HISTORIC PRESERVATION AND GENTRIFICATION: THE ARMORY DISTRICT, PROVIDENCE, RHODE ISLAND

In studying gentrification, commentators observe that upper mcome groups are attracted to historic districts inhabited by lower mcome groups, and which are forced to leave because of rising prices, rents and taxes. Does this mean that all historic preservation efforts lead to displacement of lower income groups? This study looks at the Providence Preservation Society Revolving Fund program in the Armory District on the south/west side of Providence, Rhode Island. After a review of the literature on gentrification and neighborhood change, a time series study from 1950 to 1990 traces changes in the census tract which encompasses the Armory District and three comparison tracts. Census data are used to trace changes from 1950 to 1980 to establish the status of the four tracts in 1980 when the program began.(T he indicators are: median income and income distribution; percenta e of blacks, the arg.est minoritY~r this period; po ulation ch e; housing_te__nur.e; and condition and value of housing/ These indicators are measured against national and/or city norms. Preliminar 1990 census data are u _ple.m,e ~nted by other sources: cit directories, advertised ren!s, house sales and building ins2ector records for the period after 1980. Revovling Fund


Table of Contents
public concern. While neighborhoods have always changed, this issue came into prominence beginning in the late 1970s. Many observers linked the process to the rehabilitation of buildings in historic districts. Changes in the 1970s also spurred new theoretical approaches to neighborhood change.
In looking at the causes of neighborhood change, of which gentrification is one manifestation, some theorists emphasize the role of individual household decisions, while others stress the broader economic framework. Of the economic interpretations there are two approaches, those who analyze the workings of the a free market economy, and those who see economic changes in terms of broad national and international structures.

~-Definitions
The terms gentrification and displacement need definition, because they are often confused in the public mind and differently defined by scholars. Gentrification is a term borrowed from Great Britain. Smith (1986,1) accepts the American Heritage Dictionary definition. Gentrification is the "restoration of deteriorated urban ~ property, especially m working class neighborhoods by the middle and upper classes." The Oxford American Dictionary definition, used by Nelson (1988,1) is similar. Nelson (1988, 119) does not regard new housing in an area as gentrification, only the upgrading of existing housing. London (1980, 79;Palen and London 1984, 7-9) objects to the term gentrification as value laden and prefers "urban reinvasion" which may mean "takir:ig back" areas previously upper class . Gale (1980,13) uses the term "resettlement".
There can also be gentrification of commercial areas by which upscale shops replace local services, or business districts displace residences or smaller commercial areas. This aspect of the issue is not of concern in the Armory District which is purely residential.
"Displacement" is more general, and gentrification is but one form of displacement. Displacement occurs if a move is "necessitated by housing and neighborhood related factors beyond the household's control... and these factors make continued occupancy infeasible " (Nelson 1988, 114n). Marcuse (1986, 154-7) lists a number of circumstances that can cause displacement. It may be the result of disinvestment, which cuts building services to make a building almost uninhabitable, or actually abandoned. Abandonment of a neighborhood by other residents can lead to a chain displacement.
This suggests that neighborhood change, as well as change related to a particular building, may be a cause of displacement Reinvestment may make a housing unit unaffordable because of rent or tax increases. This may come as a result of large public projects, such as highways or urban renewal, or from what Zeitz 2 (1979) calls "private urban renewal." Marcuse also counts as displacement situations in which one family moves voluntarily, but another in similar economic circumstances cannot afford to move m.
Such a situation might lead to gentrification. A more narrow definition would not accept the latter situation, and would restrict displacement to eviction or demolition (Lee and Hodge 1984, 144-6).
A further question arises with Marcuse's definition because there may be situations in which a family of similar income cannot move m because the building has become uninhabitable. To make it habitable would require investments and hence higher rents, or subsidies.
Most commentators agree that eviction for cause does not constitute displacement. This would presumably include mortgage foreclosure. Downs (1981,4, 24) raises the further problem that under normal circumstances Americans move frequently, with renters moving far more than owners. It is not easy to determine if the move is voluntary.
Displacement is the forced removal of lower income groups through a variety of causes, while gentrification refers to the actions of higher income groups which make an area unaffordable for lower income groups.

Discovery of the Gentrification Issue
During the post-World War II years, the expansion of suburbs drained inner cities of middle-income households, not offset by inmigration . Cities were left with a disproportionate number of lower income households. Prevailing theory saw this decay as an 3 inevitable process as neighborhoods aged and the more well-to-do populations moved outward.
It was therefore with considerable fascination that observers in the middle and late 1970s began to note what seemed to be a reverse phenomenon. Neglected neighborhoods were attracting reinvestment, and the "back-to-the-city" movement was discovered (Pattison 1983, 88;Marcuse 1986, 159;Nelson 1988, 17). Further investigation suggested that, in fact, those reinvesting in city neighborhoods were not abandoning the suburbs, but rather were those who stayed in the city (Gale 1980,100;Marcuse 1986, 17;Nelson 1988, Ch.3). Some saw the new investment in an optimistic light as helping to restore declining cities (Nelson 1988, 15;Degiovanni 1984, 68-7 4 ), while others expressed concern about the displacement of vulnerable groups (Gale 1979, 301;Nelson 1988, 11;Smith 1986, 3-4).
Early estimates of the extent of this influx varied, and much of it was based on "soft data" (Lee and Hodge 1984, 141-9). A 1975 mail survey of 260 central cities undertaken for the Urban Land Institute, (Black 1980, 3-12) found renovation underway in 48 percent, particularly larger cities predominantly in the South and Northeast.
There was a total of 54,000 renovation projects since 1968, a small proportion of the almost seven million new houses m the period. Renovation tended to be in small enclaves and in predominantly single family housing, with 75 percent owner occupied. Sixty-five percent of the enclaves were local or national historic districts. Other commentators also point to the attraction of historic areas (London and Palen 1984, 11;Beauregard 1986).
Historic neighborhoods play a role in many case studies: e.g.
Columbus, Ohio (Fusch 1980, 150-69); Washington, D.C. (Gale 1980, 95-112;Zeitz 1979, 70-2); Philadelphia (Houston and O'Connor 1980, 300;Levy and Cybriwsky 1980, 138-48); New Orleans (Laska and Spain 1980, 120); and Charleston (Tournier 1980, 173-84) The Gentrifiers Those identified as gentrifiers (Gale 1979,293-98) are predominantly white, with many single or two-person households and few with children. They have close to the national median mcome. The largest proportion are in the 25-34 age group, and the next largest group in the 35-44 year age range. In Gale's study, most (62-97 percent, depending on city) had four years of college, compared to only 14.7 percent nationwide, and between 61 and 87 percent had graduate degrees. More than half were professionals, and another large group consisted of managers and administrators.
Sales and clerical workers made up all but a small proportion of the rest. Other studies echo these descriptions, albeit with different percentages for the various characteristics in different locations (Gale 1980, 95-101;Palen and Nicholas 1984, 133;Clay 1980, 27;Legates and Hartman 1986).
A number of reasons for the influx of such households have been proposed. The large size of the baby boom generation increased the number of households, and therefore demand for housing (Goetz and Colton 1980,184;London and Palen 1984, 14 ).
Tax incentives and the relative cheapness of city, compared to suburban living, made it a good investment (Nelson 1988, 15, 45).
The oil crisis, which added to commuting costs, may have at times contributed to the choice of city living.
Changing social conditions also contribute to changing value systems (Beauregard 1986, 43 ). There are more women in the work force, many by necessity, which makes city living more convenient and contributes to a concomitant delay in marriage and childbearing.
Veblen's theory of conspicuous consumption may account for the attraction of urban life, with dining out, shopping in boutiques and attending cultural events. The attraction to historic houses may be part of this display (Beauregard 1986, 43;Jager, 1986).
In addition to the taste for historic buildings there is an attraction to ethnic neighborhoods (Allen 1984, 29-30) which have greater densities than most suburbs and a more participatory life style. Both attractions may reflect a desire for a link to the past, perhaps romanticized.
t There is a distinction between gentrification and "incumbent upgrading" (Clay 1983, 22-7), which is investment by existing residents. This may be caused by an improved image of the neighborhood as a place in which to invest, or by a change in household needs as a result of the stage of life. Incumbent upgrading may be fostered by a neighborhood organization, such as those established under the Neighborhood Housing Service program.
Clay (1983) maintains that such neighborhoods differ in important respects from those which attract gentrifiers. They have fewer professional and white collar residents, and more blacks and elderly.
The housing may be poorly maintained, but not dilapidated.
Gentrifiers, on the other hand, are more likely to move into more dilapidated areas where there , is much vacancy and which are predominantly white. Gentrifying neighborhoods are also more likely to be close to the Central Business District (CBD) or a half mile away, and to be in areas of special amenity, such as hillside or waterfront locations, or those with historic buildings. Thus one can more easily predict where gentrification is likely to happen than where incumbent upgrading may take place. This might happen anywhere. In London, England, gentrification occurred close to areas that were already high status (Machielse 1987, 62).
After-rehabilitation values, both sales and rentals, are higher m gentrifying areas than in upgrading ones and multi-unit buildings are likely to become condominiums (Clay 1983, 30-1). For gentrifiers, location is more important than the original condition of the building, while for upgraders the condition of the building is more important.
Upgrading may be the result of "social mobility without spatial mobility" (London 1980,78). The second and third generation, who inherit from parents or move back to the old neighborhood, may be responsible for the upgrading. This may involve a social class change.
To sum up, gentrifiers are predominantly white, young, in small households, relatively affluent and well educated. Many are m managerial or professional occupations. The causes cited are: the size of the baby boom generation, the increased number of women in the work 'place increased the demand for housing, which was often cheaper in urban areas. Other personal reasons for selecting city living, such as a need for conspicuous consumption, are less easy to document.

