Migration has been ubiquitous since the evolution of early humans. Historical reasons for large-scale movement of populations were probably related to the search for resources, safety from predators, and the need to get away from difficult living conditions. We know that in the Indian sub-continent there have been early and later migrations from Africa, South East and Central Asia, China and Europe. And there may be even more movement of people in the near future as a result of climate change, which is expected to have a defining impact on human migration.

What are the likely levels and characteristics of climate-change induced migration in South Asia? Most studies on migration at time-scales of a generation or less in India have relied on analyses of census and National Sample Survey Organisation (NSSO) data. These surveys, however, suffer from having strict and narrow categories for people to provide 'reasons' for migration.

Second, gradual change and cumulative impacts of various drivers of migration are difficult to recognise, and when people are asked why they moved they simply say, "to find a job". Environmental degradation such as deforestation and some of the effects from climate change would probably not be recognised as the main causative agents. Only immediate effects such as displacement relating to building a dam, and dramatic/profound changes, such as being forced to move out of one's country, as a result of climate related causes such as in the case of sea level rise would be remembered and cited as the primary reason for migration.

The state of migration

Migration, which has been an area of interest for academics, has increasingly begun to concentrate on rural-urban (RU) aspects, as there has been a general belief that following trends in Latin America and Africa, Asia is going along the same route. However, some people argue that such RU trends are not so evident in Asian cities. In contrast to the experience of other regions, the prediction is that in Asia only the least developed cities will see an increase if any. The contribution of RU migration to the increase in urban population in India was 21 per cent during the 1990s and census data analyses suggest that over the last few decades, overall migrant population is generally about a third of the total population.

But there is still plenty of debate on various questions: Is such RU migration accelerating in India, are the cities people move to mostly 1 to 5 million in size, smaller or larger, should people be discouraged from moving into cities even if that has been shown in general to improve economic and living standards, and what should we do. None of these are clear. Indeed there is no agreement on what constitutes the right response.

According to an NSSO report that came out in June this year, about a third of Indians are migrants, but within this there are some interesting differences in the rural-urban and also in male-female distribution patterns. Women migrate mostly for marriage, while among men the single largest reason is employment. Migration because of natural disaster, displacement due to a development project and socio-political reasons is classified as 'forced migration' and this is 2 per cent each in rural and in urban migrant households. Education, natural disasters, housing problems or house or land acquisition, healthcare, retirement, and marriage are among the other reasons for migration.

Contrary to the belief that the poor are forced to move, or that poverty is a major push factor for migration, we find that it is the educated and the wealthy who are most likely to migrate. The report indicates that the poor, the illiterate, those who are part of backward castes and tribes do not migrate as much. This suggests that lack of access to resources, social networks, and social and economic status become hindrances to migration. Migration in itself improves a person's life but those who have most to gain from it, the very poor and marginalised, are the ones least likely, or able, to do so. This particular observation is similar to what was highlighted in the 2009 Human Development Report.

For the most part, experts working on migration in the region have not considered the impacts of climate change in evaluating future trends in migration. Internationally, there has been debate on migrants and environmental refugees for a long time. Forecasts of how many will be displaced in 2050 by climate change vary widely from about 25 million to 1 billion. The problem lies not only in the uncertainty regarding future climate change impacts and coping measures, but also in the difficulty in estimating the interaction among the several factors driving migration.

What to expect with climate change

Globally, climate change is expected to result in droughts, heat waves, melting of glaciers, famine, changes in patterns of many diseases, and sea level rise. Each of these can lead to a range of effects that can then have various secondary impacts and these changes are likely to contribute to temporary migration. For instance, if precipitation drops in already drought prone-areas such as Bundelkhand, seasonal migration may increase and perhaps become permanent.

Developing regional policies on forced displacement and labour is important and since the Refugee Convention does not recognise climate change migrants as refugees, a new international agreement is required to address the plight of these people, especially those who will be rendered stateless.

