If one looks at the job description of any 'strategic' HR position, it is highly likely that 'management of change' (or 'driving change initiatives') would feature very prominently. While I fully agree that organization-wide change management efforts are important in fast changing business environment, I am finding that my interest these days is more on the the psychological process of dealing with change (the 'transition') and on developing change resilience in individuals. Having been 'at the receiving end of organization level change management efforts' many times in my career, I am not sure as to what extent these are really effective. Often they degenerate into communication programs (at best) and con games (at worst). Unless the organization can create a credible value proposition ('what is in it for me') for the impacted people the chances are that the above degeneration would happen. It can also been argued that 'second order change' can not be managed (in the usual meaning of the term 'manage'). In this context, helping the employees to become more change resilient becomes more important.
I also feel that the impact of change on the 'psychological contract' between the employee and the employer is often not given adequate attention. The violation of the psychological contact could be one of the key reasons for 'change resistance' and negative outcomes like attrition, lack of motivation etc. Often employees feel that they are 'taken for granted' in the name of 'flexibility' and 'organization responsiveness'. Of course, organizations have sound business reasons for making these changes (realignment, restructuring etc.). My point is just that often the impacted employees (who have been 'realigned') feel that the psychological contract has been violated because of what they perceive as 'unilateral changes made by the organization'. (See a related link here)
Coming back to the 'HR job description' mentioned at the beginning of this post, there could be additional factors (apart from skill set related factors) that limit the ability of internal HR professionals to manage change. For example, often HR professionals get involved too late. By that time 'emotional wounds' have already been created and what is left is more of communication and 'dressing of wounds'. While this is useful, this is not change management. This is more of 'damage control'. Of course, in many situations the internal HR professionals themselves are experiencing the same adverse effects of change and hence this could further limit their ability to carry out their 'change management responsibilities'.
Note : Another related aspect (to organization-wide change efforts) is 'culture change initiatives'. There are many 'levels of culture' (like artifacts, norms, values, basic underlying assumptions etc.) at which an intervention can be made. Technically speaking, to be fully effective, culture change has to happen at the 'basic underlying assumptions' level (as per Schein's model). This would mean that 'culture change' has to happen in a bottom-up fashion (starting with the individual) as these assumptions reside in people's minds. However, the difficulty is that often a clinical intervention is required to surface and change these assumptions. This is usually too much to manage in the context of an organization-wide change effort. I think that the 'basic underlying assumptions' & 'world view' of a person are unlikely to change unless he/she is faced with a very significant event (often a traumatic event) in life. So it might not be realistic to make an intervention at this level in the context of an organization level change. Anyway, since one is likely to change many jobs during one's career, one can't afford to get influenced by organizations at such a deep level!!!
May be what can be attempted is to create a rational reason for behavior change. This does not necessarily mean 'carrot-and-stick' in the usual meaning of the term. The 'reason' could be aimed at any level in the hierarchy (e.g. Maslow's hierarchy) of human needs (including esteem and self-actualization) and not just at the lower level needs. This would also mean creating a context (including 'role models') where the desired new behavior has a higher possibility of emerging and thriving.