Stages of Gentrification
The gentrification process takes place in stages (Pattison 1983, 77-91;Gale 1980,103-11). In the first phase, an unconventional group is the first to move m. It might be gays, artists, interracial couples, or others who have difficulty in finding conventional, affordable housing. They are oblivious to risk because they basically seek a place to live, not invest. They undertake much of the work themselves and are largely self-financed, or have a seller's mortgage. Banks are reluctant to invest.
In the next stage, local realtors, or others, market houses that become vacant through death or normal out-moving, to professionals, managers, sales, and clerical workers. At this stage, some existing residents begin to improve their properties, while absentee owners are tempted to sell or raise rents. Displacement is minor. Banks are still cautious and tend to base decisions on individuals, not property, and to finance only a low portion of the value. As acquisition costs are still low, buyers at this stage are ready to accept the risk for the sake of a bargain. A neighborhood organization may form in this second stage to press the city to address such problems as rowdy bars and houses of prostitution.
In the third stage, those who were cautious earlier feel the investment is safer; banks also become less cautious. Renters are now displaced, particularly in rooming houses. By the fourth stage the strong demand leads to the conversion of previously non-residential buildings and the in-movers are increasingly young and in the professional, managerial, clerical, and sales categories. Many rent in the area before buying. It should be noted that this presents a model for a "trickle up" theory as opposed to the more common "trickle down" theories (e.g. Burgess, Park and McKenzie 1967;Hughes 1974, 75;Birch 1974, 79-84).

Those Displaced
Estimates of the displacement of vulnerable groups vary (Lee and Hodge, 1984, 141-9). A 1979 study by the Department of Housing and Urban Development estimated 350,000 households annually were displaced between 1974 and 1976, but a study undertaken by Legates andHartman in 1981 (Legates andHartman 1986, 197) arrived at a figure of 2.5 million displaced nationally. With a fifth of renters movmg every year, it is hard to estimate how many are displaced. Lee and Hodge (1984, 149) (Nelson 1988, 18;Lee and Hodge 1984, 141).
On the other hand, concentration of gentrification in only a few neighborhoods of any one city, and its slow pace, may mean displacement is fairly low and may not be taking place only in gentrifying neighborhoods (London and Palen 1984, 13 ).
The profile of displacement families does not appear to follow the clear pattern discerned for gentrifiers (Legates and Hartman 1986, 184-96), but out-movers are not easy to track, and estimates must be made on the basis of demographic changes. Blacks are frequently cited as victims of displacement even when they are not heavily represented in gentrifying areas (Spain 1980,28-39;Legates and Hartman 1986, 184-98;Lee and Hodge 1984, 142-3) The poor elderly and households headed by single women are displaced in significant numbers (Legates and Hartman 1986, 184-98;Lee and Hodge 1984, 142-3). Characteristics of those displaced vary with size and prosperity of the city, the local housing market, housing characteristics and overall demographic composition (Lee and Hodge 1984, 154). Renters are particularly vulnerable to displacement.
Obtaining nationwide figures for those displaced has proved difficult because reasons for moving cannot be readily determined (Downs 198 I,4,24;Nelson I 988,I 8), and studies of individual neighborhoods may give an exaggerated notion of the overall phenomenon. Nevertheless, these studies show that the greatest impact is on those groups least able to bear the cost.

Nationwide Overview
A recent book (Nelson 1988) addresses the issues of where and why gentrification takes place in a broader and more long-term perspective, and seeks to discover the relationship, if any, between moving patterns and the gentrification of lower income neighborhoods. Nelson uses the 1980 census for a comprehensive study of migration patterns within and between urban areas, and between census tracts, of 40 major metropolitan areas. She finds (1988,8,17,19,52, that although out-migration to the IO suburbs slowed in the latter half of the 1970s, the out-migration of higher income residents continued and that gentrifying neighborhoods were a small proportion of the total number of neighborhoods. In-movers were not in significant enough numbers to counteract the decline of inner cities, and all cities lost population (Providence seven percent over the decade ;Nelson 1988, 128) Wishful thinking and private interests encouraged overstatements about gentrification . "Gentrification appears to have reflected shifts in investment within and among cities in the context of ongoing decline, not a major break from past trends toward city decline." (Nelson 1988, 24 ).
This does not mean that the poor were not displaced, but policies to counteract the trend need to address the reasons why people choose to stay in the city. Nelson (1988, 9) proposes four possible explanations: a renewed preference for city living, employment opportunities in service industries that made residence close to work attractive, demographic changes that created demand for different housing types, and less expensive housing in the city than the suburbs that offered a good investment. Nelson finds (1988, 57-60) correlations between stages of life and migration patterns. Single person households are more likely to live in cities, and husband-and-wife households with children more likely to live in the suburbs. Older people are less likely to migrate, so that a decrease in out-moving might reflect aging in the population, while increased out-moving could result from growth in families with young children. The slight decline in out-migration of higher income males and higher status whites was nevertheless offset by continued net out-migration in most cities (1988,72).
The decline in out-moving was among whites, not blacks, and most noticeable in the 20-year-old age category (Nelson 1988, 155).
This was particularly true of young college students, graduates and professionals, those who might fit the definition of gentrifiers, but the change was marginal and few cities benefitted by a raised total median income.
Using categories established by Noyelle and Stanbach, Nelson (1988, 98-103) finds cities with advanced service economies somewhat more apt than manufacturing cities (with Providence as an exception) to attract upper income in-movers (see also Lipton 1980).
Relative housing bargains compared to the suburbs was the next most likely factor to attract in-movers , but again the overall cumulative city decline was hard to reverse.
Although out-migration lessened, this does not prove that migration patterns contributed to gentrification. To study this question, Nelson (1988, 113-27) looks at specific census tracts in 10 cities to determine which type of neighborhood is most vulnerable.  (1988,125). The income growth at the neighborhood level was less than the total to be expected. She finds that, "when above average income growth occurred in the 1970s, it took place more often m lower income than in upper income tracts in all cities . When neighborhood mcome rose at above average rates, gentrification of poor tracts was more common than further gains in pockets of plenty" (Nelson 1988, 125-6).
Loss of population in a census tract might be due to reduced density in gentrifying areas as small units are combined into larger (Nelson 1988, 128-34). This did not seem to be a factor in the 1970s because cities were "thinning out" and gentrifying tracts even gained population in some cases. In most cities, however, there was an increase of the poor in non-gentrifying tracts. There was also a loss of population in tracts with a high poverty rate in 1970 into tracts which in 1980 had a somewhat higher income. This suggests there may have been abandonment in the poor 1970 tracts. That the poor population shifted more dramatically than the population as a whole is consistent, Nelson found, with the view that they were being disproportionately displaced. They moved more frequently into tracts with slower income growth than to poor tracts, but there was also evidence of concentrations of the poor.
The Providence-Pawtucket-Warwick metropolitan area was one of the few metropolitan areas where upper income in-movers increased in the 1970s and increased more than that of lower income in-movers (Nelson 1988, 149). "Upper income movers were 20 percent more likely to choose the central cities of the Providence-Pawtucket-Warwick area than the cities' share of metropolitan population would suggest" (Nelson 1988, 97 Nelson's study shows that all cities lost population during the latter part of the 1970s, although at a slower rate than earlier in the decade. Those who remained were likely to be young and educated. Cities with advanced service economies and those with relative housing bargains were likely to attract in-movers and stall outmovers (Nelson 1988, 125). This led to gentrification in a few neighborhoods, but the overall impact was slight. In Providence the census tracts near Brown University saw a rise in median incomes.
These figures predate the Armory District project, and do not show this tract as gentrifying.

