0.99 billion as income tax.
The results can be made more precise if data on exports made by each firm is made available to us. Some firms export a bigger fraction of their output while others export a small fraction of their output or do not export at all. Therefore, firms exporting a bigger fraction of their output pay sales tax on a smaller fraction and end up with a higher profit, experts added.
Audit officers need desegregated input-output ratios for conducting an effective sales tax audit. The firms in the cement sector may understate (overstate) inputs that are exempt (not exempt) from input tax. Therefore, share of desegregated input factors like fuel, electricity, packing material, spare parts, raw material, and capital, etc need to be assessed independently. The experts presented the estimated effective input-output ratios by taking data from the Census of Manufacturing Industries (CMI) 2005-06 of cement factories and verify the ratios with the regression analysis. The estimated input-output ratios were also verified from independent experts. The report particularly focused on those factor inputs on which input tax exemptions are admissible to the cement factories.
The result of the analysis revealed that the largest share in the value of production is of the user cost of plant (this is not the cost of new machinery, but the user cost of the entire plant and machinery). This share comes out to be around 32 percent. The second largest share in the value of production in cement is of electricity at around 27 percent. Fuel has 11 percent share in the production process. Other inputs with significant shares in the production value include raw materials with a 7.3 percent share, packing materials with 5.4 percent share, and labour with around 5 percent share. Other inputs like user cost of capital (3.3 percent) and spare parts (3.2 percent) have smaller shares than 5 percent of the production value. Lubricants have the smallest share at just 0.4 percent.
The second approach used to verify these shares was based on regression analysis on firm level data of 29 cement factories obtained from Manufacturing Industries (CMI) 2005-06. These results confirm our findings that fuel and plant unambiguously have the largest shares in output. The shares for both, plant and fuel come out to be around 29 percent and are almost identical to the shares displayed above. The share of raw materials comes out to be 13 percent share of lubricants is 2.3 percent whereas the composite input has a 26.9 percent share in output.