 •  Unemployment and migration
 •  Migrant labour, migrating debt In case of famine, it may be difficult to indicate the extent to which climate change forced the migration, but this is relatively easier to understand with sea level rise (SLR). Most experts now expect at last a metre rise by 2100 with effects such as salt-water intrusion, an increase in the intensity of storms, fresh water shortage, reduction in agricultural yield and inundation. All these difficulties will force many people to move well before they are inundated. Globally, SLR will lead to migration of people out of their low-lying islands, delta regions and coasts. Around the world, about 60 million live within 1m of mean SLR and roughly ten times that number live in an area that will be inundated by a 10m SLR.

In the South Asian region, climate change is expected to have dire consequences: variability in monsoon patterns with a likely increase in the number of days with heavy precipitation and flooding, a severe shortage of fresh water and the melting of the Himalayan glaciers leading to an exacerbation of drought in many parts of the Indian subcontinent, reduction in agricultural yield eventually leading to widespread famine. Diseases such as cholera will intensify with water shortage and vector borne diseases such as malaria will rise.

Since the region has small islands, delta regions and long coastlines with densely populated cities, SLR will certainly force people to move inland or out of their islands and delta regions. Being one of the most densely populated regions of the world, the adverse impacts of climate change will be experienced by millions of people.

Based on the information we have it is difficult to predict how migration numbers will change with the impacts of warming. Given the scale of uncertainties, most reported numbers are good guesses. Perhaps more people will move to other rural and urban areas with an increase in droughts, fall in agricultural yield, water shortage, and floods. People will be vulnerable to many effects: globalisation, water shortage, famine, poverty and so on. They will experience them simultaneously and respond to them based on how these impinge on their daily lives. Seasonal migration during floods or droughts may intensify and then become permanent.

We need to understand the tipping points for migration and the numbers of people who are already migrating in different parts of the country as a result of environmental causes. While triggers of migration are complex, there are certain patterns that are becoming clear. Migration tends to take place in waves to places where people have existing family or social networks, cultural affinities and economic opportunities. The young tend to move early, creating further stresses in the remaining population and intensifying future migration. Old people tend to leave last, but do not necessarily return first.

In the unique instance of SLR where numbers might be relatively clearer, how many people do we expect will migrate? With its long coastline and densely populated coastal cities in addition to the delta region of the Ganges, India can expect tens of millions of people to be forced to move. Up to 80 million people living along the coast may be forced to migrate inland as a result of SLR. In case of a 3 to 5 metre sea level rise, Dhaka, Mumbai, and Kolkata could potentially be significantly de-urbanised.

Another important consideration is the SLR expected in Bangladesh. About half the population of Bangladesh lives within 10 metres above sea level and flooding currently displaces 500,000 people each year. Although flooding and temporary migration are a way of life for many people in this region, the frequency of 100-year storms is also expected to rise and with that the number of people who are forced to move is expected to increase. Since Bangladesh has a large delta region in a small country, many of the close to 80 million who might be affected by rising seas may have no place to go except their hinterland and their neighbours.

What should we do?

India cannot afford not to take a proactive approach to climate change and migration. Adaptation measures in key sectors such as water conservation, agriculture, urban planning and coastal management will improve resilience and reduce the pressure on migration from climate change.

One may consider, for example, that coastal planning and infrastructure development need to be regulated. While existing investment of crores of rupees is not resilient to SLR, infrastructure development continues along India's coastline as ports, highways, airports, special economic zones, industrial parks, upscale hotels, and housing developments are proposed. Mal-development along the coast already leads to salt water intrusion and destruction of coastal ecosystems. Thus in the case of SLR as in other areas of climate change, development challenges are not always distinct from adaptation, but some of the specific problems that we will encounter with warming will be relatively distinct while at the same time exacerbate development problems if they are ignored.

There are particular knowledge gaps on internal migration in India, which need to be located. We need to do a better job of tracking migration and its causes, identify indicators and thresholds, so we understand how decisions on migration are made. We need an understanding of multiple vulnerabilities in particular ecological zones in India in order to adapt better to warming. In situations in which migration cannot be avoided we need to have policies in place to prepare for a proactive phased movement of people. Provision of skills in advance of migration, civil and legal rights and appropriate labour policies are all relevant to preparing for the changes in migration patterns.

If the global community and South Asia are to survive climate change, it will require us to learn that we cannot engineer our way out by merely using new technologies. We may have to re-imagine a more cosmopolitan South Asia, redefine our identities and what we mean by culture.