Theories of Neighborhood Change
Whatever the strength of the "back-to-the-city movement", it was one of several developments that stimulated new thinking about neighborhood dynamics (Palen and Nicholas 1984, 128). The rise of the neighborhood organization in response to urban renewal and highway construction was another.
In looking at causes for neighborhood change, theorists have taken a number of different approaches, although there is overlap between them. Some (Goetz and Colten 1979;Sternlieb et.al.1974;Leven et.al. 1976) see the perception of the neighborhood by various actors as significant. These analysts often assume the workings of the free market, without addressing the broader reasons for its ups and downs. Another approach looks at the issue from the point of view of individual decision making. Rossi (1980) pioneered the study of mobility, which has important implications for neighborhood change. More recently, Galster (1987) proposes a complex integration of individual and social forces on housing decisions. Others (Williams andSmith 1986, Beauregard, 1986) put more emphasis on the broader economic structure of the economy.

Traditional
Theory.
Neighborhood theory in America began in the 1920s with the Chicago School, notably the work of Park and Burgess (Zeitz 1979, 9;Sternlieb 1974, 322), which continues to influence thinking despite -many qualifications and disagreements. This theory viewed the city as a series of concentric rings, with the CBD at the center. There is a continual outward movement of the more well-to-do as they move from older housing to newer, and to larger lots. Market forces trigger a "filtering" process as families rise in the social scale, move out and are replaced by lower income groups.
There have been many amendments to this basic thesis and criticisms of its assumptions and determinism. The theory did not foresee the segregation that followed the influx of blacks to the North after World War II which did not follow the patterns of the Park-Burgess model (Zeitz 1979, 9). It offered little explanation for the new back-to-the-city movement.
Nevertheless, the image of an inevitable decline of inner city neighborhoods was prevalent in the literature and a number of descriptions of the stages of decline were well known (Hoover and Raymond 1962). Government policy accepted the model in such matters as Federal mortgage insurance policies which based 1 6 evaluations on the social make-up of a neighborhood (Bradford 1979, 324-31 ). The theory became a self-fulfilling prophecy.

Neighborhood
Change.
More recent students (Leven et. al. 1976) draw attention to the impact of changes in one neighborhood on adjacent neighborhoods, rather than expansion of the CBD, as the impetus for change. Unlike Burgess and Park, they maintain no inevitable direction for change.
They put more emphasis on the decision of individual households rather than market forces . Expectations about the future of the neighborhood proves more important than any actual changes (Leven at. al. 1976, 94, 110, 118).
,Conceiving of housing as a bundle of characteristics, Leven et al (1976, 34-42)  These transition areas first expenence a drop in values in advance of local change, either of racial of socio-economic character, reflecting expectations of middle-class residents as they observe the adjacent lower income neighborhood expand. Exodus of the higher income group then propels further decline. This is a reversal of the gentrification process.
Some neighborhoods resist the process, such as cohesive ethnic neighborhoods, those close to universities and similar institutions, or those screened from adjacent neighborhoods by a park or other desirable features. (Leven et. al. 1976, 110-8) When a neighborhood begins to change, or 1s already derelict, owners, and even more so landlords, neglect maintenance.
Reinvestment in neighborhoods must therefore include the nonstructural aspects as well. Historic preservation can be an approach to neighborhood revitalization.
Inter-neighborhood reactions, particularly expectations of the future, and "who lives in the neighborhood", play a role in neighborhood change. This might support the finding that gentrification takes place close to areas of stability. (Machi else 1987, 62;Nelson 1988, 114). Cohesive neighborhoods, often ethnic, also show signs of stability.

Role of Landlords.
A number of studies concentrate on the motivations of landlords. A landlord's decision to abandon is not purely an economic one, but based more on the expectation of the future of the neighborhood than on the economics of the building (Sternlieb et. al. 1 8 1974, 321-9 In a rising market, traders or speculators enter the scene, usmg leverage to gain in resale from the appreciation of the market.
Selection of tenants and maintenance are of secondary importance.
In a weak market, owners seek the highest annual return, with the mm1mum of investment. These are the slumlords who milk the property, are often delinquent in taxes, probably have a seller's mortgage and virtually ignore tenant selection procedures.
When rehabilitation becomes necessary and feasible, two types appear. "Shareholders" are professional investors who favor limited partnerships and hope to make money on tax advantages.
"Rehabbers" are adept at political manipulation and the red tape 1 9 required to take advantage of federal and other programs involving tax breaks and subsidies. They "make their money at the front end and ownership is incidental" (Goetz and Colton 1979, 86).
Condominiums may be at the end of this process.
"In the long run, even responsible interests bow to economic realities and sell to interests that can profit" (Goetz and Colton 1979, 88). This may be arson-for-profit at the low end of the market, or in situations of rapid shifts.
Neighborhoods change with the ebb and flow of the market, and gentrification is part of this process. Two factors can exaggerate these trends (Goetz and Colton 1979, 11, 25, 95-103 (1979,95). Where demand exceeds supply, public policy could drain off excess profits through rent control or tax. policy, but if these policies are applied in a declining market they will be counterproductive.
Landlords play an important role in neighborhood dynamics.
The type of landlord can be an indication of the local market, and economic stabilization must include ways to attract more responsible landlords, such as the professional managers and locally based blue collar owners. This would also include the need to consider the broader market conditions in which landlords operate.

Mobility Theory
Mobility theorists point to a household's changing housing needs over time as the major force in neighborhood change. There are stages in a household's decision to move (Rossi 1980, 24): 1) dissatisfaction with the existing dwelling; 2) decision to move; 3) search for an alternative; 4) the actual move. There is a remarkable stability in mobility patterns over the years (Rossi 1980, 28-9;Rossi 1981, 148-71), although this can vary widely by neighborhood in a given city (Downs 1981, 27-33).
Some moves are forced through eviction, fires and similar causes, or induced by other factors such as divorce or new job, but about two-thirds are independent decisions (Rossi 1980, 34 ). Rossi (1980, 61) sums up the relationship between housing and mobility: The findings of this study indicate the major function of mobility to be the process by which families adjust their housing to the housing needs that are generated by shifts m family composition that accompany life cycle changes.
The stage in the life cycle plays a critical role (Rossi 1980, 58, 120-5).
Most moves take place in the first decade of a household's existence.
Large families are more likely to move than small, young families than older. Changes in household makeup, such as death, divorce, the addition or maturing of children trigger moves. Renters are more likely to move than owners, particularly renters who wanted to own.
Of the sources of dissatisfaction that lead to a decision to move, (Rossi 1980, 60, 131-5) housing related problems, particularly space and layout of the unit, are more significant than neighborhood related causes. Of these, the social makeup of the neighborhood 1s most important.
Neighborhood location, nearness to jobs, friends or relatives is of little importance (Rossi 1980, 58, 120-5). The more complaints a household has, the more likely it is to move.
Annual income has "practically no relation" (Rossi 1980 130, 202) to the probability of wishing to move in the face of the other factors. Only when households reach the last stage does cost affect the alternatives found in the search for new housing (1980,138,202).
The fit between household needs and the dwelling unit has important implications for neighborhood change. This precedes, although is constrained by, economic considerations. The propensity for young households to move makes the observation that gentrifiers are predominantly young less pertinent.
Changing household needs will affect the makeup of neighborhoods. In any neighborhood some are moving out for family, job or other reasons. The question is not who moves out, but what households move in (Myers 1983, 113-18). Age and racial profiles can help predict the future of a neighborhood. For example, a neighborhood with a significant number of young black families and of elderly whites, is likely to become predominantly black. A central neighborhood with a high proportion of retired people, may be replaced with in-moving professional people (Henig 1984, 179) which may be normal turnover rather than displacement.

