Translator style has been commonly understood as having both textual and contextual dimensions (Baker 2000, Munday 2008, Ng YLE 2009, Saldanha 2011). At the textual level, translators tend to have specific stylistic profiles in their translations that make them different from other translators' style, despite their similarities; whereas at the contextual level, the translators might have ideological perspectives that motivate them to produce a target text (TT). This paper, however, discusses these issues within the context of English-Indonesian translation involving non-literary text in particular. Investigation is based on a comparative model, theories of translator style and a corpus-based approach. The research adopts a qualitative method, particularly a textual analysis taking the form of a case study. Analyses of translation as a product are based on two sets of parallel corpora (Zanettin 2000) taken from two different editions of a similar economics textbook (the source text sub-corpora) and their respective translations in Indonesian (the target text sub-corpora). The elicitation of linguistic patterns in the TTs repeatedly used by the translators starts from producing a list of keywords in the study corpora by using WordSmith Tools version 5.0. These textual characteristics of each translator's work are then cross-checked with the translator concerned through in-depth interviews to uncover the ideological motivation in selecting specific language patterns in the translation. This research reveals two major findings: (1) at the textual level, each of the two translators has his own specific linguistic choices in the translated economics text (apart from similar linguistic features); (2) at the contextual level, translator adopt an ideological perspective in translating the source text (ST), either sourcelanguage oriented (i.e. overt translation) or target-language oriented (i.e. covert translation) (House 2006). These findings are also supported by the information obtained during the interview concerning the contextual features.