After a few weeks of commentary on wider issues relating to land and agriculture, this week I return to a focus on our field data, and the lessons this tells us for pressing policy issues in Zimbabwe. In this blog, the focus is on the implications of the complex outcomes of land reform for land administration.
As readers of this blog will know, we have been tracking 400 farms since the early 2000s across Masvingo province. One of the things we have been looking at is how plots change hands over time. This is important for a number of reasons. From this data we can examine the institutional mechanisms that have evolved for land transfer; we can look at patterns of land inheritance and the gender dynamics involved; we can identify emerging informal markets in land; we can identify the role of different authority structures in land allocation; and we can assess the level of turnover and the reasons for plot ownership transfer in the new resettlements.
All of these issues are important in thinking about new land administration and management systems in the new resettlements. With the land reform, such issues were not high on the policy agenda; the imperative was to distribute land and not think too hard about its longer term administration. But 14 years on, these issues are now significant, and ones that need to be addressed, as land administration systems are established. Understanding the existing situation and building on it makes much sense, but also future policies may wish to address anomalies and inequities that have arisen.
So in 2012, we looked at all households across the 16 sites (N=389), including in A2, A1 self-contained and A1 villagised (the former ‘informal’ sites have since been formalised as villagised A1 schemes), and compared the household listing made in advance of our mid-2000s surveys and the list pertaining in 2012. The period then relates to the time from establishment and ‘formalisation’ of ‘fast-track’ areas, including the issuing of ‘offer letters’, and 2012. This does not include the initial year or two of turmoil following land invasions when there was quite a bit of movement into and out of sites, as we described in our book (pp. 72-6). In this most recent survey, we asked about the reasons for the change in plot holder, what happened to the original holder, what relationships the new holders had to the previous one, the gender of the new plot holder, whether the new ownership arrangement had been confirmed in an ‘offer letter’, and the mechanism through which the new holder had got the land.
Overall, 19% of plots had changed hands over this period, with A2 sites seeing the highest change (29% of plots), followed by A1 self-contained (22%) and A1 villagised (15%). Across all sites, 57% of these changes were due to the death of the original owner. Exits for other reasons were 30% of transfer cases in the A1 sites and 75% in the A2 sites. In one A2 farm there was a particularly dramatic level of turnover, with a series of politically-motivated evictions occurring around the time of the 2008 elections. Overall though only 8% of all plots in the sample had changed hands during this period for reasons other than death (5% in A1 sites and 22% in A2 sites).
In nearly a decade this level of turnover is quite small, including when compared to previous resettlement experience. In 1983, for example, Bill Kinsey commented on the implementation problems of early post-independence resettlement, with many new settlers abandoning their plots due to inadequate planning and support. By the mid-1980s resettlement area populations had settled down and began attracting people, both kin and labourers, and there was a significant growth in household sizes. But after a long period of relative stability, old resettlement areas studied by Kinsey and colleagues have seen a significant out-migration in the period of economic collapse in the 2000s. In this period, people sought alternative options when agriculture became increasingly challenging, with shortages of inputs and absence of cash markets during hyperinflation, and household sizes shrank. This has included movement to new resettlement areas under the fast-track programme, often together with assets such as cattle which showed significant declines in the old resettlement areas.
In our ‘fast track’ land reform sites in Masvingo, death of the original plot holder (in all cases men) was the primary reason for change in ownership in 57% of all cases. In the A1 self-contained sites, death was the reason for 78% of changes, while in A1 villagised sites it was 68% and A2 it was 25%. Of the others, 28% left for other rural areas, mostly to original rural homes, although a few got new, better land in the resettlements. The remainder left for town (8%) or were evicted (7% – the A2 cases noted above).
Overall 31% of plots were taken over by women, very often the wife of a deceased husband. This varied between 44% (A1 villagised) and 17% (A2). There has therefore been an increase in women plot holders in the A1 sites. Yet few of these have been recognised in changes in ‘offer letters’, with only 23% of new holders being named in a new land occupancy registration document. Overall, 63% of all new plot holders were relatives, with 41% being immediate relatives (wife, son or daughter of the former plot holder). The remaining 37% of transfers were to people with no kin relation to the former owner.
49% of transfers were described as through inheritance, mostly to close kin (wife or children). A further 12% were passed to other relatives as ‘gifts’. 2% of transfers to relatives were mediated through a church or via the chief. There were a couple of cases in an A2 site where ‘purchase’ was identified, although no others admitted to money changing hands. However for some of those identified as ‘gifts’ of land, there were exchanges made for the infrastructure developed, with between one and four cattle being paid for such improvements. The remaining 34% of transfers were through other routes – 19% via the District Land Committee and 15% through Committees of Seven. However these patterns varied across sites with ‘inheritance’ dominating in the A1 sites (59%), while it was only the mechanism in 25% of cases in the A2 sites. Here the formal District Land Committee process was the most important at 58% of all cases. The local site based Committee of Seven was important in 29% of A1 villagised cases, but much less so in other sites.