A Composite Theory.
Other factors besides the adequacy and adaptability of the housing unit in meeting household needs affect housing decisions, and, in turn, neighborhood stability (Galster 1987, 4-26 Housing may be both an article of consumption and an investment; owners will act differently to maximize their well being, depending on which aspect is foremost (Galster 1987, 15-16). A time element is introduced because individuals or households are affected by their family plans or by expectations for the future of the neighborhood. An owner who intends to remain in a neighborhood pays less attention to the investment aspect of housing, while one who intends to move may make cosmetic improvements to enhance housing value.
Expectations for the future are subjective and may not reflect the actual state of affairs (Galster 1987, 228). There are two components of future expectations (Galster 1987, 125-142) An owner's intention to stay or to sell, to improve or to ignore a property, depends on what others in the neighborhood do; there is a circularity since everyone awaits the neighbors' actions, an example of the prisoner's dilemma (Galster 1987, 22). If no one acts, deterioration will take place. The degree of social interaction and neighborhood cohesion thus become important.
Social interaction and neighborhood identification can be measured by behavior, such as frequency of discourse, and by attitudes, such as considering neighbors as "friendly" or having common interests (Galster 1987, 87, 125). Attitudes prove to be more significant than specific action. Social cohesion is measured by a high degree of individual and aggregate social integration.
There is a complex set of relations between satisfaction with a dwelling unit and the degree of neighborhood cohesion (Galster 1987, 18-19, 112, 170-220). The stage in the life cycle is a critical component, with those in the earlier stages less likely to be satisfied with the dwelling, and more likely to move, but satisfaction with the neighborhood and neighborhood social cohesion significantly improve satisfaction.
Household characteristics most influence expectations of the future quality of the neighborhood (Galster 1988, I 34-42), but the actual present physical and demographic makeup play little role as does racial composition. Satisfaction with the housing and with the neighborhood are virtually unrelated to expectations of quality, a finding hard to explain (Galster 1987, 162). For expectations about future property value, the higher the degree of social cohesion the more optimism. In both types of expectations, the elderly are more pessimistic about the future, possibly because they are less informed.
The stage in the life cycle plays an important role in mobility (Galster 1987, 154-70), on satisfaction with the dwelling (Galster I I \ 1987, 112) and on the decision to invest in property maintenance (Galster 1987, 220-5 There are some less obvious reverse effects. Those who do not intend to move tend to neglect the exterior of buildings. The planned length of tenure affects the nature of the investment, rather than the size of the expenditure. Investment patterns differ significantly between those who perceive the neighborhood will change in quality and those who expect property values will change. Those who expect neighborhood quality to decline spend less, but if they expect property values to decline, they spend more, "bucking the trend" (Galster 1987, 224). On the other hand, if they hope to reap capital gains from rising prices, they spend less -they are "free riders". Galster's findings ( 1987, 235) suggest that filtering to lower income groups does not necessarily lead to deterioration, although he notes his study deals only with owner-occupants, not absentee owners, who are less affected by the key factor of neighborhood identification.
The process itself of filtering may make a difference, because higher income households that . remain may become pessimistic about the future of the neighborhoods and reduce upkeep (1987; 236). Pessimism becomes self-fulfilling.
Entry of higher income groups, gentrification, can have two possible, opposite effects (Galster 1987 ,238). Existing residents may be more apt to invest because they become more optimistic about the neighborhood, or they may reduce maintenance expecting to achieve capital gains. The intrusion of a new group may also disrupt existing cohesiveness, leading to pessimism about the neighborhood.
Galster does not discuss the possibility of both outcomes in the same neighborhood.
Galster, while not ignoring economic factors, puts his emphasis on individual decisions in the improvement or decay of neighborhoods.
He also puts great emphasis on neighborhood cohesion but warns this is not easy to create (1987,219,246) and can lead to parochial and exclusionary neighborhoods, such as the ethnic neighborhoods cited by Leven et. al. (1976, 110-13).

Economic
Interpretations. Kolodny (1978, 93-110) identifies the winners and losers in neighborhood change. There are two types of theory. Some posit inevitable cycles of decline, whereby successively lower income populations succeed upper mcome ones with the inevitable deterioration of the housing stock. This can be hastened by public policies, but the very theory of inevitable decline stifles remedial action.
The other type of theory sees decline, the filtering to lower income groups and ultimate dereliction, as caused by outside forces or events, many of which are beyond the control of local government, such as attraction to the suburbs or economic changes. Local government policy may either accept, stop or slow change. Each policy will benefit different actors: those who are forced to move involuntarily (the displaced), those who move voluntarily for personal reasons, those who would like to move in but are discouraged, those who are unwilling stayers, such as the elderly. In looking at alternative policy approaches, Kolodny finds each benefitted some actors but never all. He sees no equitable solution to stem neighborhood decline so long as economic disparities and racial discrimination exist.
This is also the conclusion of a number of theorists on gentrification, who look at the issue in terms of broad economic and political structures. Gentrification is a defensive economic action to protect investment in an uncertain economic climate (Beauregard 1986, 35). Because of the need for sound investment, and because of the need for conspicuous consumption, gentrifiers do not select just any low-cost areas. While many low-cost areas remain working class or "filter down", areas most likely to be gentrified are those close to the CBD, or with special attractions (Beauregard 1986:53). This investment decision is an expression of the structural changes in the city brought about by global economic changes.
According to this line of argument (Smith 1986, 1-7;Williams and Smith 1986, 208;Williams 1986, 208-222) the internationalization of capitalist production has changed the function and hierarchy of cities. Old industrial cities are declining, while new global cities are attracting the new financial element. In these cities, the more well to do are displacing the poor both by their demand for housing and by the expansion of commercial areas into former I I I residential areas. Many declining cities, however, will become attractive locations for investment in the period of economic crisis caused by global economic changes. This is because bargains may be found, and other investments are not safe. The degree to which this will happen and when it will happen, however, will depend on local investment conditions such as the cost of mortgage money, tax structures, subsidies and other government policies (Williams 1986, 214-16) This structural interpretation finds evidence of a polarization between classes, with managers, professionals and administrators becoming richer, while the old working class is increasingly impoverished with the decline of manufacturing. This leads to physical deterioration of old working class neighborhoods. Thus gentrification and abandonment are seen as the product of the same economic forces. Gentrification is the product of the market (see also Smith and LeFaivre 1984, 55;Marcuse 1986, 172), and the solution cannot be found in the existing market sy~tem. Poor and deteriorated neighborhoods can only be saved by reinvestment, but reinvestment leads to gentrification. The only solution (Williams and Smith 1986, 222) is the "decommodification of housing." The argument does not directly address the question why reinvestment will be in inner city neighborhoods; m many locations suburban ~ investment might well be a safer one.
The answer to the question of location for reinvestment 1s given by the "rent gap" theory (Smith and LeFaivre 1984, 49-54), by which the real value of city property is not represented by the market value. Neighborhoods become "prepared for gentrification" because there is a gap "between the ground rent actually capitalized with a given land use at a specific location and the ground rent which could potentially be appropriated under a higher and better land use at that location" (Smith and LeFaivre 1984,50). It is the state or financial institutions that make the new investment feasible.
Individuals come later. Thus where Galster stresses the role of the household in neighborhood change, Smith sees the household, and particularly the working class, as the victims of "those most able to control the real estate market" (1984,49; see also Smith 1986, 4-29).

Summary
In the mid 1970s, observers discovered the unexpected phenomenon of reinvestment in inner-city neighborhoods. A section of Providence's West End neighborhood was selected as a target area.

