This paper proposes a simple but efficient method which is free from memorizing side information, for histogram shifting-based reversible data hiding (HS-RDH). To hide data into an image, a HS-RDH method modifies particular pixel values in the image, based on its tonal distribution, and the method not only extracts hidden data but also recovers the original image from the distorted image which conveys hidden data. Most of HS-RDH methods excepting a few methods should memorize a set of image-dependent side information. The proposed method is free from memorizing side information by introducing two mechanisms; one is estimation of a side information based on a simple statistic and the other is hiding a part of side information as well as data to the image. This approach makes the proposed method is superior to the conventional blind methods in terms of the quality of images which carry hidden data.