So what are the implications of these data?
- Turnover in plot holding arrangements has been surprisingly low in the A1 areas in particular. This indicates a strong commitment to the new land, with only about 15% of all transfers involving people leaving for their former rural homes, and even fewer back to jobs in town. When a plot holder’s death occurs, transfers are usually to close relatives.
- The A2 sites show a different pattern, with higher turnover. There was a pattern of evictions (in one site, due to politically-motivated grabs), and in others death followed by kin-based inheritance was not the dominant pattern. Instead there were reallocations by the District Land Committee on vacation of the plot.
- The possibilities of effective transfer of underutilised land to new settlers seems more possible in A2 areas, where some indications of incipient ‘vernacular’ land markets are evident (probably underreporting of this due to illegality). By contrast in A1 areas transfers tend to be within families, and the opportunities for government to reallocate are relatively smaller.
- Despite the fact that District Land Committees are supposed to have control over land and oversee transfers, this is less important in A1 areas, where ‘traditional’ inheritance (or ‘gifting’) arrangements generally apply on the death of plot holder.
- The ‘offer letter’ system of land registration is not keeping up with land transfers, partly because so many are outside the formal system. Women are gaining land through inheritance on the death of male plot holders, and through transfers to daughters and aunts (relatively few). However women’s new ownership arrangements are largely not recognised in changes in ‘offer letters’, raising questions about the security of such tenure.
- The Committees of Seven, established soon after land invasions for local management of new resettlements, seem to be on the wane, particularly outside the A1 villagised areas where they seem to hold some land administration function for transfers outside kin-based arrangements.
- Chiefs, churches and other institutions for land administration seem to have a limited role, although ‘traditional’ leadership (sabhukus etc.) are often part of Committees of Seven in many sites, so there is an overlap in both people and roles.
- Outside the one A2 case where evictions occurred, conflicts over land transfer were very rare, and in most sites none were reported. This is surprising given the intensity of competition over land. We can expect such conflicts to increase over time, however.
- Across all land transfers identified no subdivisions of plots were recorded, and all were recorded as being handed over in full. This again is surprising given growing land competition and the need to inherit land by the next generation. However subdivisions are occurring in land still held by the original holder who is still present. Sons are allocated portions of a field and have their new homes within or nearby the original home compound. As described for the earlier phase of resettlement, such ‘vertically extended’ household arrangements are becoming increasingly common as a form of de facto subdivision.
Overall, as the Government of Zimbabwe attempts to define a new land administration system in line with the Constitution, these findings are clearly pertinent. There is a need for more effective registration of land and ownership, and also a need for a mechanism to keep this up to date. Ensuring registration of women as plot holders is essential. To encourage turnover and the making use of underutilised land is key, especially in A2 areas, transfers can potentially be facilitated through introducing some form of land market, assisted by land taxation; although again close monitoring and transparent registration (perhaps through leases) will be vital, especially to avoid the sort of evictions seen in one site.
Subdivision is always a controversial issue, as the thorny question of what is a ‘viable’ farm is raised. However this varies considerably and as demand for land continues to increase, there are potential subdivision options within sites, and indeed between them, including moving A2 farms to smaller A1 self-contained arrangements. In our sites, as witnessed by the lack of subdivision on transfer, and the limited amount of conflict (outside the particular site that was subjected to external grabbing), this may be a next-generation challenge, but clearly something that must be responded to.
As Marleen Dekker and Bill Kinsey argue in relation to the first round of resettlement, the direct benefits of land reform are significant in the first generation, but as things unfold, there are new challenges requiring different forms of responsive innovation. They note patterns we are already seeing in our sites, including the growth of household size through attracting labour, for example, as well as the de facto subdivision through ‘vertical extension’. These result in new land uses associated with new demographic patterns and changing social relations.
Trying to plan everything and imposing a narrow ‘straitjacket’ of rules and regulations only undermines potentials, they argue. There are, as Thomas Sikor and Daniel Muller note, serious limits to top-down state driven land reform. While new systems for land administration in Zimbabwe are clearly needed, they must be flexible, responsive and adaptable to new circumstances.
This work has been carried out together with the Masvingo field team: BZ Mavedzenge, Felix Murimbarimba and Jacob Mahenehene