Background
Preservationists became more actively interested in neighborhoods as a result of the same forces that fostered the neighborhood conservation movement and that spurred much of the new attention to neighborhood theory. The post-World War II movement to the suburbs, particularly by the more affluent, and the increase of lower income groups in the cities were cause for considerable alarm as city tax bases shrank. Urban Renewal was seen as a way to bring reinvestment to the cities. However, the excesses of that program leveled many city areas without bringing 3 1 new development.
Massive high~ay programs also brought devastation to many neighborhoods. These programs gave nse to a I < ~ new politica 1 l force -the neighborhood organization which fought, often successfully, to curb urban renewal and highway construction, and to seek government support for its own interests. Many such organizations took the next step of developing their own housing, economic development, health and social programs to address the needs of inner city neighborhoods ( Baroni 1983, 177-88). The Model Cities Program sought to harness this energy with requirements for neighborhood participation in Community Action Programs. The Carter administration established a neighborhood conservation section in the Department of Housing and Urban Development, and carried out the Neighborhood Strategy Area program (Rosenthal 1988). Neighborhood-based planning became fashionable.
Urbari Renewal and highway expansion also activated the embryonic historic preservation movement (Kay 1986, 50), and because historic buildings were likely to be in inner city areas that had suffered population changes, preservationists were faced with problems of deteriorated housing in low or moderate income areas.
Preservation efforts take a number of forms. The National Register of Historic Places _, establishes criteria for historic buildings that are eligible for federal assistance, and many localities have established registers with somewhat difference criteria. Significant buildings may be listed separately on these registers, or individual buildings, less significant in themselves, may together form an historic district which reflects a particular past way of life or 32 settlement patters.
Protection of such districts may become neighborhood conservation.
Preservation efforts range from simply stabilizing a building to prevent further deterioration, to preserving the original structure, or carefully replacing lost parts which are undertaken for museum quality . Selection of th~strategy is determined by the importance of the building, intended uses, and also by financial considerations.

Formation of the Revolving Fund
In the late 1970s, the Providence Preservation Society, which had previously worked only on Providence's East Side, began to consider the rich collection of historic buildings on the west and south sides of the city. These buildings were the heritage of In the 1970s, Providence was losing population It stood at about 150,000, down from its largest population of 250,000 at the beginning of this century. The economic condition of the city was critical.
Lower income neighborhoods showed increasing signs of deterioration. It soon became evident the Preservation Society could be more effective if it had funds of its own, providing an ability to act rapidly in the real estate market in cases in which a significant building was threatened.
In deciding to establish its own Revolving Fund, the Preservation Society studied the operations of dozens of other similar efforts, but focused its attention particularly on four . 1 Leaders of the four were invited to address preservation audiences and hold workshops in Providence.
In Savannah (Gratz 1989, 33-67;Rosenthal 1988, 117-20) where Historic Savannah had been working 1 Records of the initiation of the Revolving Fund may be found in the uncatalogued archives of the Providence Preservation Society in the library of the Rhode Island Historical Society. for a decade to salvage the decaying early 18th century inner city, a banker, Leopold Adler III, undertook a new effort, the Savannah Landmarks Rehabilitation Project, to save the Victorian section of the city. This section had become home to a poor and black population.
By skillful use of Section 8 housing and CETA job training funds, Adler was able to restore the buildings as homes for the existing population.
In Pittsburgh, Arthur Ziegler (Gratz 1989, 73-80), after first spearheading restoration of a working class district, persuaded the city's redevelopment agency to help in a lower class neighborhood by purchasing facade easements on historic buildings which would then provide funds for owners to repair their houses. In Cincinnati, a black neighborhood leader, Carl Westmoreland (Gratz 1989, 70-3;Westmoreland 1972Westmoreland , 1976, took over a Victorian section of landlord-owned houses and converted them to neighborhood owned houses for low income residents who were predominantly black. The Hartford Architectural Conservancy had also undertaken preservation for lower income residents, although on a smaller scale.
Pittsburgh and Hartford, relying less of federal funds, used revolving funds. These operated on the philosophy that the fund should strengthen a neighborhood by conservrng its housing stock and find ways to use problem buildings that were considered uneconomic by their owners. A revolving fund might therefore lose money on a given project, but the investment would be repaid by overall improvement in the neighborhood.
Moneys repaid would be recycled.
The tax advantages of donations to such non-profits could help in fund raising. The main differences between a Revolving Fund loan and other city programs is that the loan is repaid into the Fund for new loans.
The Fund requires all main structural problems to be addressed. No paint is slapped on until leaking roofs and gutters are repaired, foundations stabilized, and so forth. If an owner cannot afford to do all repairs at once, a series of loans are made over a period of time.
Those not wishing to make such commitments are referred to other city programs. Most recipients put as much or more of other funds, and private or bank loans, into the project.
The Revolving Fund also differs from city programs m providing technical assistance to analyze structural problems, to recommend remedial action, to find qualified workmen and to oversee the work. All of this can reduce the cost of rehabilitation.
The Revolving Fund also has some similarities with the  (RIHPC 1979, 5-9). This southern section of Elmwood, originally in Cranston, was ceded to Providence in 1868.
Cooke's development forms the core of the Adelaide Avenue National Register Historic District. In the West End, a merchant and philanthropist, Ebenezer Knight Dexter, left ten acres to the city for a military drill field (RIHPC 1976,9). This is the Dexter Training Ground in the present day Armory District.
Development of public transportation and increasing city population spurred settlement of these neighborhoods . Horse-drawn omnibuses introduced on the main thoroughfares in the 1850s were supplanted by horse-drawn rail cars in the 1860s, and later by trolleys. In the twentieth century, the automobile gave access to those parts of the neighborhoods that were less accessible from public transport.
After the Civil War, the city's population developed rapidly, growing from 69,000 in 1870 to 225,000 in 1916. The increase was swelled by many immigrants . By 1910, seven in ten residents were immigrants (RIHPC 1979;7-8;Woodward and Sanderson 1986;38-9).
As the city grew in population and in economic strength, new industrial districts appeared. Olneyville, an early mill settlement just west of the Armory District, expanded in the nineteenth century.
During the course of the nineteenth century, a new industrial center appeared around a former pond at the junction of present day Bucklin and Dexter Streets in the West End, and firms were scattered throughout the southern part of the city. Many workers and managers in these businesses found it convenient to live within walking distance (RIHPC 1979, 7-8, 29;Woodward and Sanderson 1986, 38 Gentrifiers are typically characterized by small households, predominantly young with incomes above the national median, and, by definition, with incomes above those of the neighborhood's existing population.
In the process of renovating houses and improving the neighborhood, gentrifiers cause prices, rents, and taxes to rise, which leads to the displacement of existing residents.
Certain groups are considered at particular risk of displacement. To establish whether gentrification has taken place in the Amory District, it is necessary first to determine the income of existing residents (Gale 1979, 293-98), the size of groups at risk (Legates and Hartman 1986, 184-98), and the price and condition of housing (Clay 1983, 22-70). The black population is taken as an indicator of those  (Figure 4-3). Median income was also roughly the same (Figure 4-1) Many theorists (Lipton 1980, 48,52-4;Hodge 1980, 193;Downs 1981,40,79;Nelson, 1988, 135)  -Population change. Differences in population changes by tract provides some insight into the relative attractiveness of the tract.
-Housing tenure. Many consider resident owners contribute to stability in a neighborhood. These may be homeowners or landlords who are neighborhood residents (Goetz and Colton 1979, 80-5;Galster 1987 ,.19, 294 ). Renters are more at risk of displacement as neighborhoods gentrify.
-Condition and price of housing. This is both an indicator of the socio-economic makeup of a tract and a basis to assess the effectiveness of the historic preservation program, which is concerned with improving building conditions.
Four of these, building value and condition, owner occupancy, and race are used by Galster (1987, 80-1 ), and each is cited by one or more theorists on gentrification (Zeitz 1979,39-55;Nelson 1988, ch.5;Gale 1980;Downs 1981, 25, 84-98;Marcuse 1986, 166-70). These indicators are used to compare the four census tracts to each other and/or to city or national norms to establish the relative standing of the tracts over time. Data on such indicators as housing prices and rents in 1990 are also used to compare the tracts, but city-wide data for 1990 are not currently available.

From 1950 to 1980
The legacy of the early history recounted above may be found

Income distribution
Median incomes can hide changes in a tract. A bimodel distribution, with some high incomes and many low incomes can be an indication of change in a neighborhood (Leven et. al. 1979, 137;Lipton 1980,50).  In the three decades, there was thus a dispersion of much of the black population from the East Side to the south side changing dramatically the racial make-up of Tracts 2, 3 and 14. Tract 13 was an anomaly with its small increase.

Total Population Shifts.
The growth of the black population and its dispersion through the city took place against a backdrop of decline in total city population which also was uneven (   1980 1950-80 1970-80 1950-80 - and urban renewal projects may account for some of these changes .

