diff --git a/changes/77.changes b/changes/77.changes
new file mode 100644
index 0000000..17cd7b7
--- /dev/null
+++ b/changes/77.changes
@@ -0,0 +1,2 @@
+Refactor handling on empty values for multivalued fields.
+
diff --git a/search/management/commands/import_data_csv.py b/search/management/commands/import_data_csv.py
index d13a984..5db431c 100644
--- a/search/management/commands/import_data_csv.py
+++ b/search/management/commands/import_data_csv.py
@@ -69,22 +69,41 @@ def set_empty_fields(self, solr_record: dict):
 
     def set_empty_field(self, solr_record: dict, sf: Field):
         # look for empty fields, empty multi-value fields, and empty multi-value codes and set them
-        if (sf.field_id not in solr_record and sf != 'default_fmt') or (solr_record[sf.field_id] == '') or \
-            (isinstance(solr_record[sf.field_id], list) and len(solr_record[sf.field_id]) < 1) or \
-            (isinstance(solr_record[sf.field_id], list) and solr_record[sf.field_id][0] == ''):
-            if sf.default_export_value:
-                default_fmt = sf.default_export_value.split('|')
-                if default_fmt[0] in ['str', 'date']:
-                    solr_record[sf.field_id] = str(default_fmt[1])
-                elif default_fmt[0] == 'int':
-                    solr_record[sf.field_id] = int(default_fmt[1])
-                elif default_fmt[0] == 'float':
-                    solr_record[sf.field_id] = float(default_fmt[1])
-                else:
-                    solr_record[sf.field_id] = ''
-            if sf.solr_field_is_coded:
-                solr_record[f"{sf.field_id}_en"] = "-"
-                solr_record[f"{sf.field_id}_fr"] = "-"
+        val = solr_record.get(sf.field_id)
+        is_empty = (
+            (sf.field_id not in solr_record and sf.field_id != 'default_fmt') or
+            val is None or val == "" or
+            (isinstance(val, list) and (len(val) == 0 or all(str(i).strip() == '' for i in val)))
+        )
+
+        if is_empty:
+            fallback_value = ''
+
+            if sf.default_export_value and '|' in sf.default_export_value:
+                try:
+                    data_type, default_value = sf.default_export_value.split('|', 1)
+
+                    if data_type in ['str', 'date']:
+                        fallback_value = default_value
+                    elif data_type == 'int':
+                        fallback_value = int(default_value)
+                    elif data_type == 'float':
+                        fallback_value = float(default_value)
+                except (ValueError, TypeError):
+                    fallback_value = ''
+
+            # check for multivalued field
+            is_multivalued = getattr(sf, 'solr_field_multivalued', False)
+            final_value = [fallback_value] if is_multivalued else fallback_value
+
+            solr_record[sf.field_id] = final_value
+
+            # check for localized fields
+            is_bilingual = getattr(sf, 'solr_field_lang', '') == 'bi'
+
+            if sf.solr_field_is_coded or is_bilingual:
+                solr_record[f"{sf.field_id}_en"] = final_value
+                solr_record[f"{sf.field_id}_fr"] = final_value
 
     def handle(self, *args, **options):
 
@@ -383,7 +402,7 @@ def handle(self, *args, **options):
                                                 solr_record[csv_field + '_en'] = "-"
                                                 solr_record[csv_field + '_fr'] = "-"
 
-                                                # Ensure all empty CSV fields are set to appropriate or default values
+                            # Ensure all empty CSV fields are set to appropriate or default values
 
                             solr_record = self.set_empty_fields(solr_record)
 
diff --git a/search/tasks.py b/search/tasks.py
index 5189027..bb427cf 100644
--- a/search/tasks.py
+++ b/search/tasks.py
@@ -21,10 +21,14 @@ def cache_search_results_file(cached_filename: str, sr: SolrResponse, rows=0, fi
       
         with open(f"{cached_filename}.tmp", 'w', newline='', encoding='utf8') as csv_file:
             cache_writer = csv.writer(csv_file, dialect='excel', quoting=csv.QUOTE_NONNUMERIC)
-            #### the values from the dict docs[0] can be used to find the column headers
-            ## in the field_list dict.
+
+            headers = []
+            for doc in sr.docs:
+                for field in doc:
+                    if field not in headers:
+                        headers.append(field)
+
             label_header = []
-            headers = list(sr.docs[0])
             for colname in headers:
                 colname = colname.strip(" _,").lower()
                 if colname.lower() == 'id':
@@ -61,23 +65,23 @@ def cache_search_results_file(cached_filename: str, sr: SolrResponse, rows=0, fi
             csv_file.write(u'\N{BOM}' + data)  
 
         ######  @TODO  - Need to remove header_file immediately
-
         
         # Use a CSV writer with forced quoting for the body of the file
         with open(cached_filename, 'a', newline='', encoding='utf8') as csv_file:
             cache_writer = csv.writer(csv_file, dialect='excel', quoting=csv.QUOTE_ALL)
             c = 0
-            for i in sr.docs:
-                if rows != 0 and c > rows:
+            for doc in sr.docs:
+                if rows and c >= rows:
                     break
-                try:
-                    for j in i.keys():
-                        if isinstance(i[j], list):
-                            i[j] = ",".join(i[j])
-                    cache_writer.writerow(i.values())
-                    c += 1
-                except UnicodeEncodeError:
-                    pass
+
+                values = []
+                for colname in headers:
+                    value = doc.get(colname, "")
+                    if isinstance(value, list):
+                        value = ",".join(str(v) for v in value)
+                    values.append(value)
+                cache_writer.writerow(values)
+                c += 1
     return True
 
 ## OPEN-4055: Accept a list of Fields for the exported search results