For the mid to late 1990s and early 2000s, the basic neoclassical growth theory predicts a steady Japanese economy, when in fact the Japanese economy was depressed. This study applies the new theory with intangible investment and non-neutral technology proposed by McGrattan & Prescott (2010) to the Japanese economy, and finds that the predictions derived from the new theory are much closer to the actual data. The improvement of this extension remains robust when tangible investment adjustment costs are added. This study is the first to apply the new theory to a country other than the US, and compared with existing literature on the lost decades of Japan, this study better explains the depression in labour hours by avoiding introducing a large, unjustified labour wedge or other exogenous inputs.