Owner-Occupancy.
The owner occupancy rate declined in all the tracts, as it did for the city as a whole (Figure 4-3 (Table 4-2).
They can provide only a rough picture of the relative standing of the four tracts. The 1950 census measured condition by the designation "no private bath or dilapidated" and "no running water or dilapidated". These are combined in Table 4-2. In 1960, the categories of "sound", "deteriorating," and "dilapidated" were introduced, with further subdivisions in each category concerning the presence or absence of plumbing facilities. These categories are "not entirely equivalent" to the 1950 categories (Census 1962, 6).     The emphasis was to be on rehabilitation, but some lots were slated for acquisition, and the plan was to be updated periodically.
The plan also rezoned much of the area. The whole of Tract 13 was rezoned R-4, multiple residential, but excluded apartment houses, hotels, fraternities and rooming houses. There were a number of rooming houses in the area (Polk 1980 The great concentration of multifamily buildings were in the area west of Cranston Street. The plan proposed rehabilitation, with a loan program, but also listed buildings for acquisition. It was later amended to acquire about 18 abandoned structures, with hopes of reselling them for rehabilitation. In 1985, a group of consultants (Stull and Lee et. al. 1985) undertook a new study of the same area, although no formal plan followed. They described the area's housing as "predominantly working class". They noted the widespread existence of vacant land, about 12 percent of the "West Elmwood" area, resulting from the demolition of abandoned buildings, and more buildings stood abandoned.
The area was losing population, although there was an increase rn the number of families with young children since the last census. In 1980, four percent of the units were condominiums, and the vacancy rate was 16.5 percent, of which some 45 percent were abandoned or "out of useful production" (1985, Section D). There was a large number of female headed households, many with children.
Only 57 percent of the males were employed or looking for work.
Seventy percent of workers were blue collar.
Another indication of housing conditions in the neighborhood 1s the number of demolitions ( although not to the extent of Tracts 3 and 14. Abandonment was occurring, which also brings displacement.
All the indices of housing condition and the fact that urban renewal projects were undertaken in all but Tract 2, suggest a decline in the housing conditions, although this is difficult to quantify precisely.

Armory District in Tract 13
This discussion has referred to Census Tract 13 as a whole, but the Armory District represents only the eastern half of that tract.
Block statistics give some indication of the differences between the two halves. They are roughly equivalent in total population and number of units (Table 4-4 ). While the total black population is small, as noted above, there are proportionately twice as many blacks in the Armory District as in the western section, 7 versus 3.4 percent respectively.
Similarly there are more hispanics, 11.6 versus 8.6 percent. The rate of owner-occupancy 1s slightly higher in the western segment than in the Armory, 22 and 20 percent of the units respectively.
In this western section there are more addresses with only one unit. Except for Marvin Gardens Elderly Housing, there is only one structure in the western section with ten or more units, while there are 51 m the Armory District. In other words, the buildings in the Armory District tend to be larger in scale and older than those in the western sector. Some originally sizeable singI°e family houses had been broken · up into apartments or rooming houses. Many of the buildings were originally built as two-and three-family houses and may have been further divided.

Summary
In the three decades before 1980, the four census tracts shared the declining fortunes of the city to a disproportionate extent.
Median incomes in 1950, above that of the city and even the nation in some tracts, all fell below that of the city by 1980. Tracts 13 and 14 lost a larger proportion of their population than the city as a whole, although Tract 2 was relatively unaffected by population loss. The   1989. This is shown as a ratio to the city-wide median. The latter was $11,452 in 1980, and $19,623 in 1989  upper three income categories, and a distribution close to that of the city as a whole. For 1989, the distribution in Tract 13 remains close to that of Tracts 3 and 14, while Tract 2 has close to the same proportion of lowest income persons, but a larger portion in the upper two income groups to account for the higher median.
Occupations: The Providence Directories (Polk 1980(Polk , 1990 offer some insight into the occupational distribution of the tracts, but the sample is too small to provide a definitive comparison. A random spatial sample (Ebdon 1977)  This section unfortunately provides data on only a third    10 to 599. The large increases in these groups have changed the racial and linguistic makeup of all the neighborhoods, as shown in Figures 5-7 ,5-8. In the preliminary Census data, the count for all racial groups is less than the total population for the Tract. This is because of incompletely processed and suppressed data. The gap has been assigned to "other", which also includes Native Americans.
In 1990, almost half of Tract 13 is minority. Whites are in the minority in the other three tracts. These results may be read as greater stability in Tract 13, or as greater resistance to an influx of new groups, but they do not indicate displacement of existing minorities

Owner Occupancy
The Providence Directories (Polk 1980;1990) were used to establish the proportion of owner-occupiers ( Figure 5-10 Source: Polk 1980, 1990 comparison of the number of owners reported in the 1980 Directory with similar figures from the 1980 Census (Table H-1), however, raises some questions (Table 5-2).   were most severely affected, but Tract 13 also lost seven percent of its units. Until block data are available, it is not clear whether these were in the Armory District or the West Side . It is also possible that some of the loss was due to combining smaller units into larger, but data to test this hypothesis are not available . The addresses were assigned to the tracts based on Directory listings.
As the small size of the Armory District produced a small sample of sales for 1980 and 1990, listings for the four tracts for 1979 and 1989 were added to increase the sample siz~.
To adjust for different building sizes, the selling prices are divided by the number of rooms, although this does not adjust for the size of the rooms or the condition of the building, which will affect sale pnce. The mean price-per-room by building type for 1979-80 is shown in Table 5-5. In general, the price-per-room declines as the number of units increases, for both years. The average room price for single family houses m Tract 13 is below Tract 2 and above Tracts 3 and 14. It has the lowest price of all the tracts for two-unit buildings, but the small sample leaves a question here. The larger samples for three-units is a better indicator , with Tract 13 again below 2 and above 13 and 14. The prices for the 1989-90 period are in Table 5-6. In 1989-90, Tract 13 had a mean price per room for single family houses well above Tract 2. This category is skewed by the fact that one of the five sales was for $259,000. This house is a 13 room, architect-designed mansion on a large lot facing Dexter Training Ground, a unique building in the Armory District for its size, Tract 2 $90,883 Tract 3 $62,500 Tract 13 $80,500 Source: MLS 1989. Numbers in parenthesis are the sample size.
Tract 14 $71,416 Ren ts: Classified advertisements in the Providence Journal for unfurnished apartments were used to estimate rents. As is true for the MLS listings, ads do not reflect private rentals. Rentals to friends or relatives, for example, are often special arrangements. The contract rent may differ from the advertised rent. Ads that were listed as "Armory District" or where location could otherwise be identified were used. However, unlike the MLS listings, where addresses were always given, real estate boosterism may lead landlords to list as "Armory" addresses that are well removed from the district designated by the Providence Preservation Society.
As actual addresses are seldom given, the Armory, Elmwood, Federal Hill and East Side areas were used as controls instead of the Census Tracts . The designation Elmwood would include Tract 2 and part of Tract 3. The West End, Tract 14, is seldom cited. Federal Hill, north of the Armory District, with similar housing, is used instead .
The East Side, the most affluent part of town, was used to indicate what rents might be sought in an area that was gentrifying.
Ads for the four area were drawn from one Sunday in each month from January to April. To adjust for its small size, additional ads for the Armory District were taken from three other Sundays, a week after the larger sample. Phone numbers were checked to avoid duplication of the Armory ads on these subsequent weeks. Only ads that listed the number of bedrooms were used, which was usually the case.
Sometimes utilities are included. Where they are not, it is assumed that they are not included. Winsor Associates (1990, 2- The average and median rents for the four areas are shown in Table   5-8, and 5-9 for one, two and three bedroom apartments. In most categories rents in the Armory District are slightly higher than for Federal Hill and somewhat more above Elmwood, but well below that of the East Side.
Winsor Associates (1990,Ch.2) undertook a detailed study of the city rental market by neighborhood in the summer of 1989. The results (  provide a larger sample by which to assess the program, while separating the Armory and non-Armory categories. Characteristics of the clients are summarized in Table 5-11. Incomes were determined at the time of the loan on the basis of tax returns, and have not been adjusted for inflation. In 1989, the median household income 2 for the Tract 13 was $14,345 (CACI,1990) and the median family income was $18,234. A very high proportion lived in their own buildings and aJI lived in the neighborhood Resident owners are more responsive to neighborhood norms and a stabilizing influnce (Goetz and Colton 1979, 80-5, 91;Galster 1988, 19). Renters move more often . Age was given at the time of the loan. A great majority of recipients in the Armory District were in their 20s and 30s and were single, married couples, or couples with few children. Hence recipients match the description of gentrifiers (Gale, 1979, 293-8).
This is also the stage of life people are most likely to move (Rossi 1980, 58, 123, 125 ), which puts in question this characteristic of In the other areas, which have more resident minorities, 30 percent of the recipients were minority.
The occupational makeup of the recipients (Table 5-10) includes working spouses since they were also residents of Tract 13.
Retired persons and students are not included in the percentages, and the occupations of a few were not known. Several were unemployed and were categorized according to their last job. Most of these buildings were multi-unit, and rehabilitation added to the rental stock. Projects using Community Development funds were required to rent to income-eligible tenants, which would have enabled existing tenants to remain.
In the Armory District only six buildings were single-family.
In  (1983), or rehabilitation by other newcomers, was undertaken with other funds. The nature of this housing stock means that most buyers in the Armory District must also become landlords. Since much gentrification takes place m single-family houses (Black 1980, 9) the predominance of multi-unit buildings may be an inhibiting factor.
The average loan was $13,122 in the Armory District, but most were for $10,000. The total was $669,000. The average loan in the extended area was somewhat larger, $21,999, for a total of $518,319.
These sums were matched by the recipients. Interest rates were geared to the borrowers ability to pay.

