The Result of The Programme of International Student Assessment (PISA) on 2012 shows that Indonesia is on 64'th position from 65 countries in Mathematics Mean Score. The 2013 Learning Curve Mapping, Indonesia is included in the 10th category of countries with the lowest performance on cognitive skills aspect, i.e. 37'th position from 40 countries. Competency is built by 3 aspects, one of them is cognitive aspect. The low result of mapping on cognitive aspect, describe the low of graduate competences as an output of Indonesia National Education System (INES). INES adopting a concept Eight Educational System Standards (EESS), one of them is graduate competency standard which connected directly with Indonesia's students. This research aims is to model INES by using KSIM cross-impact. Linear regression models of EESS constructed using the accreditation national data of Senior High Schools in Indonesia. The results then interpreted as impact value on the construction of KSIM cross-impact INES. The construction is used to analyze the interaction of EESS and doing numerical simulation for possible public policy in the education sector, i.e. stimulate the growth of education staff standard, content, process and infrastructure. All simulations of public policy has been done with 2 methods i.e with a multiplier impact method and with constant intervention method. From numerical simulation result, it is shown that stimulate the growth standard of content in the construction KSIM cross-impact EESS is the best option for public policy to maximize the growth of graduate competency standard.