Summary
The data leave some gaps in tracing changes during 1980s.
Preliminary 1990 census data and substitute data leave some The rent data suggest rents rn the Armory are comparable to west/south side rents, but not those of the East Side. With averages and means only somewhat above those of Federal Hill and Elmwood, the differences may be accounted for by a few higher priced units.
Housing prices also do not show a large increase in Tract 13. In 1989-90, the tract stands roughly in the same relationship to Tracts 2, 3 and 13 as it did in 1979-80, and the Armory District is no more expensive than the western section, although the sample is too small to make a final verdict.
Housing conditions are difficult to document. The Directories (Polk 1980(Polk , 1990  The neighborhood may have fitted descriptions (Leven et.al.. 1976, 110-13;Galster's (1987, 219, 246) of a cohesive ethnic neighborhood that would also be exclusive. Rental units would be held for family members. The Armory District was a small and relatively clearly defined area, adjacent to the larger Italian community of Federal Hill, and all part of the same parish.
By the 1960s and 1970s, long-term residents had paid off their mortgages and, as "empty nesters", rented to reliable and long-term tenants at low favorable rates, sufficient to cover insurance, taxes, utilities.
One such long-term resident reported that even in 1990 such rents could be as low as $100-150 a month. This could be one explanation for the lower median rents in this census tract. Any new owner needing to cover mortgage costs obviously could not continue such rents. One of the early Armory settlers under the Revolving Fund program described the area when she came as "a sleepy neighborhood": a building with two apartments would have one old lady in each.
The community changed. The children moved out in the wake of federal housing and highway programs which favored the suburbs (Checkoway 1986, 119-36). The largest losses of population were in the 1950s and the 1970s (  (Polk 1980(Polk , 1990. "Slumlords", as defined by Goetz and Colton (1979, 80-8), were buying in the early-to-mid-1970s, an indication the market would not support other owner types. The median value of units was dropping relative to the city throughout the 1960-80 period.
As the city population declined, presumably demand dropped. A number of absentee owners abandoned their buildings.
The description of a "sleepy neighborhood", the number of older long-term tenants, and the degree of abandonment suggest a vacuum in this housing market at the time the Revolving Fund began. The area was "not that bad" but little or no new investment was taking place.
Existing residents saw the pilot project of the Revolving Fund at 103 Parade Street, a highly visible Victorian house, which had become a burned out rooming house, as distinctly catalytic. It was a "psychological boost", "The young people saved the neighborhood," two informants reported. The young people were ready to take risks and also worked together and with older residents for a variety of city improvements.
Who were the "young people"? Many of the handful of pioneers came from out-of-state, and "didn't have the negative attitudes of Rhode Islanders. They appreciated the houses, which, like so much Providence housing, were real bargains at the time." Although these people belonged to the baby boom generation, they were not so much the young urban professionals mentioned in the gentrification literature as seekers of the "moral alternative," suggesting a community-oriented and art-centered way of life. They found a quiet, racially and ethnically mixed neighborhood. The large Dexter Training Ground, dominated in those days by a men's bocca club, provided peace, quiet, space and greenery. Many of the newcomers were attracted to Providence by the colleges and universities, a phenomenon noted in other cities (Nelson, 1988, 157;Gale, 1979,295-7). They were graduates, students, or college applicants. For many buying and fixing up a house with sweat equity was both an investment and a way to obtain housing. The two-to three-unit housing provided a place to live and some rental income. Although the units were not rented to family members, . the housing solution was similar to the three-decker pattern of so many immigrant groups.
After the Armory program started, sale pnces remained about the same as before for several years, but began to rise about 1985, IOI according to the long-term landlord. This was also the time that prices across the city were rising, on the west side far more than on the East Side (Winsor 1990, 1-3 to 1-7). It is hard to determine therefore how much the price rises in the Armory were the result of Revolving Fund activity, and how much resulted;ted from the citywide boom. The boom was already slackening by 1988-9, and as of 1991, the market was very slack. People were "discouraged with Rhode Island" and many of the Revolving Fund recipients were "hanging tight," the landlord reported. Still, he added, it is "a great area".
Some of the Revolving Fund newcomers remained and are currently raising families. Some moved to take new jobs elsewhere.
Some found the typical two-or three-bedroom apartments cramped as families grew. The schools were considered a real problem. This is a community issue the neighborhood organization has not addressed . in a systematic way, and perhaps lacks sufficient numbers to deal with effectively.
What happened to the older long-term owners? Many sold their houses as prices rose, particularly during the housing boom.
Some are still there. "The only ones gentrified were those who died," quipped one long term resident.
What of tenants? There are still some who enJoy the favorable private rents. While the norm of rents advertised ( Several commentators (Levy and Cybriwsky 1980;Spain and Laska 1984,123;Zeitz 1979,74;Palen and London 1984,10) speak of conflicts in goals and values between new and existing residents.
While the Revolving Fund recipients in the Armory District are probably better educated than the remaining older residents (data on educational background were not available in detail), the craft orientation of many of the newcomers seems to have blended well.
One informant said they were better educated "in what they know".
Conflict appears to revolve around personalities more than issues.
Like other neighborhoods (Rossi 1980, 36-7, 120;Downs 1981,28), there has been considerable flux of tenants in the Armory District. One Revolving Fund recipient, who feels well established in the neighborhood after nearly a decade, raised the question of who was new and who was old and how long a newcomer is "new". Residents also observed considerable shifts in the minority population over the decade. In the early 1980s, the Asian population appeared to grow rapidly, but declined since 1985. On the other hand, the number of blacks seemed to drop in the early part of the decade, and now is seen as increasing.
These are both trends that will not show in the final census counts .

Introduction
The main findings of this study are summarized and general conclusions presented below. These include assessment of the Revolving Fund program and its limitations. Recommendations follow.

Summary of Findings
Observers who drew attention to the issue of gentrification, identified gentrifiers as predominantly young, in small households and with median incomes close to, or above, the national median, and by definition above that of the neighborhood in which they settle.
They are disproportionately in managerial and professional Some (Goetz and Colton 1979,80-5;Galster 1987, 19) maintain that resident owners are more responsible and increase neighborhood stability. Owner-occupancy is a measures of one aspect of this but does not include neighborhood-based landlords.
Owner-occupancy of units was highest in Tract 2 in the 1950-1980 period and Tract 13 stood next (Figure 4-4 ), with rates in the earlier decades well above that of the city. The rate was lowest in Tract 3, and declined in all tracts over the period.
Since many buildings in all the tracts have two or more units, an owner-occupied unit implies ownership of the whole building but precise figures for the percentage of owner-occupied buildings are elusive. Directory (Polk 1980(Polk ,1990  More recent data on rents rest on advertisements and cannot be compared to census data for 1950-80. City-wide medians are not available, and cases cannot be assigned to census tracts. Median and mean rents in the Armory District (Tables 3-8 All these indicators show the four tracts declining relative to the city, and the diminution of the . differences between them that appear in the earlier years. There are no indications of an influx of higher income groups into Tract 13 in the 1980s. Still the tract data may hide counter trends within the Armory District.

The Armory District
In 1980, the Armory District was comparable to the western side of the tract, with a slightly larger proportion of the tract's few minority households. In the Armory District there are fewer singlefamily houses and more two-to four-unit buildings than in the western section. The housing stock is also older.
Data are available on Revolving Loan recipients (Table 5-11), but not on other in-movers. The median income of the former, existing residents and newcomers, is below or slightly above the median for the entire tract, depending on whether family or household medians are used. As the median for the tract as a whole fell relative to the city, there appears not to have been an influx of higher income households in the wake of the Revolving Fund program sufficient to alter the overall figures.
The age and household size of recipients fit into the pattern of gentrifiers, but as households are most likely to move in the early stages of the life cycle, this characteristic should be discounted.
The The 1989-90 sale prices of houses in Tract 13 (Table 5-6) seem fairly typical of the tract as a whole Of the few house sales in the western section of Tract 13, many are above the median for the house type, suggesting that not all the higher priced buildings were m the Armory District.
Evidence on displacement is sparse. While by anecdotal reports one or two elderly persons were displaced, the requirement that half the residents of a building receiving Community Development funds meet income guidelines would permit lower income tenants to remam. Furthermore, 39 percent of the rehabilitated units were m abandoned or unoccupied buildings, where no one was displaced. The large in-migration of minorities certainly does not indicate displacement. Data are not available on the distribution of the minorities in the two halves of Tract 13, but informants report noticeable increases in the Armory District.

Conclusions
There 1s little evidence of gentrification in the Armory District m terms of the influx of higher income households or the displacement of existing residents. In 1980 it was a "sleepy neighborhood". The flight to the suburbs of the upwardly mobile members of the ethnic population and subsequent tearing of the social ties centered on the parish created a slack housing market.
There were "real bargains".
The pattern of events does to some extent fit the early stages of the gentrification process described by Pattison ( 1983) and Gale (1980, 103-112). The first comers came seeking affordable places to live. Many are artists and those interested in a more urban life style.
As craftsmen and artists they were also attracted by historic buildings. As banks were reluctant to lend in the early 1980s, the Revolving Fund fulfilled this function, together with private savings and "sweat equity". In the second stage, two of the newcomers joined with an experienced preservationist to form the Armory Revival Company to rehabilitate 23 existing units, and build 29 new units, which the Rhode Island Housing and Mortgage Finance Corporation financed. These met the agency's price guidelines for low-and moderate-income households. There has been some marketing of the neighborhood as a result. There is no evidence of the class tensions (Gale 1980, 106) that arise in the second stage.
Evidence on displacement 1s lacking.
The next two stages m Pattison's and Gale's model, leading to a complete turn around, have not taken place. A number of factors may impede this outcome. In Nelson's study (1988) Providence was one of the cities in which the rate of out-migration of higher income whites declined in the late 1970s, but did not cease. Providence has continued to lose its white population, while attracting more minorities (Census 1990). Spain (1980, 28) considered change m the racial makeup of cities an indirect measure of housing change, smce whites tend to have higher incomes than blacks. According to the 1980 Census (Tables P-13, P-15, P-17, P-19), in 1979 the median income of white households in Providence was $12,000; of black households, $9,067; of Asians, $9,537; and of hispanics, $8,892. It 1s unlikely that these discrepancies have declined.
While Providence is a small financial center, it is currently experiencing the effects of a nationwide recession, and the future 1s unclear. The demographic changes together with the weak local economy would not suggest a large increase in the market for higher priced housing.
The houses in the Armory District are predominantly two or four units. While these have proved fine starter houses for newcomers, and many families remain today, the units are not always appropriate for growing families. Black (1980, 3-12) found most gentrifying neighborhoods were single family, and Clay (1983,30) found they were less likely to be in areas with multi-unit buildings.
In London, Machielse (1987, 62) found gentrifying areas tended to be close to existing high status neighborhoods, and Nelson (1988, 97) found that tracts adjacent to Brown, that is on the East Side, showed a rise in median incomes. The Armory District is isolated from the prestigious East Side, and while many newcomers welcomed this distance, it may prove to be a deterrent in the future.
Finally, it is difficult to assess the future impact of the rapid increase in the minority populations in Tract 13 and neighboring tracts. Cultural conflict might precipitate out-moving of long term residents and Revolving Fund recipients, if they perceive the new groups as contributing to crime or exhibiting "unacceptable standards of behavior" (Leven 1976, 88, 144). With language barriers, social cohesion (Galster 1987, 20, 87, 142, 223) may be disrupted, leading to diminished expectations for the future quality of the neighborhood.
Behavior, often related to socio-economic background rather than to race or ethnicity per se, appears to be the significant factor in triggering out-moving. Fears about decline in housing values also play a role (Galster 1987, 142-3;Leven et.al. 1976, 134-6)

Contribution of the Revolving Fund
The -It concentrated on an area small enough to make visible progress, which encouraged existing residents to invest.
-It tackled "problem" houses, either those so damaged or so entangled in legal or bureaucratic red tape that individual owners were daunted.
-It adjusted loans to the owner's ability to pay. This involved adjustment of interest and phasing of loans.
-It worked closely with owners on the best construction techniques and concentrated on basic structural soundness.
-Together with neighborhood organizations, it looked at the neighborhood as a whole to address other community problems.

Limitations of the Program
A small area, such as the Armory District, is nevertheless subject to larger social and economic forces (Smith and LeFaivre 1984;Smith and Williams 1986;Smith 1986). Local conditions and demographic changes (Leven et.al.. 1976, Ch.7;Goetz and Colton 1979, 80-8;Gale 1979, 293) also play a role, as do the needs of particular households (Rossi 1980, 61;Galster 1988). In the past, federal housing programs have helped mitigate the impact of these forces on lower income households, but in the last decade (Nenno 1987), these programs have been poorly funded. Neighborhood programs, like the Providence Preservation Society Revolving Fund, can play only a marginal role in the face of such large forces. Like local governments (Yeates and Garner 1980), they are limited in their ability to deal with forces that currently dominate the housing market.
The larger economic trends have led to greater disparities of wealth and also to periodic recessions, which affect jobs and hence families' ability to pay for housing. The median incomes (Figure 2-1) of the region and the city relative to the national median show the relative decline of the Providence SMSA (see Harrison 1984). With numerous blue collar and low-paid service personnel, the four tracts under consideration declined relative to the city as a whole. These areas were also affected by the federal highway programs and housing programs (Checkaway 1986) which drew many residents to the suburbs.

Recommendations for Further Research
The completed 1990 census will permit a better comparison of the indicators used between tracts and over time since the data will be compatible. Block data will also provide a better comparison of the two sides of Tract 13. It would also be useful to have a more detailed tracking of house prices during the period of the 1980s to assess the impact of the housing boom and subsequent recession.
More needs to be known about the hispanic and Asian populations; their incomes, occupations, and tenure status. Language presents a barrier to interviews, but local welfare and ethnic organizations could provide data.
It would also be useful to compare the experience of the Armory District that of similar programs in other cities, both slow growth cities like Providence, and others, such as the much-studied 1 1 6 Washington D.C., to determine more precisely the conditions which lead to gentrification. Many gentrification studies are a decade or more old.

Recommendations
If the goal of the Revolving Fund is to stabilize a neighborhood by preserving the historic buildings, it must be concerned with demographic changes. Rapid change is particularly likely to have adverse effects (Galster 1987, 236-8) as residents lose confidence.
Monitoring change and slowing it would therefore be desirable. This may mean adapting the preservation strategy to accommodate lower mcome groups. Needs of the new population should be studied . Such a strategy would mean outreach efforts to the new constituency, and possibly developing alternate forms of tenure, such as cooperatives .
There are problems, however. Adler's efforts in Savannah to preserve buildings for low income tenants used large infusions of federal housing and other program funds. These are currently m short supply. At the same time efforts should be made to retain existing owners, both long-term residents and revolving fund recipients.
As a small organization, the Revolving Fund can better work as part of a city-wide and city-led coalition addressing housing issues.
The city's Human Relations Commission could undertake efforts to mitigate cultural conflicts. The city is aware of the need to hold middle and upper income residents to maintain the tax base. To this end, prime problems that need to be addressed are education and crime, or the perception of crime. to an improved quality of life in one small neighborhood, although not equipped to deal with larger social and demographic changes which may face the neighborhood in future. It is important to study such efforts to find methods to improve neighborhoods for all income groups.